Product Personality

Transcription

Product Personality

Product Personality
ISBN 90-77595-46-5
© Pascalle C.M. Govers 2004
All rights reserved. No part of this book may be reproduced or transmitted in any form
or by any means electronical or mechanical, including photocopying, recording or by any
other storage and retrieval system without permission from the author.
Product Personality
Proefschrift
Ter verkrijging van de graad van doctor
aan de Technische Universiteit Delft,
op gezag van de Rector Magnificus prof. dr. ir. J.T. Fokkema,
voorzitter van het College voor Promoties,
in het openbaar te verdedigen op dinsdag 15 juni 2004 om 15.30 uur
door Pascalle Cornelia Maria GOVERS
doctorandus in de psychologie
geboren te Goirle.
Dit proefschrift is goedgekeurd door de promotoren:
Prof. dr. J.P.L. Schoormans
Prof. dr. P.P.M. Hekkert
Prof. dr. W.M. Oppedijk van Veen
Samenstelling promotiecommissie:
Rector Magnificus, voorzitter
Prof. dr. J.P.L. Schoormans, Technische Universiteit Delft, promotor
Prof. dr. P.P.M. Hekkert, Technische Universiteit Delft, promotor
Prof. dr. W.M. Oppedijk van Veen, Technische Universiteit Delft, promotor
Prof. dr. ir. A.C.J.M. Eekhout, Technische Universiteit Delft
Prof. dr. S.M.J. van Osselaer, Erasmus Universiteit Rotterdam
Prof. dr. B. de Raad, Rijks Universiteit Groningen
Prof. dr. W.F. van Raaij, Universiteit van Tilburg
5
Contents
Acknowledgements
8
1.
Introducing Product Personality
11
1.1
1.2
1.3
1.4
1.5
1.6
1.7
Personality characteristics and products
Previous research
Defining product personality
Products and product personality
Contribution of product personality to product design
The contribution of this thesis: three research issues
Overview of the thesis
11
12
15
16
18
19
20
2.
Describing Products as People
22
2.1 The symbolic meaning of products
2.1.1 Products as symbols within a social context
2.1.2 Products as symbols of the self-concept
22
23
24
2.2 Human personality
2.2.1 Consistency of human personality
2.2.2 Five factors of human personality
26
27
28
2.3 Person perception
2.3.1 The influence of the perceiver: accessibility effects
2.3.2 The influence of the target-person: the role of appearance
29
31
33
2.4 Discussion and implications regarding the research issues
2.4.1 Implications for the perception of product personality
2.4.2 Implications for the influence on consumer preference
2.4.3 Implications for the assessment of product personality
35
35
36
37
3.
38
The Influence of Appearance on the Perception of
Product Personality
3.1 Product appearance
38
3.2 Differentiating between product variants
3.2.1 Method
3.2.2 Results
3.2.3 Discussion
40
40
41
44
3.3 Designing product personality
3.3.1 Method
3.3.2 Results
3.3.3 Discussion
46
48
50
53
3.4 Conclusions
55
6
4.
Product-Personality Congruence and Consumer Preference
57
4.1 The similarity-attraction relationship
57
4.2 The influence of product personality in a pre-purchase situation
4.2.1 Method
4.2.2 Results
4.2.3 Discussion
59
60
65
67
4.3 The influence of product personality in a post-purchase situation
4.3.1 Method study 1
4.3.2 Results study 1
4.3.5 Discussion
69
72
75
77
79
80
4.3 Conclusions
81
5.
Assessment of Personality
83
5.1
5.2
5.3
5.4
Assessment of human personality
Assessment of product personality
Scales used to assess the personality of products
Conclusions
83
85
86
89
6.
Development of a Product Personality Scale
91
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
General steps in scale development
Steps in developing a product personality scale
Format of measurement
Personality characteristics from existing measures
Personality characteristics from qualitative research
Reviewing the items with respect to definition
Grouping together items with the same meaning
Selection of items to describe both people and products
Selection of items to form the concept scale
6.9.1 Method
6.9.2 Results
6.9.3 Discussion
91
93
95
96
98
99
101
103
107
107
109
113
6.10
Conclusions
115
7.
Testing the Product Personality Scale
119
7.1
7.2
7.3
The final step in scale development
Differences in the interpretation of items
Describing the lexical meaning of the items
119
120
121
7
7.4
Visualizing the meaning of the items
7.4.1 Development of the visuals
7.4.2 Validation of the visuals
124
124
127
7.5
Testing the product personality scale
7.5.5 Discussion
132
133
134
138
139
141
7.6
Conclusions
142
8.
Discussion and Implications
144
8.1
8.2
8.3
8.4
8.5
8.6
8.7
Product personality
The symbolic meaning of product variants
Perception of product personality
The influence of product personality on consumer preference
Actionability of product personality
Suggestions for future research
144
145
147
151
154
157
159
Summary
163
Samenvatting
166
References
169
Appendices:
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
Appendix G
Appendix H
Appendix J
Appendix K
Appendix L
Appendix M
Curriculum Vitae
183
184
186
187
189
194
198
208
209
213
215
221
223
8
Acknowledgments
Writing this thesis is something I had never imagined myself doing. Before I
started as a Ph.D. student, I didn’t think science was for me. However, once I had started
my Ph.D. I found that I felt very at home. I loved the puzzling, trying to get to the core
of things and creating new ways to study them. I have always been intrigued by the
“irrational” side of consumer behavior, and this Ph.D. provided me a perfect opportunity
to making it my daily work. Nevertheless, doing a Ph.D. proved to be a challenge, during
which many people stimulated, helped, and supported me.
First, there is my (ex-)colleagues at PIM. Everybody in the department, thank
you for being such fine colleagues! It was nice to work in the friendly and stimulating
atmosphere that you constitute together. Amina, Erik, Niels, Serge, and Sylvia thank you
for all the effort you put in the process of reducing 1200 personality descriptors! Dirk,
you and your knowledge of literature as well as your book collection always inspired me.
Sandra and Karin, thank you for your day-to-day help. Karin, I enjoyed our many talks
and I hope you enjoy your new job as well as your own home. Agnes, your help as
manager of the PEL has been essential. Gitty, thank you for all the jobs you did for me
during the first years. Hadewig, Jasper, Marieke, Marjolein, Mayke, Roos, Yvonne, thank
you for being so enthusiastic about my research that you dedicated your time and effort
to it. It was stimulating to work with you. Dicky, Gert, Ianus, Jeroen, Lucia, Madelon,
Mirjam, Pieter, Remco, Rudolf, Thomas, and Lucienne thank you for your time and input
in gathering the pictures. Lucienne, I also want to thank you for your hospitality.
Next to practical help, many people supported me in other ways. Amina, Dirk,
Erik, Erik-Jan, Friedo, Jan, Kaj, Karin, Mariel, Niels, Ruth, Serge, Sylvia, I loved our
lunches and our walks afterwards. Marja, thank you for the fun talks and support during
the hard times in the first year. Amina, Erik, Maaike, Niels, Ruth, and Serge, you starting
your Ph.D.’s encouraged me in doing mine. Sharing the process of doing a Ph.D. and
everything surrounding it, has been important to me. Niels, I have had a great time
sharing a room with you. You taught me to be more pragmatic. Ruth, our conscientious
nature makes us good working partners. I hope we will continue to work together in the
future. Dirk, I miss my car-pool partner!
I am also greatly indebted to my promoters, Jan, Paul and Walle. Thank you for
convincing me to do this Ph.D. project. You believed that I could do this even when I
didn’t. I learned a lot from all of you. Jan, it was a great pleasure to work with you. One
of the best things about you is that you always put people first. I think this makes you a
great person. Your involvement and belief in me, together with your interest in my
research, have been a great stimulation. You teach without lecturing. You made me
9
explore new ideas at times that I would have found a dead end had I continued focusing
on details. More importantly, you taught me to put it all in perspective. Thank you! Paul,
you were the one that continued to put the design perspective under my attention. The
discussions we had about which step to be taken next, or what way to go now, challenged
me to go beyond my own insights. Thank you for your input, it helped shaping this
thesis. Walle, unfortunately your role changed after the first year of my Ph.D. Instead of
being prominently present, you became a (very important!) background force. I could
always count on you and you were there at every important crossroad. Your involvement
and support have been an essential motivating force. I am grateful that you are one of my
promoters.
Finally, I want to put in a word to my family and friends. First, thank you all for
listening to me and stimulating me. You helped me achieving this! Gerard & Ria, I am
very grateful to be a part of your family. Jeroen, the memory of you making fun of my
books and my eagerness to learn nowadays makes me smile. You saw the scientist in me
before I even considered going to University! I know that you are proud, and I am proud
that it is you who stands besides to me during the defense of this thesis. Dear Christel, it
is good to have you in our family. Ernest; in the last ten years you have become a close
friend and I am very happy that you agreed to be my paranimf. I appreciate our talks and
discussions about life, love, religion and other important issues. Hopefully, many others
will follow. Rijnie (and Naomi), you stepped into this process when it was already
running. Yet, you supported me with great compassion. It feels good that you are here to
share in the ‘great final’. Dearest Kees & Lia, you were the ones that never seemed to get
tired of my research, who were always prepared to think with me in order to find
solutions, and who comforted me when I was about to give up. Your love and belief in
me helped me through some tough times. I am proud to share this accomplishment with
you. I know that you are proud too. I also know that you love me for just being me, and
that is the most wonderful feeling! My dear Fred, I know that the stress of me writing my
thesis also put some weight on your shoulders. The wonderful thing is that though you
got tired of it sometimes, you never stopped supporting me. I am grateful that I could do
this with you next to me, the feeling that we share makes me stronger. You promised me
you would wait for me at the finish line, and you did. I love you, thank you for loving (all
of) me.
Pascalle Govers
10
1
“The friendliest toilet paper” (Dutch magazine advertisement, 2001)
1.1
Personality characteristics and products
Personality characteristics are those words or descriptions that describe “stable
and durable, non-physical qualities of a person on which he/she discriminates him/herself from others”
(Doddema & De Raad, 1997, p. xiv). Among other things, people use personality
characteristics to describe their first impression of another person. The first impression
people form of other people is important because it influences the likelihood of a relation
(Zebrowitz, 1990). In analogy, the first impression of products is important because it
influences the likelihood of purchase (Van Raaij, Antonides, Oppedijk van Veen &
Schoormans, 1999). This thesis studies how people describe this impression of products
with personality characteristics and how it affects their consumer behavior.
The word “personality” in product personality refers to the fact that people use
human personality characteristics to describe their impression of a product. Like people,
products have stable and durable qualities that discriminate them from other products.
Product personality refers to the set of personality characteristics that people use to
describe a specific product variant and to discriminate it from others. The “non-physical”
in the definition of human personality characteristics can be translated into “intangible”
when used with respect to products. The qualities of a product that are described with
personality characteristics cannot be reduced to a single tangible attribute of that product.
Product personality is a high-level description of the total product. It is what designers
refer to as “character” (Janlert & Stolterman, 1997). The fact that the toilet paper in the
advertisement shown at the start of this chapter is described as friendly does not refer to
a specific attribute. It refers to the sensory characteristics (it feels soft), the quality
12
characteristics (the paper is strong), and the decorations (small flowers). The totality of
the product makes it friendly.
An important difference between human personality and product personality is
that product personality is more explicitly related to the product exterior. Human
personality describes the internal qualities that guide a persons’ behavior. Product
personality on the other hand describes more external qualities. It describes the overall
impression that a product makes upon the consumer (see also chapter 2).
1.2
Previous research
The concept of product personality fits in the tradition of symbolic
consumption. The idea that products can serve as symbols is at least a century old
(Veblen, 1899) and might even be traced back before 1834 (Leibenstein, 1950). The
majority of literature on this topic is developed in the second half of the 20th century after
Levy (1959) and Goffman (1951, 1959) indicated to marketers and social scientists that
products have a symbolic value. The range of topics studied since is very broad, varying
from the cultural meaning of products (e.g. Kleine, Kleine & Kernan, 1993; McCracken,
1986), semiotics of consumption (e.g. Holman, 1981; Mick, 1986), products as tools for
self-expression (e.g. Belk, 1988; Prentice, 1987), and impression formation based on
possessions (e.g. Belk, 1978; Dittmar & Pepper, 1994; Gosling, Ko, Morris & Mannarelli,
2002).
Part of this literature is concerned with the use of personality characteristics to
describe non-human entities. It indicates that people not only use personality
characteristics to describe and discriminate between people. They also use them with
respect to brands (Aaker, 1997, Biel, 1993), stores (d’Astous & Lévesque, 2003; Sirgy,
Grewal & Mangleburg, 2000), products (Janlert & Stolterman, 1997; Jordan, 1997, 2000;
Levy, 1959; Sirgy, 1982) and computers (Nass, Moon, Fogg, Reeves & Dryer, 1995).
Brand personality refers to the set of personality characteristics associated with a
brand (Aaker, 1997). It is considered an important way to differentiate a brand in an
otherwise equal product class (Biel, 1993). Aaker defines brand personality as “the set of
human characteristics associated with a brand” (p. 347). Brand personality is considered to be a
multi-dimensional construct consisting of five factors: (1) sincerity, (2) excitement, (3)
competence, (4) sophistication and (5) ruggedness. Recent studies have shown that the
content and the amount of brand personality dimensions appear to depend on culture.
There seem to be some global dimensions: sincerity, excitement, and sophistication, and
some culturally specific ones: peacefulness (Japan) and passion (Spain) (Aaker, BenetMartínez, & Garolera, 2001). Dutch research into brand personality provided seven
13
dimensions, three of which are considered specific for The Netherlands: (1) gentle, (2)
annoying, and (3) differentiating. Three of the seven dimensions resemble the global
dimensions and one dimension is shared with the United States (ruggedness) (Smit & Van
den Berg, 2002).
The use of personality characteristics to describe retail stores is referred to as the
retail-patron image. The retail-patron image reflects the stereotypical image of the
clientele of a store (Sirgy, Grewal & Mangleburg, 2000) and is a concept from congruity
theory (Sirgy, 1982). It is the equivalent of product-user image that reflects the
stereotypical image of users of a product class or a brand (Sirgy, Grewal, Mangleburg,
Park, Chon, Claiborne, Johar & Berkman, 1997; Sirgy et al., 2000). Congruity theory
shows that a match between the product-user image and the consumers’ self-concept
positively influences the attitude towards that product. This self-congruity effect is
extended to retail stores, where a match between the retail-patron image with the
consumers’ self-concept has a positive effect on retail patronage (Sirgy et al., 2000).
Recently a store-personality scale has been developed (d’Astous & Lévesque, 2003) that,
like the brand personality scale of Aaker (1997), consists of five dimensions. The five
dimensions of store personality are: (1) sophistication, (2) solidity, (3) genuineness, (4)
enthusiasm and (5) unpleasantness. Store personality as measured by this scale is defined
as “the mental representation of a store on dimensions that typically capture a individual’s personality”
(d’Astous & Lévesque, 2003, p. 457). It is considered to be inferred from sources like
advertising, sales personnel and shoppers.
An assumption that is present in most of the previously mentioned studies is that
the impression of a product or a store as described by personality characteristics, actually
describes the user or shopper. Both the retail-patron image and the product-user image of
self-congruity theory are considered to be a reflection of the image of the shopper or the
user. Brand personality is said to describe the “user component of brand image” (Biel, 1993, p.
71). Figure 1.2 shows an example of an advertisement where these concepts come
together. The Mexx advertisements show attractive, caring and successful women (their
personalities are described in the ad) wearing Mexx jewels or sunglasses. These two
women communicate the user image or brand personality of Mexx. W omen who identify
with these women are expected to prefer Mexx products.
Brand personality has also been defined as reflecting the brand itself (Aaker,
1997; Biel, 1993). In that meaning, it resembles the concept of product personality. In
product personality the personality characteristics are used to describe the product itself.
This use of personality characteristics to describe the impression of the product itself has
not yet been widely studied. Some have studied the use of personality characteristics to
describe the impression of the product itself, such as Janlert and Stolterman (1997), and
Jordan (1997, 2000). Janlert and Stolterman wrote a conceptual paper about “the
14
character of things”, and Jordan conducted two applied studies asking people to think
about products as if they were people and to assign human personality characteristics to
them.
Figure 1.1: Advertisements for Mexx jewels (left) and sunglasses (right)
A special case of using personality characteristics with respect to products is
described by Nass et al. (1995). Nass et al. tested the hypothesis that people use different
personality characteristics to describe computers depending on their behavior. They
developed computers that interacted with the user in either a submissive or dominant
manner, by manipulating the style of communication of the computer. The results
showed that people indeed differentiated between the computers based on their behavior.
In this case personality characteristics were used to describe the product itself. However,
they referred to an evaluation of an agent inside the computer. For this reason, it falls
outside the scope of this thesis.
In conclusion, product personality is different from existing concepts because it
refers to the product itself and not to its users. Though it resembles brand personality,
product personality describes a single product and not a brand. This distinction is
relevant because brands feature different product variants (see section 1.4), and different
product variants from the same global brand can have different product personalities. For
example, the results of one of our studies (see section 6.9) showed that two Siemens’
vacuum cleaners are described with different personality characteristics (see figure 1.2).
Happy
Conscientious
Friendly
Cute
15
Conscientious
Masculine
Serious
Honest
Figure 1.2: Two Siemens vacuum cleaners with different product personalities
1.3
Defining Product Personality
Product personality refers to the profile of personality characteristics that people
use to describe a specific product variant and to discriminate it from others. Product
personality is an overall description of a single product and is strongly influenced by a
product’s appearance (Janlert & Stolterman, 1997). For example, the appearance of a
product variant can make the product look tough or happy. However, product
personality does not only refer to personality characteristics used to describe the humanlike features that are explicitly incorporated in products. The fact that the products in
figure 1.3 can be described as cheerful is partly due to the incorporation of a smile into
the appearance of the products, which is an aspect of cheerful people. Product
personality comprehends more, it describes the overall impression of a product. The use
of personality characteristics to describe the overall impression of a product is
demonstrated in figure 1.4. In these examples personality characteristics are used to
communicate both the style of the appearance and the non-visible attributes of a product.
This is the use of personality characteristics with respect to products that we refer to as
product personality.
Figure 1.3: Smiling products
16
“Stylish, smart and sexy….”
“…a television without frills….
modest, neat, and trendy….”
“…friendly and inviting,
though at the same time robust and sporty..”
Figure 1.4: Personality characteristics used with respect to products in advertisements.
1.4
Products and product personality
Products encompass anything that can be offered to satisfy a need or a want.
They can be classified into three groups based on their durability and tangibility (see table
1.1). Fast-moving products are tangible products that are consumed in one use or a few
uses, such as shampoo, soft drinks, and perfume. Durable products are tangible products
that are used over an extended period of time and that normally survive many uses, such
as cars, coffeemakers and mobile phones. Services are intangible products, for example:
healthcare, insurance, and advice (Kotler, 1997). We restricted the research on product
personality to durable products used by consumers1. This means that in this thesis the
term “product(s)” is used to represent only consumer durables.
1
The focus on consumer durables is based on the fact that the field of Industrial Design
Engineering is predominantly concerned with durable consumer products.
17
Table 1.1
Classification of products based on tangibility and durability
Tangible
Intangible
Durable
Durables
Services
Non-durable
Fast-moving
Since product personality describes the product itself, we studied durable
consumer products at the lowest level of the product hierarchy, the level of product
variants (see table 1.2; Kotler, 1997). The use of this level means that the yellow and the
blue Siemens vacuum cleaner, shown in figure 1.2, are seen as two different products.
Table 1.2
Seven levels of the product hierarchy (Kotler, 1997)
Name
Description
1. Need family
The core need that underlies the existence of a product family.
2. Product family
All the product classes that can satisfy a core need with
reasonable effectiveness.
3. Product class
A group of products within the product family recognized as
having a certain functional coherence.
4. Product line
A group of products within a product class that are closely
related because they perform a similar function.
5. Product type
A group of items within a product line that share one of several
possible forms of the product.
6. Brand
The name, associated with one or more items in the product
line that is used to identify the source or character of the items.
7. Product variant
A distinct unit within a brand or product line that is
distinguishable by size, price or appearance.
A final comment about the products used in product personality research
concerns their nature. Studies on symbolic consumption have primarily concentrated on
publicly consumed, high status products. The product type most frequently used as the
stimulus product is a passenger car (Malhotra, 1988). However, as is suggested by Kleine,
Kleine and Kernan (1993) the ordinary products we use in everyday life also have
symbolic meaning. Moreover, product personality is supposed to discriminate between
the range of product variants offered to consumers. There is a great variety in product
variants in both high status and ordinary product classes. Thus, product personality is
believed to be relevant to all durable consumer products. As a consequence, this thesis
includes both high status products and ordinary product classes.
18
1.5
The contribution of product personality to product design
Most people do not consciously think about how they choose among the
diversity of products that is offered to them. If asked, they will probably say that they
consider price, quality and functionality. Nevertheless, beyond price, quality and
functionality there are symbolic reasons that guide consumer behavior. People use
products to communicate their identity, status, and their membership of social groups.
Psychology and consumer research have a history of studying the symbolic
meaning of products. The primary goal of both of these fields is to understand human
behavior. Therefore, most of the research is not directed at products. Products merely
serve as stimuli. As a consequence, the knowledge gained from these fields of study is too
general to be applied in product design. A designer may know that people use products as
symbols to communicate their identity, but this knowledge does not provide insight into
how (s)he can design a product with a certain identity. What should a “serious” and
“honest” product look like? It is important to know the answer to this question because
the symbolic meaning of products has become increasingly important. Nowadays,
differentiating products based on their technical functions or quality is difficult
(Dumaine, 1991; Veryzer, 1995). Since the wave of the quality controls in the 1980-ies,
products can be expected to fulfill their functions reasonably well. Symbolic meaning
provides another way to differentiate products, or as a designer put it in an interview:
“designers can visualize the values that a product or organization represents and are therefore capable of
differentiating them” (Bosveld & Van de Kolk, 2003).
However, the question remains how this symbolic meaning can be incorporated
in product development. Literature addresses this problem as one of actionability.
Actionability of product attributes means, “that the attributes indicate specific actions the
manufacturers must take to build such a product” (Shocker & Srinivasan, 1974, p. 922). This
thesis addresses the actionability of symbolic product meaning by addressing the
actionability of product personality.
The contribution of product personality lies in the fact that product personality
describes the impression of a single product variant. Information about how product
personality is perceived and which product personality is preferred provides designers
with a basis for adapting the products that they design to the wishes of the consumer.
Product design works at the level of product variants. Designers create product variants,
not product classes or brands. They need information at this specific level. Studying
product personality contributes to product design because it provides knowledge at the
level of product design.
1.6
19
The contribution of this thesis: three research issues
The aim of this thesis is to raise the actionability of product personality by
addressing three issues: (1) the perception of product personality, (2) the influence of
product personality on consumer preference, and (3) systematic assessment of product
personality. These issues are addressed by applying knowledge from psychology, about
humans and human interaction, to products and human-product interaction.
The perception of product personality
The first issue that this thesis will address is the perception of product
personality. In order to establish whether product personality is indeed a concept worth
studying, we first need to find out whether the assumption holds that people describe
products using personality characteristics. The first question to be answered is: Do people
perceive product personality? If the answer is yes, then the question becomes: How do
people perceive product personality?
Based on literature, product personality was predicted to be a concept strongly
influenced by the product’s appearance and used by consumers to discriminate between
product variants. Two empirical studies are reported that confirm the assumption that
people ascribe human personality characteristics to products. It appears that people
differentiate product variants based on product personality. These studies also show that,
like in person perception, the appearance of a product is a strong determinant of
perceived product personality.
The second issue that will be addressed is the influence of product personality on
consumer preference. Knowing that people use product personality to distinguish
between product variants, we set out to investigate how product personality influences
consumer preference. To make using product personality in product design worthwhile, it
should not only be used to describe products, it should also influence consumer
preference.
We assumed that, since it is known that people evaluate other people more
positively when they perceive them as similar to themselves, people might also evaluate
products more positively when they perceive them as similar to themselves. Studies in
consumer research have also reported findings that support this assumption. We
conducted two studies to investigate the influence of this similarity-attraction relationship
with respect to product personality and largely found confirming evidence. Similarity
between a persons’ personality or self-concept and the product’s personality positively
20
influenced people’s preference for a product variant in a pre-purchase as well as in a postpurchase situation.
Systematic assessment of product personality
The fact that product personality influences consumer preference makes it
relevant to assess it. Moreover, since the appearance of a product is a determinant of
product personality, it is possible to create products with a particular product personality.
This is an attractive strategy because it seems that people prefer products with a
personality that is similar to their own. In order to design products with a pre-determined
personality that is perceived correctly by consumers, we have to know what appearance
characteristics consumers associate with a particular personality characteristic. Systematic
assessment of product personality provides a tool to find that out.
Systematic assessment of product personality involves rating the personality of
many product variants according to many people in a way in which the results are
comparable. To this end a product personality scale was developed. Theory about
personality assessment and scale development is used as a base for developing this
product personality scale. Several studies are conducted to gather and select items for the
scale. A final product personality scale consisting of 18 items resulted. The items contain
verbal personality characteristics extended with lexical descriptions and visuals.
1.7
Overview of the thesis
The thesis consists of eight chapters. Chapter two and three address the
perception of product personality. Chapter four is concerned with the influence of
product personality on consumer preference. Chapter five, six and seven describe the
development of a product personality scale. Chapter eight closes with a general discussion
of the findings and implications for future research.
The first chapter following this introduction is chapter two, a theoretical chapter
that gives an overview of the relevant literature from psychology and consumer research.
It re-formulates the ideas from these fields of study to the context of products and
product personality and describes the implications this has for the three research issues
described above.
Chapter three investigates the influence of product appearance on the perception
of product personality. Analysis of free descriptions of watches showed that there is a
relationship between product appearance and product personality. A second study
reported in chapter three, investigates whether this relation can be used to design product
21
personality. The results suggest that consumers indeed recognize a pre-determined
product personality based on product appearance.
Chapter four is concerned with the influence of product personality on
consumer preference. Two studies are reported, which indicate that the effect of product
personality on consumer preference can be predicted from principles of human
interaction. The similarity-attraction principle is tested in a pre-purchase and post
purchase situation. In both cases similarity between the consumer and the product
personality has a positive effect on consumer preference.
Chapter five is a short theoretical account of human personality assessment and
product personality assessment. The conclusion of this chapter is that none of the
existing measures can be used to validly assess product personality. If we want to assess
product personality we will have to develop a scale that is especially attuned to this goal.
Chapter six describes the development of this product personality scale. It
describes the selection of the relevant items. The chapter starts with the gathering of
personality characteristics from literature and qualitative research, and subsequently
reduces them to a relevant and manageable set of 20 items according to general steps in
scale development.
Chapter seven tests the concept scale of product personality that results from
chapter six. First the items are further developed and then they are tested with respect to
their validity and reliability. An 18-item product personality scale resulted. The items exist
of verbal personality characteristics that are extended with lexical descriptions and visuals.
Chapter eight provides a general discussion of the complete thesis, and concludes
with some implications regarding the three research issues.
2
2.1
The symbolic meaning of products
Many products have a significance that goes beyond their functional utility. This
significance stems from the ability of products to communicate meaning (Hirschman,
1981; McCracken, 1986). Products are symbols by which people convey something about
themselves to themselves and to others (Holman, 1981; Solomon, 1983). This symbolic
meaning influences consumer behavior or consumer preference. “All commercial objects have
a symbolic character, and making a purchase involves an assessment – implicit or explicit – of this
symbolism….” (Levy, 1959, p.119).
Research about the symbolic meaning of products in consumer behavior has
been developed from different perspectives. On the one hand, the symbolic meaning of
products is studied from a social perspective; people are members of social groups and
they use products as tools to communicate membership or distinction of these groups
(e.g. McCracken, 1986; Solomon, 1983). On the other hand, the symbolic meaning of
products has been studied from a more individualistic perspective. This perspective of the
symbolic meaning of products focuses on products as symbols of the self-concept (e.g.
Belk, 1988; Kleine, Kleine & Kernan, 1993; Sirgy, 1982).
2.1.1
23
Products as symbols within a social context
The symbolic meaning of products from the social perspective can be seen to
fulfill two major functions that stem from two human motives. The first function is to
gain belonging to and acceptance of a social group. This function is motivated by the
need for (love and) belongingness (Maslov, 1943). The second function relates to
distinction within this group motivated by the need for uniqueness (Fromkin & Snyder,
1980; Tepper Tian, Bearden & Hunter, 2001).
The need for belongingness is rooted in a fundamental fear of isolation and
loneliness. People are social beings that seek the company of other people. They desire a
place in their social group and they need to feel accepted by its members (Maslov, 1943).
The need to belong is served by conformity. When people dress according to fashion or
buy the latest mobile phone, they communicate that they are part of a group. This need is
referred to in common language as “keeping up with the Joneses”. In consumer behavior
it is referred to as the bandwagon effect. The bandwagon effect describes the desire to
purchase something only because many other people have purchased it. “It represents the
desire of people to purchase a commodity in order to […] conform with the people they wish to be
associated with” (Leibenstein, 1950, p. 189).
Next to the need for belongingness, people also have a need for uniqueness.
People want to belong to a social group but they also strive to distinguish themselves
within this group, they want to be positively distinct from other people (Fromkin &
Snyder, 1980; Tepper Tian, Bearden & Hunter, 2001). This need for uniqueness is
comparable to Maslov’s self-esteem needs. Maslov (1943) reasoned that people, once they
feel that they belong, also desire for a stable, high evaluation of themselves. According to
Csikszentmihalyi (2000) this need motivates people to “…purchase goods that show our
uniqueness, and separate us from the rest of the crowd” (p. 269). The need for uniqueness
influences consumer behavior because people try to achieve a sense of personal identity
by contrasting their consumption pattern to that of others (Csikszentmihalyi, 2000;
Thompson & Haytko, 1997). The need for uniqueness as a motive in consumer behavior
is recognized in the snob effect. The snob effect refers to the phenomenon that the
desirability of a product decreases because others are consuming the same product. “It
represents the desire of people to be exclusive; to be different; to dissociate themselves from the common
herd” (Leibenstein, 1950, p. 189).
A related concept is the Veblen effect. The Veblen effect refers to the effect that
some products become more in demand when they have a higher, instead of a lower price
(Leibenstein, 1950). Veblen (1899) introduced the term conspicuous consumption to
refer to the consumption of this kind of products by the upper class aimed to
demonstrate wealth, showing that one can afford luxury. Once common people started to
24
buy the same products it was no longer exclusive (the price lowered). The upper class
would then find a new way of distinguishing themselves, showing other rare and
expensive possessions. This Veblen effect can be motivated by both of the previously
discussed human motives. It can be motivated by the need to belong. People want to be a
member of and be accepted by the upper class. But at the same time the Veblen effect
can be motivated by the need to be unique. By buying expensive products people want to
communicate that they are different from (and better than) other people.
2.1.2
Products as symbols of the self-concept
The individualistic oriented study of the symbolic meaning of products has
largely focused on products as symbols of the self-concept. The self-concept is defined as
“the totality of an individual’s thoughts and feelings having reference to himself as an object”
(Rosenberg, 1979, p. 7). Belk (1988) has argued that the products that we own contribute
to and reflect our identities. He reports evidence that our products are part of who we are
and how we see ourselves.
An important source of confusion with respect to the self-concept is the use of
the terms “self” and “self-concept” as synonyms. A concept is an idea about something;
the entity to which the self-concept refers is the self (Baumeester, 1998). Self-concept thus
is the picture of the self (Rosenberg, 1979).
Everybody has a self-concept and with the self-concept come the motives of
self-consistency and self-enhancement (Baumeester, 1998; Rosenberg, 1979). Selfconsistency refers to the motive to confirm one’s self-concept (Swann, Stein-Seroussi &
Giesler, 1992) and to protect it against change (Rosenberg, 1979). Self-enhancement is
also called self-esteem and refers to the desire for information that reinforces a positive,
self-concept (Baumeester, 1998). Maslov (1943) referred to these motives as the esteem
motives. People have a desire for a stable (self-consistency) and high (self-enhancement)
evaluation of themselves.
Self-consistency is seen to influence consumer preference. Consumers prefer
products associated with an image that is consistent with their self-concept (e.g. Belk,
1988; Levy, 1959; Malhotra, 1988; Sirgy, 1982). Sirgy (1982) suggested in his selfcongruity theory that consumers make a psychological comparison between their selfconcept and the stereotype of the general user of a product. If a person identifies with
this product-user image (s)he will experience high self-congruence which positively
influences product evaluation. Studies concerning this self-congruity hypothesis often
report confirming evidence (Dolich, 1969; Ericksen & Sirgy, 1989, 1992; Grubb & Hupp,
25
1968; Heath & Scott, 1998; Hong & Zinkhan, 1995; Landon, 1974; Malhotra, 1988; Sirgy,
1985). Aaker (1999) also found a congruity effect for brand personality.
Self-enhancement encourages consumers to prefer products with a positive
image and to avoid products with a negative image (Sirgy, 1982). Theoretically, it could
occur that a person prefers a product with a negative image because it confirms his/her
negative self-concept, yet will reject it because the negative image conflicts with the selfenhancement motive. There is some discussion about which motive is stronger
(Baumeester, 1998; Swann, Stein-Seroussi & Giesler, 1992). Since (healthy) people
generally have a positive self-concept, the motives of self-consistency and selfenhancement are rarely in conflict (Rosenberg, 1979). As a consequence, the influence of
self-consistency upon consumer behavior is often reinforced by the influence of selfenhancement.
“You are what you drive”
“Cars differentiate themselves
by means of character and
expression.”
(Metro, June 22nd 1999)
“Mobile phone
shows who you are”
“…..People use their
mobile phones, consciously
or unconsciously to show
their personality.”
(Metro, October 19th 2001)
“Your home is who you are!”
“…..how you decorate your
home tells everything about
your personality…..”
(Flair, January 29th 2002)
Figure 2.1: Examples in the media on the use of products as symbols of personality.
26
It may be concluded from the preceding sections that people want to belong to a
social group, yet at the same time want to positively distinguish themselves in a manner
that is consistent with their self-concept. Product personality allows people to do just
that. For example, a straightforward person may want products without frills. An
extroverted person may choose colorful, eye-catching product variants. Figure 2.1 shows
three examples that confirm that people show who they are through the choice of a
particular product variant.
2.2
Human personality
The term personality can be used to mean two different things. One meaning of
personality is that it refers to a quality that some people have and that others do not have.
It refers to the presence of a certain temperament that makes a person interesting or
attractive (Oxford Dictionary, 2000). In this meaning one can say that “Robin has a lot of
personality”, “Brain lacks personality”, and “Joe has a great personality”. It is in this
meaning that schools advertise courses promising “more personality” to those who
enroll.
The other meaning of personality is the meaning of personality that is used in
personality psychology. It is based on the idea that everybody has a (unique) personality
that differentiates him/her from others. Within personality psychology there is not a
universally accepted definition of personality, “in fact, there may be as many different definitions
of personality as there are theorists who tried to define it” (Hjelle & Ziegler, 1981, p. 6). Despite
substantial differences, the various personality definitions have some common aspects. A
first aspect that is common to all personality theories is that personality is viewed as being
concerned with individual differences. Every person has a unique personality. No two
persons are exactly alike in temperament or behavior (Carver & Scheier, 1996; Hjelle &
Ziegler, 1981; Murphy & Davidshofer, 1994; Pervin, 1975). Another aspect that
personality definitions have in common is that they view personality as the cause of
consistency in behavior. Personality is something within the person causing individuals to
behave consistently in different situations (Carver & Scheier, 1996; Murphy &
Davidshofer, 1994). Finally, personality is seen as an overall description of a person. It is
an abstraction based on information about a person’s behavior, thoughts and feelings
(Carver & Scheier, 1996; Hjelle & Ziegler, 1981; Murphy & Davidshofer, 1994).
This conception of personality in personality psychology resembles the
conception of product personality in the current research in that both view personality as
an overall description and both deal with individual differences. Human personality
27
describes people, and differentiates between them, whereas product personality describes
products, and differentiates between those.
2.2.1
Consistency of human personality
A key assumption in personality psychology is that personality is stable and
influences a person’s behavior in a wide range of situations. This assumption of
consistency has been the base of considerable debate since Mischel (1968) showed that
people generally show far less cross-situational consistency in their behavior than had
been assumed. It appeared that the correlation between personality ratings and
personality related behavior (e.g. generous and donating money to charity) averaged
around .30 (Mischel, 1979). This is very modest because it means that personality
accounted for less than 10% of the variation in behavior, leaving 90% unaccounted for.
Currently, personality psychologists seem to have reached a consensus that
people show a relative, rather than an absolute, consistency in their behavior (Pervin,
1985). Relative consistency means that the behavior of one individual may differ over
different occasions, yet the behavioral differences between individuals that vary on a
certain personality characteristic remain constant, both over time and in varying situations
(Schmitt & Borkenau, 1992). For example, if one person is out-going and another person
is introverted, the behavior of the first individual may be outgoing when (s)he is at a
party, but less so when (s)he is at a business-meeting (resulting in the low correlations
reported by Mischel). This out-going person should however remain more out-going than
the other person in both situations (resulting in relative consistency). Moreover, Shoda
and Mischel (2000) showed that behavior that seems inconsistent at first, might show a
pattern that can be considered a stable profile of personality. For example, a child may
not react aggressively when teased by other children, yet may become very aggressive
when warned by adults. Though this behavior is not consistent in terms of having an
aggressive personality (not aggressive in one situation and very aggressive in another), the
profile of aggressive behavior (not aggressive when teased, but very aggressive when
warned) can be consistent.
Like human personality, product personality is assumed to be stable. Different
people, at different times are expected to describe the same product variant with similar
personality characteristics. Nevertheless, the impression people form of a product might
also be influenced by the situation in which it is perceived. Although important, this
influence is not part of this thesis. We will focus on the perception of product personality
without the influence of the situation. Once we understand that, future research can
investigate the role of the situation.
28
2.2.2
Five factors of human personality
After the crisis caused by Mischel (1968), the field of personality psychology was
reenergized in the early 1980’s by the emergence of a five-factor model of personality
(Goldberg, 1981). It appeared that the basic personality characteristics could be described
by five superordinate factors, which are referred to as either the Big Five (Goldberg,
1981), or the five-factor model (McCrae & Costa, 1996). These five factors are necessary
and sufficient to represent human personality and are highly robust. They have been
found in various age groups, in different linguistic groups, and in different cultural
groups. They have also be replicated using different methods of measurement (Digman,
1990). Table 2.1 shows them arranged in order of magnitude of explained variance
accounted for in various personality ratings.
Factor
1.
Table 2.1
The five-factor model of human personality
Description
Extroversion
People high on this factor are spontaneous,
assertive, talkative, and active.
2. Agreeableness
People high on this factor are good-natured, polite,
considerate, and supportive.
3. Conscientiousness
People high on this factor are neat, serious,
ambitious, and precise.
4. Neuroticism
People high on this factor are nervous, anxious, and
high-strung.
5. Openness to experience
People high on this factor are original, curious,
intellectual, and open-minded.
The first factor “extroversion” is generally seen as similar to Eysenck’s
introversion/extroversion dimension (1947), representing a tendency towards sociability
and an open expression of impulses. The second factor “agreeableness” incorporates a
sense of warmth and emotional supportiveness together with a kind of compliance. The
third factor “conscientiousness” represents qualities like responsibility and correctness,
but also a purposeful striving towards goals. The fourth factor “neuroticism” again
represents one of Eysenck’s dimensions, namely emotional stability/neuroticism. The
essence of this factor seems to be the experience of anxiety. The fifth and last factor
“openness to experience” merges a sense of intelligence or intellect with various aspects
29
of openness (openness to new ideas, flexibility of thought and readiness to indulge in
fantasy (Carver & Scheier, 1996; Digman, 1990).
These five factors represent the most abstract level of the personality taxonomy.
The five-factor model is an integrative framework that merges several personality
characteristics into higher order dimensions. Each of the five factors incorporates typical
personality characteristics, which incorporate typical behaviors or habits. These behaviors
and habits in their turn incorporate specific responses (Digman, 1990). The
superordinate, five-factor structure is considered to be too abstract for use with respect to
product personality level. Because of the high level of abstraction, the five factors may
not differentiate between product variants. For example, if one product variant is
perceived as cheerful and the other is perceived as outspoken, both can be described as
extravert.
The issue of losing discriminating power when higher order dimensions are used
is also relevant to human personality (Carver & Scheier, 1996). The superordinate level
does not serve to discriminate between people, it serves as a summary of human
personality in general. The five-factor structure is derived from personality characteristics
used in ordinary language (McCrae, Costa & Piedmont, 1993). Arguing that the most
prominent differences between people would have found their way to ordinary language,
human personality was studied by looking at how people describe the personality of other
people (e.g. Goldberg 1981). Likewise, we will study product personality by first studying
the way in which people describe products.
2.3
Person perception
Next to personality psychology, social psychology is another field that is
concerned with describing people and differentiating between them. The main difference
between social psychology and personality psychology is that social psychology does not
investigate personality as an internal causal force. Instead of studying personality as an
independent variable, social psychology studies personality as a dependent variable
(Pervin & John, 1999). Social psychology does not study how personality affects
behavior, but how personality is perceived, and which factors influence this perception.
The field of study within social psychology that we are interested in here is person
perception. Person or social perception studies the way in which lay-people conceive,
understand and interpret other people and form a personality impression of them
(Gilbert, 1998). Person perception is relevant to product personality because product
personality is concerned with the way in which consumers perceive the personality of a
product. Knowledge about the impression formation process with respect to people may
30
provide useful insights with respect to how consumers form personality impressions of
products.
Research in person perception revolves around two variables, the perceiver and
the target person (the object of perception). Research into the influence of these two
variables on person perception derives from different theoretical backgrounds. Research
on the influence of the perceiver is dominated by the constructivist approach to
perception, sometimes called theory-driven perception. This approach states that
perceptions are structured in the mind of the perceiver. It emphasizes subjective
perceptions and investigates how the perceivers’ internal, mental structures, called
schemas, influence person perception (Zebrowitz, 1990). For example, it has been shown
that assimilation and contrast effects with respect to activated schemas influence
perception. Therapists who expected to interview a patient, and thus activated the
“patient schema” assimilated the information of a target person into this schema. The
target person was perceived as more disturbed by these therapists than when (s)he was
evaluated by therapists who were expecting to interview a job applicant (Langer &
Abelson, 1974). If that what is perceived strongly conflicts with what is expected, then a
contrast effect will occur. For example, the male stereotype includes competence. Males
who display the conflicting behavior, i.e. incompetence, are contrasted to this stereotype
and, as a result, are rated more negatively than equally incompetent females (Deaux &
Taynor, 1973). In this case, the target person is not included into an active schema (the
male stereotype) but the behavior is measured against the stereotype as if it is a normative
standard (Stapel & Koomen, 1998). With respect to products, this would mean that the
activity of a schema also influences the perception of products.
Research on the influence of the other variable in person perception, the targetperson, originates from a structuralist approach to perception. The basic assumption in
this approach is that perceptions derive from elementary sensations. Perception is
considered to be data-driven and “objective”. Research in this area concentrates on
identifying the aspects of the target-person influencing person perception. Characteristics
of the target person that influence person perception include both the person’s behavior
and appearance (Zebrowitz, 1990). For example, a person that helps an elderly woman
across the street is perceived as helpful, and a person wearing glasses is perceived as more
intelligent than a person without glasses. Though behavior is important, the impact of
appearance is hard to overestimate. Many experimental results testify to the strong impact
of physical appearance features to the perception of a person’s personality (Jones, 1990).
2.3.1
31
Influence of the perceiver
Various characteristics of the perceiver influence his or her impression of the
target person, including his/her cognitive schemas, affective and demographic factors
(Zebrowitz, 1990).
Cognitive schemas
Perceivers have knowledge about people in general and use this knowledge in
developing impressions of specific target persons. This pre-existing knowledge is
organized and structured in “cognitive schemas” (Vonk, 1990). Cognitive schemas are
defined as cognitive structures that represent organized knowledge about a given concept
(Fiske & Taylor, 1984). Schemas that contain knowledge about personality are called
“implicit personality theories” (Vonk, 1999) and include (a) the stable attributes of
persons and (b) the expected or inferred associations among these attributes (Vonk,
1990). For example, an implicit personality theory may associate the concepts
independent, self-confident and persistent with each other, and dissociate them from
insecure and irresolute. As a result of these implicit personality theories the perceiver can
“go beyond” the properties of the person perceived (Bruner, 1957). Based on what is
observed people make predictions about what is not perceived. This way, they can make
inferences about the person’s underlying personality characteristics (Park & Judd, 1989).
Based on their implicit personality theory, people may infer that a person who is
independent is also self-confident and persistent, but not insecure or irresolute.
One way in which cognitive schemas influence person perception is through
priming or accessibility effects. When a schema or construct is primed or made accessible,
perceived information can be more easily interpreted in accordance with this schema.
Accessibility of a schema means that there is only a minimum of input necessary for
something to become assimilated into this schema. The more accessible a schema, the
sooner something will be interpreted in terms of that schema and the more stimuli will be
assimilated into it, causing other (more relevant) schemas to be masked (Bruner, 1957).
Thus, as the accessibility of schemas in memory increases, the more easily new stimuli will
be interpreted in accordance with these schemas (Higgins, 1996). Schemas and constructs
become accessible due to recent use and frequent usage. Recent use makes a schema
temporarily more accessible and frequent use makes it chronically accessible (Vonk,
1999).
32
Stereotypes
Stereotypes are a special case of cognitive schemas influencing person
perception. Stereotypes are defined as cognitive schemas of a particular social group and
its members (e.g. women, beta students, blacks, Muslims) (Augoustinos & Walker, 1995).
Stereotypes influence person perception because people categorize other people as
members of certain social groups, and subsequently apply their implicit personality
theories to these individuals. For example, in a study by Duncan (1976) perceivers
watched a videotape of two people engaged in a heated discussion, which culminated in
one person pushing the other. White perceivers were much more likely to perceive the
“pusher” as violent when he was black, than when he was white, in which case he was
perceived as “playing around”. Duncan concluded that this effect was consistent with the
(1976) American stereotype that blacks were more aggressive than whites.
A stereotype can only influence person perception if the perceiver is aware of the
distinction and the stereotype. If the perceiver is unaware of a stereotypical distinction
between beta and alpha students then these stereotypes cannot influence his or her
perception. Moreover, a stereotype needs to be activated. A person has to be recognized
and categorized as member of a particular social group for a stereotype to be applied.
Affective factors
Among the affective factors that influence person perception are the perceiver’s
goals and emotions. Goals will determine which information is attended to and how this
information is interpreted (Jones, 1990). For example, the short-term goal of a date
positively influenced a target-person’s perceived personality. When men were committed
to taking a woman out on a date they formed a more uniformly positive personality
impression of her than when they were not (Zebrowitz, 1990). The perceiver’s goals not
only bias his evaluation but also influence to which information (s)he attends. If a
person’s most important value is honesty (s)he will pay special attention to cues related to
honesty and integrity. On the other hand, if a person’s most important concern is
“getting ahead in the world” (s)he will be especially alert to symbols of power and status
(Jones, 1990).
Person perception can also be biased by the perceiver’s emotional state. Several
studies have documented the influence of moods on judgments of the attractiveness of
others. People are perceived to be more attractive by those whose mood has been
elevated and less attractive by those whose mood has been depressed (Zebrowitz, 1990).
33
Demographic factors
Although there is little research concerning the influence of demographic factors
on the perception of other people, it seems reasonable to assume that a perceiver’s age,
gender and cultural background will have a significant effect on person perception by
influencing cognitive factors (Zebrowitz, 1990). This influence may be exerted via the
accessibility of schemas and traits. The perceiver’s cognitive schemas and stereotypes, as
well as the perceivers goals, may each covary with age, gender and culture (Zebrowitz,
1990).
2.3.2
Influence of the target-person: the role of appearance
Our perception of other people is not only determined by the perceiver’s
cognitive schemas, affective factors, and demographic characteristics. It is also
determined by characteristics of the target person (Zebrowitz, 1990). Even though we
like to think otherwise, the impact of appearance is hard to overestimate. Many
experimental results testify to the strong impact of physical appearance features to the
perception of a person’s personality (Jones, 1990). For example, if different perceivers
only have visual appearance to base their judgments on (silent film or photographs), their
first impressions of a target person agree with each other beyond chance level (Albright,
Kenny & Malloy, 1988; Borkenau & Liebler, 1992a).
This section discusses the influence of the visual appearance characteristics of a
target person on person perception. The visual appearance characteristics can be divided
into visual-static or visual-dynamic characteristics (Borkenau & Liebler, 1992a).
Visual-static characteristics
The visual static characteristics of a target person that influence person
perception include stature, clothing and facial characteristics. “Stature” has two meanings.
In the first meaning it is used to refer to the importance and respect that persons have
because of their ability and achievements. In the second meaning it refers to a person’s
height (Oxford Dictionary, 2000). This double meaning is reflected in the impression of
short and tall people. Tall stature is related to perceived status and dominance
(Zebrowitz, 1990), the higher the status of the target person the greater the perceived
height (Stewart, Powell & Chetwynd, 1979). Lower status people are perceived to be
shorter (Macrae, Stangor & Hewstone, 1996). Notably, most of the American presidents
have been several inches longer than the norm for their times.
Clothing leads to value judgments and expectations about social class, occupation
and income (Stewart, Powell & Chetwynd, 1979). Holman (1980) has conducted an
34
experimental study using photographs of women dressed in several different clothing
ensembles, and found differences in perceived sexiness, fashionability and
masculinity/femininity. Other studies that show “way of dress” influencing inferences
about personality traits (Borkenau & Liebler, 1992a, 1992b, 1995) indicate that a showy
dress correlates positively with perceived extroversion, and negatively with agreeableness
and conscientiousness. Formal dress correlates positively with conscientiousness.
With regard to facial qualities, a diversity of features has been studied as
determinants of perceived personality, such as: beardedness, hair color, make up and the
wearing of glasses. All of these are shown to affect the perceived personality (Cook, 1979;
see also Borkenau & Liebler, 1992a, 1992b, 1995). A well-documented effect of the
target’s face on perceived personality is babyishness. People with babyish facial features
such as large round eyes, a small nose, a large forehead or a small chin are perceived as
physically weaker, more submissive, more naïve, warmer and more honest than maturefaced people of the same age, sex and attractiveness (Zebrowitz, 1990). On the other
hand people with a mature face are perceived as conscientious (Borkenau & Liebler,
1992a).
Visual-dynamic characteristics
The visual-dynamic characteristics that influence person perception can be
further subdivided into gait, mimic and gestures. The influence of gait information on
impressions has been studied using “point-light”. This technique has been employed to
study reactions to a person’s gait, independently of other aspects of appearance. The
pattern of a person walking is represented as small luminous dots placed on the joints,
moving across a black background. Results of this kind of research indicate that gait
influences the impression of the walker’s personality. For example, youthful walking
adults are perceived as happier and more powerful than their older walking peers
(Montepare & Zebrowitz-McArthur, 1988). Using silent film, Borkenau and Liebler
(1992a) also found that “way of sitting” was one of the variables affecting impressions of
people. Where “controlled way of sitting” correlated with impressions of
conscientiousness, “relaxed way of sitting” correlated with impressions of emotional
stability.
With respect to mimic, Borkenau and Liebler (1992a, 1992b, 1995) studied the
influence of the extent of smiling and facial expression (either friendly, unconcerned or
self-assured) on strangers’ impressions. They found that extroversion is related to a
friendly and self-assured expression, accompanied by extensive smiling. The impression
of agreeableness is more related to a friendly expression rather than a self-assured
expression. In the case of emotional stability, the relation is the other way around.
35
Emotional stability is more strongly related to a self-assured expression than to a friendly
expression.
Factors relating to gestures in the same studies show the same kind of results.
Frequent and fast movements are positively related to extroversion, not related to
emotional stability and negatively related to conscientiousness.
Summarizing the preceding, it can be said that physical appearance factors have a
strong impact on the perception of a person’s personality. In analogy with person
perception we assume that product appearance is a major determinant in the perception
of product personality.
2.4
Implications regarding the three research issues
This section summarizes the described literature and discusses the implications
for the three research issues as formulated in section 1.5. The first section discusses the
implications for the perception of product personality when insights from person
perception are applied to product personality perception. The second section discusses
the implications for the influence of product personality on consumer preference. The
third section discusses implications for the systematic assessment of product personality.
2.4.1
Implications for the perception of product personality
Like human personality, product personality is assumed to be relatively stable
over time and in varying situations. Although context might influence the perception of
product personality, it will not be part of this thesis. Neither are the aspects of the
perceiver that could influence the perception of product personality. The goal of this
thesis is to start raising the actionability of product personality, where actionability refers
to, “… the specific actions the manufacturers must take …” (Shocker & Srinivasan, 1974. p. 922).
Insight into how context or perceiver characteristics influence the perception of product
personality would not increase actionability because both fall outside the actions a
manufacturer can take. The only aspect that falls within the scope of a manufacturer is
the product itself. Product designers influence the product’s characteristics. Therefore,
the focus of this thesis will be on how product characteristics influence the perception of
Another argument for a focus on product characteristics is our position in the
debate about perception being a reflection of reality (structuralist approach) or in “the eye
of the beholder” (constructuralist approach). As with almost all modern theories, we
36
assume that elements of both approaches interact to form perception. Though perception
is grounded in external stimuli, it cannot be reduced to individual elements of the target.
Perception is also influenced by peoples’ cognitive structures. It is however not
completely constructed in the mind of the perceiver (Zebrowitz, 1990). With the
knowledge that both the perceiver and target influence perception, we are left with the
question of the relative importance of the two variables. If perception is strongly
influenced by characteristics of the perceiver, consensus among perceivers should be low
(Vonk & Heiser, 1991). If perception is mainly a function of stimulus’ characteristics,
consensus among perceivers should be high (Kenny, Albright, Malloy & Kashy, 1994).
Since we are studying the perception of product personality within one country (The
Netherlands), we believe that the consensus among perceivers will be reasonably high,
thereby justifying a focus on target characteristics. This does not mean, however, that
given a particular product variant we assume different people will form a similar product
personality impression. We believe the effect of differences between the products to be
stronger than the effect of differences between perceivers. The question of how
differences between perceivers influence product personality is left open for future
research.
From the product characteristics that might influence how product personality is
perceived, we concentrate on the role of appearance characteristics. People that differ in
their appearance are perceived to have different personalities. Analogously, products that
differ in their appearance are also expected to have different personalities. The visual
appearance of a product includes all the product characteristics that people can perceive
by looking at a product. This includes the product’s shape, texture and color (Creusen,
1998). All these appearance aspects influence a product’s personality. A variation of one
of these aspects may already lead to a change in a product’s personality. If products have
completely different appearances they probably also have very different personalities.
That is to say, the more products differ in appearance the more their personalities are
expected to differ.
2.4.2
Implications for the influence on consumer preference
Implications regarding the influence of product personality on consumer
preference mostly stem from the literature about the symbolic meaning of products.
Literature suggests that people may use product personality to express a consistent and
positive view of themselves. As a consequence, people prefer product variants with a
personality that is consistent with their self-concept. The concepts of brand personality
(Aaker, 1997) and product-user image (Sirgy, 1982) also describe this aspect of consumer
37
behavior but they do not explain symbolic consumer behavior at the level of product
variants. For example, why would a person prefer one Siemens vacuum cleaner over the
other (see figure 1.2)? Product personality differs per product variant and describes part
of the symbolic meaning related to the product itself. Product personality can thus
explain why a consumer prefers the happy and cute vacuum cleaner instead of the serious
and honest one, even though these two products have a similar product-user image and a
similar brand personality.
Although product personality is a concept that works at a lower level of
abstraction than the known concepts, the mechanisms that influence preference are
expected to remain the same. People are expected to prefer products with a product
personality that is similar to their self-concept. This effect is also known in human
interaction as the similarity-attraction effect. People are attracted to people who are
similar to themselves. Both a review of this specific literature and some hypotheses about
a similarity-attraction relationship between consumers and products are discussed in
chapter four.
2.4.3
Implications for the assessment of product personality
The implications for the assessment of product personality mainly originate from
personality psychology. The definition of human personality resembles the
conceptualization of product personality in two important ways. First, both consider
personality to be an overall description. Secondly, both concepts are used to describe
individual differences. The assumption that these individual differences can be measured
is present in all theories of personality (Pervin, 1975) and it also applies here. Personality
assessment is a way to gain information about a person. In analogy, product personality
should be assessed in order to gain information about product personality and how
product personality relates to product characteristic. The main difference with the
measurement of human personality is that we will not be looking for higher order
dimensions that summarize product personality, but rather for a profile of everyday
personality adjectives that can be used to describe and compare the personality
impression of consumer durables. More relevant literature with respect to personality
assessment will be discussed in chapter five. Chapter six and seven report the steps taken
to develop a product personality scale.
3
The Influence of Appearance on the
Perception of Product Personality
“Sportiness is neither in the length nor in the thickness of your spoiler”
(Saab magazine advertisement, 2003)
3.1
Product appearance
The appearance of a product includes all the product characteristics that people
can perceive by looking at a product (Creusen, 1998). It has also been referred to as
product design, product form or product shape (e.g. Berkowitz, 1987; Bloch, 1995;
Kotler & Rath, 1983). We will use the term product appearance in order to distinguish
product appearance from product design as an activity within the product development
process (Roozenburg & Eekels, 1998).
Product appearance is considered to be a powerful tool to gain competitive
advantage in the market place (Berkowitz, 1987; Bruce & Whitehead, 1988; Kotler &
Rath, 1983; Trueman & Jobber, 1998; White, Salter & Gann, 2003). Variety in appearance
between product variants is a tool for differentiation. Product variants are offered in
different sizes, colors, and/or shapes. These differences create the needed differentiation
for competitive advantage (Berkowitz, 1987). Competitive advantage is accomplished if a
product variant differs from the other variants in such a way that it will better fit
consumer needs and requirements than others (Bruce & Whitehead, 1988). The role of
product appearance is also present in its symbolic meaning. People seek to express their
status and to convey specific images through the appearance attributes of the products
they purchase (Kotler & Rath, 1983).
Differentiation through variety in product appearance is not unique to current
time. Montgomery Ward and Co, a mail order company, already offered a range of 131
variants of pocket knifes in 1895 (see example in figure 3.1).
The Influence of Appearance on the Perception of Product Personality
39
Figure 3.1: Pocket knifes from Montgomery Ward & Co, Catalogue no. 57, 1895, p. 440.
According to Forty (1995), it was already quite common in the nineteenth century to
offer products in such a range of variants. Forty argues that this variety “gave customers a
degree of choice and enabled them to feel more sure of their own individuality. A masculine-looking pocket
knife might underline the purchaser’s view of himself as manly, but as long as it was the only men’s knife
available it would do nothing to make him feel different from other men. What would do this would be the
opportunity to choose from a range of knives or to have a particular design what he alone among his
acquaintances might possess” (p. 87). It may thus be concluded that product appearance helps
to differentiate between product variants. By creating choice, companies allow the
consumers to choose the product variant that best fits his/her needs. As such, companies
distinguish their products from the competitor and gain consumer preference.
It is hypothesized that the variety in appearance can be described by personality
characteristics. Products that differ in their appearance are expected to have a different
product personality. The current chapter reports two empirical studies testing this
prediction. The first study investigates whether the differentiation between product
variants based on appearance leads to differences in the perception of product personality
(section 3.2). The second study investigates whether differences in product personality
can be designed through differences in product appearance (section 3.3).
40
3.2
Differentiating between product variants2
This section reports a preliminary study investigating whether people use
personality characteristics to describe products and, if so, whether these personality
characteristics distinguish between different product variants from the same product type.
In this study we investigated the differentiation of wristwatches based on the personality
characteristics people used to describe the watches and we also explored the influence of
product appearance. If product personality is influenced by product appearance, then
differentiation between product variants based on personality characteristics should result
in differentiation in product appearance. We chose to use watches as stimuli because the
use of watches made the study more accessible. Watches are a highly personalized
product type and often seen as part of ones “self” (Belk, 1988). As a consequence, people
may more easily describe watches using personality characteristics, than they would some
other, more utilitarian product types. An additional advantage of choosing watches as
stimuli is that watches are offered in a wide variety of appearances.
Since part of the goal of this study was to explore the influence of differences in
product appearance on product personality, we preferred a large stimulus set with a
variety in appearance, while functional differences were minimal. This was easier with
(wrist)watches than it might have been with many other durable products.
Twelve respondents were asked to describe their impression of a set of 30
watches. Their responses were content analyzed and the answers that contain personality
related characteristics were selected. Subsequently, a homogeneity analysis was performed
on the personality characteristics and the watches in order to study the differential value
of the personality characteristics.
3.2.1
Method
Respondents
Twelve respondents (7 males and 5 females) were randomly selected from the
consumer household panel of the Faculty of Industrial Design Engineering of Delft
University of Technology. The consumer panel contains an a-select sample of the
population of Delft and the surrounding area (see Tan, 2002). Their age varied from 27 to
40 years, with an average of 33. Participation was voluntary. Respondents were rewarded
with a financial compensation. They received a written debriefing afterwards.
2
With some alterations, the study in this section has been published previously in Govers and
Schoormans (2000).
41
Stimuli
Based on an analogy with person perception, we expected that differences in
product appearance would generate differences in product personality. We therefore
created a stimulus-set of 30 wristwatches with a large variety in appearances.
Variety in product appearance includes more than design variety within a product
class, it also includes visual functional differences. For example, a coffeemaker does not
only have a different appearance than a screwdriver, it also has a different function. Due
to this difference in function, every coffeemaker has an inherently different product
personality than any screwdriver. Since we did not want this kind of (functional)
appearance difference to affect our results, our stimuli were all from the same product
type: wristwatches. All stimuli provided the consumer with the same standard function.
Color pictures of the wristwatches served as the stimuli in the study. The pictures are
represented in appendix A.
Procedure
During a face-to-face interview, the 30 color pictures were shown to the
respondents one at a time. Respondents were instructed to describe “their impression” of
each watch. We used a free-response format in order not to force respondents into the
use of personality characteristics. The interviewer wrote down the words and phrases the
respondent used to describe each watch. Before turning to the next stimulus, the
interviewer read the answer the respondent had given out loud so that additions or
alterations could be made. The complete interview lasted approximately half-an-hour and
was taped on video.
3.2.2
Results
A simple count of the answers leads to a total of 766 verbal descriptions given by
the 12 respondents. A verbal description is a word or phrase used to describe (at least)
one of the stimuli. These verbal descriptions were content analyzed by two independent
judges (both female, ages 23 and 25). The judges were instructed to group together the
verbal descriptions they considered as describing the same aspects (e.g. sporty, sporting,
sportsmen, and gentleman, manly, masculine, businessman). Each judge first grouped the
verbal descriptions individually. Afterwards, they discussed their solutions until consensus
was reached. This way, the data set was reduced to 68 categories containing all the verbal
descriptions. An overview of the categories is shown in appendix B.
The 68 categories differed in the number of verbal descriptions that they
encompassed. Some verbal descriptions were very idiosyncratic in that they were only
42
given once. These verbal descriptions could not be categorized with other verbal
descriptions. They were each treated as a separate category (15 categories, 2% of the total
number of verbal descriptions). Other categories were more common and included up to
48 verbal descriptions (e.g. “sporty”, 6% of the total number of verbal descriptions). The
nature of the categories also differed. Some categories described tangible attributes of
watches, such as “gold” or “digital”. Other categories described more intangible aspects,
such as “eccentric” or “boring”.
In order to find out whether the respondents used personality characteristics to
describe the watches, we asked five other independent judges to evaluate the 68
categories. Two of them were male, and three were female. Their ages ranged from 22 to
25, with an average of 23. They were asked to classify the categories according to the
definition of personality characteristics based on Doddema and De Raad (1997): “A
human characteristic used to differentiate between people excluding physical characteristics” (p. xiv).
They were given the list of category names and were instructed to indicate which ones
they thought described a personality characteristic according to the given definition.
Nineteen categories were judged by the majority of the judges (>3) to describe a
personality characteristic (ri = .39 and α = .76). These categories account for 36%
(274/766) of the original verbal descriptions and are presented in figure 3.2 (in appendix
B they are depicted with a *). Figure 3.2 also indicates the frequency of use of each of the
personality characteristics.
Figure 3.2: The 19 categories describing a personality characteristic and the number of
times they were used to describe the stimuli.
43
The results show that the frequency of use differs per personality characteristic. Some
personality characteristics (e.g. “sociable”, “boring”) are rarely used to describe this set of
watches. Others (e.g. “sporty”, “girlish”) are more commonly used.
It appears that one third of the verbal descriptions is personality related. This
confirms our expectation that people use personality characteristics to describe products.
However, the use of personality characteristics does not necessarily originate from
differences in product appearance. The concept of product personality includes the
assumption that people use personality characteristics to discriminate between product
variants with a different appearance. We conducted a homogeneity analysis (HOMALS)
in order to find out whether the personality characteristics used to describe the watches
also discriminate between the watches.
HOMALS describes the relationship between two or more nominal variables in a
low-dimensional space. This space contains the variable categories (the personality
characteristics), as well as the objects (the watches). The personality characteristics that
are often used together will be close to each other and to the watches they describe.
Watches that are described using different characteristics are further apart. A first
HOMALS analysis, which included the 19 categories that are considered personality
characteristics (the ones in figure 3.2), showed that two characteristics, “formal” and “not
sporty”, did not discriminate between the watches. They had very low (< .1) discriminant
measures on both dimensions (Van den Berg, 1987). A second HOMALS analysis was
conducted without these two characteristics. This resulted in a two-dimensional solution
that explained about 80% of the variance (fit = .79). Figure 3.3 represents this solution.
awkward
girlish
eccentric
childish
conspicuous
boyish
sporty
peculiar
informal
tough
masculine
skilful
feminine
oldfashioned
plain
Figure 3.3: Personality characteristics and watches
boring
44
The results show that it is possible to distinguish three groups of personality
characteristics and watches. The circles drawn in figure 3.2 indicate these groups. The
three groups more or less represent masculine3, feminine and youthful watches. The
masculine and youthful watches are being more conspicuous than the feminine watches.
The masculine watches are robust and big with generally more additional functions than
the other watches. The masculine watches that are least robust and sober in functions are
the ones closest to the feminine watches. The results also indicate that these watches are
also considered to be more boring than the others. The youthful watches are colorful and
have unusual shapes, but generally seem to have less added functionalities (e.g. stopwatch,
alarm) than the masculine watches. Immature maybe a better description for this group.
The watches in this group are not only childish and girlish, but also awkward and
peculiar. The feminine watches are refined, chrome colored and offer the consumer only
the basic functionality (indicating time). These watches are not only described as feminine
but also as old-fashioned.
Twenty-one watches can be said to fall into these three groups, which is two
third of the total stimulus set that contained 30 watches. These watches are often
described using these personality characteristics and can be seen as visual examples of the
personality characteristics surrounding them. Nine watches fall outside the three groups.
This means that the respondents either did not describe these watches with the selected
characteristics, or did not agree about the use of these personality characteristics for these
watches. Post-hoc analysis of the watches that fell outside the HOMALS solution
indicates that these watches were frequently described as “chic” (“chic” in Dutch) and
“sleek” (“strak” in Dutch). Neither of these descriptions is categorized as personality
characteristics, which explains why the watches described and distinguished on these
terms are not in the HOMALS solution.
3.2.3
Discussion
The goal of this study was threefold. First, we wanted to explore whether people
use personality characteristics to describe product variants. Second, we wanted to see
whether these personality descriptions discriminate between the product variants. Third,
we wanted to explore if there was a relation between product appearance and the
personality characteristics that people use to describe them.
The results suggest that a reasonable proportion of the initial verbal descriptions
of watches, gathered in a free-description task, is indeed personality related. The
3
Due to limited space, three watches from the “masculine” group are not displayed.
45
proportion of personality related answers (30%) is quite remarkable. Research in person
perception has shown that personality characteristics “… make up only a little more than 40
per cent of Western adults’ open-ended person descriptions…” (Zebrowitz, 1990, p. 15). This
means that the amount of personality characteristics in a free description of products is
not very different from a free description of people. Moreover, like personality
characteristics that are used to describe people, personality characteristics that are used to
describe watches also distinguish the watches. Three groups of watches can be
distinguished based on the personality descriptions, each group with its own appearance
characteristics. Because of the similarities in appearance of the watches within a group we
may conclude that product personality is influenced by the product’s appearance. Yet, as
suggested, the “design variety” between the watches is not the only aspect that seems to
influence the association of personality characteristics with the watches. Small functional
differences also influence the perception of product personality. The appearance of the
masculine watches, for example, suggests additional features that are not present in the
feminine watches. These functional differences are part of the product’s appearance.
Product appearance includes the functional differences that are visible, such as additional
buttons or clocks in the case of watches. As a whole this influences the perception of
In conclusion, the results of this preliminary study suggest that people use
personality characteristics to describe products, and to discriminate between product
variants that vary in appearance. Together, these results provide a sufficient base to
investigate the concept of product personality and the influence of product appearance
more extensively.
Limitations
The results of this study are based on a small sample and one product type:
watches. This may be seen as a shortcoming, especially since watches are a special case of
consumer durables. Watches, like clothing and shoes, are part of a persons’ appearance. It
is a highly personalized product type often seen as part of ones “self” (Belk, 1988). As a
consequence, people may more easily describe watches with personality characteristics
than they will describe other product types. However, it is for this reason that we
considered watches a good product type to start our exploration into product personality.
The fact that people use personality characteristics to describe watches does not prove
that people generally use personality characteristics to describe durable consumer goods,
but it is a first indication. Future research will have to look into the use of personality
characteristics with respect to other, more utilitarian product types, and try to replicate
these findings with a larger sample.
46
Another limitation of this study is that we used free descriptions. Free
descriptions were seen as the best format for discovering whether people would use
personality characteristics to describe product variants. Asking respondents to rate the
product variants on several personality characteristics would not have provided an answer
to this question. However, due to the free-description procedure we cannot be
completely certain about the differences in perceived product personality that resulted
from this study. The fact that a personality characteristic appears close to a specific watch
and not to another is because it was only mentioned with respect to that particular watch.
This does not mean, however, that the personality characteristic is not descriptive of the
other stimulus. It simply was not mentioned in the free descriptions.
A final limitation of this study is the fact that we used pictures of real life
products and did not remove brands or set price limits. As a consequence, the impression
of the product variants might also have been influenced by other aspects, such as brand
associations and anticipated price. However, the results do show that there is a relation
between product personality and product appearance. Watches that are described using
the same personality characteristics have comparable appearance features. This may imply
designers can use product appearance to create a certain product personality. If products
that are robust are consistently described as masculine, one can create a masculine
product by designing a robust product variant.
3.3
Designing product personality4
Now that we have established an empirical base for the existence of product
personality that suggests a relation between product personality and product appearance,
we set out to further investigate the relation between product personality and product
appearance. The assumed relation with appearance is not unique to products and product
personality. From person perception we know that the impression other people form of a
person’s personality is strongly influenced by their appearance (see section 2.4.2). People
can use this to “manage” the impression they want to make. Whether a person is going to
a job interview or to a nightclub will have consequences for the impression (s)he wants to
make and will probably result in a different “look” for that person. People use their
appearance (and behavior) to create a particular image of themselves (Brehm & Kassin,
1990). Designers use a similar tactic. They use the appearance of products as a tool to
create products with a certain expression.
4
With some alterations, the study in this section has been published previously in Govers,
Hekkert and Schoormans (2004).
47
Designers use the relation between impression and appearance by manipulating
the product appearance, trying to evoke certain beliefs in the consumer (Janlert &
Stolterman, 1997). Designers make choices regarding appearance characteristics, such as
shape, proportion, material, color and texture, and decide how to mix these elements
(Bloch, 1995). In making these decisions, they create a product with a certain personality.
However, in order to create products with a certain personality designers should
translate an intangible concept into visual and material features in such a way that it is
recognizable to other people. If consumers do not recognize a robust product variant as
tough, the communication of tough through product appearance is ineffective. There are
many examples in daily life that support the suggestion that consumers generally
recognize a pre-determined personality. Levi’s, for example, changed their product
portfolio to be more innovative. As a result the brand is now (again) perceived as
“young”, “cool” and “trendy” (Van Buiten & Jans, 2002).
Smets, Overbeeke and Gaver (1994) indicate that designers can create a
recognizable, pre-determined impression through product appearance. Design students
created dessert packages designed to express a particular taste, portable cassette players
designed to express a particular music style, and sculptures designed to express a
particular scent. Subsequently, respondents had to choose which one of three designs
(one correct and two distracters) best represented the original concepts. In all three cases
the proportion of correct choices exceeded chance level. The respondents recognized the
original taste, music style or scent in the designs. In another study, the participants
(design students) were asked to design a pager that would evoke sensual feelings
(Hofmeester, Kemp & Blankendaal, 1996). Ratings of the two “sensual” pagers were
compared to the ratings of a reference model. The results show that the pagers designed
to express sensuality are indeed rated higher on adjectives such as “erotic”, “organic”,
“strokeable” and “sensual”. In both of these studies the impression of the designed
products was tested with respondents that have a design background. This can be
considered to be a limitation, since it appears that consumers and designers perceive
product design differently (Hsu, Chuang, & Chuang, 2000). In order to test whether
consumers understand the impression of the product variant as it was intended, it is
recommended to use respondents that are naïve with respect to product design.
The study described in this section investigates whether designers can create a
predetermined product personality that consumers can recognize. Stimuli are created to
express a certain personality characteristic. Respondents are then asked to rate the stimuli
products (irons) on these personality characteristics (happy, cute or tough). We expect
that respondents will recognize the personality characteristic that is used as a basis for the
design. We have, therefore, formulated the following hypothesis:
48
Hypothesis
Respondents rate irons designed to be happy [cute] {tough} as happier [cuter] {tougher}
than irons not intended to be happy [cute] {tough}.
3.3.1
Method
Respondents
Eighty-eight respondents (46 males and 42 females) were randomly selected from
the consumer household panel of the Faculty of Industrial Design Engineering of Delft
University of Technology. The consumer panel contains an a-select sample of the
population of Delft and the surrounding area (see Tan, 2002). Ages of the respondents
ranged from 19 to 74 years, with an average age of 48. Participation was voluntary.
Respondents were rewarded with a financial compensation.
Stimuli
Eighteen graduate students of Industrial Design Engineering (11 males and 7
females) specially created the stimuli for this study. The students were given 45 minutes
to sketch an iron expressing a particular personality characteristic; happy, cute or tough.
The personality characteristics happy, cute and tough were selected because they can all
be applied to people as well as products (see section 6.8). Each student received a written
instruction and a picture of a prototype iron with the functions that had to be
incorporated into the design (see figure 3.4). The instruction emphasized that it should be
possible to use the iron, but that they did not need to worry about manufacturing
problems. They were not allowed to use symbols or icons and only had a black fine-liner
and a grey marker at their disposal.
Switch
Handle
Temperature
Metal base
Flex
Water reservoir
Figure 3.4: The prototype iron
After they had finished sketching, the students filled out a questionnaire asking
to explain their design choices (with respect to product form) and to indicate which
49
colors and textures they would have preferred for their design. It took the students
approximately 15 minutes to fill out the complete questionnaire. Participation was
voluntary and the students were not paid. Students could sign up to participate via a
registration form on a notice board.
This task resulted in a set of 18 black-and-white sketches of irons, consisting of
six “happy” irons, six “cute” irons and six “tough” irons (see appendix C). A pre-test (n =
10) showed that when respondents had to rate all 18 irons, they lost interest about
halfway through and started to fill out the questions arbitrarily (e.g. all similar answers).
The answers about the last stimuli were, therefore, unreliable. Due to this fact, we
randomly selected 9 sketches to serve as stimuli (see figure 3.5). For every personality
characteristic, we randomly took three of the six irons designed to express this
characteristic.
Happy
Cute
Tough
A
B
C
Figure 3.5: The stimuli irons in order of personality characteristic
In the remainder of the text we will refer to the individual irons that are designed to be
happy as happy iron A, happy iron B, and happy iron C (presented from top to bottom in
figure 3.5). A similar notation will be used to refer to the irons designed to be cute and
the irons designed to be tough.
50
Procedure
The nine selected irons were arranged in three random orders, resulting in three
different versions of the questionnaire. Respondents rated each iron on the three target
characteristics, “happy” (“blij” in Dutch), “cute” (“schattig” in Dutch) and “tough” (“stoer”
in Dutch). It took the respondents approximately 10 minutes to rate the nine irons.
Measures
All three characteristics were measured using 5-point scales (1 = “not
happy”/“not cute”/“not tough” and 5 = “happy”/“cute”/“tough”). Instead of matching
data as Smets et al. (1994) did, we used rating scales such as used by Hofmeester et al.
(1996). We viewed this as a stronger indication of recognition. Smets et al. asked their
respondents (design students) to choose which one of three designs best represented a
goal characteristic. We asked our respondents to rate the stimuli on the same personality
characteristics that were used as a basis for the designs.
3.3.2
Results
For use in the analysis, we calculated a group mean for the three irons that represent each
characteristic. Table 3.1 illustrates the individual and group means of the irons for the
three target characteristics. The results show that especially the tough irons score high
means for their target characteristic, both individually and as a group. The cute irons have
the lowest means (both the individual means and the group mean) for their target
characteristic. The mean scores of the happy irons (the individual and group means) for
their target characteristic all approximate three. The difference between the individual
means is smallest for the happy irons (maximum difference .35). The individual means of
the happy irons do not differ significantly from each other (F(2, 261) = 2.16, n.s.). The
maximal difference between the individual means for the cute irons is .79 and .75 for the
tough irons. These individual means do differ significantly from each other (cute: F(2,
261) = 9.1, p < .001; tough: F(2, 261) = 10.1, p < .001). In the case of cute and tough, the
irons intended to express the same personality characteristic are not perceived to express
this personality characteristic to the same degree. Nonetheless, we will use the group
mean as calculated over all three irons in the test of the hypothesis because we think this
is a more robust test than selecting the best examples and testing these.
51
Table 3.1
Individual and group means of the irons for their target characteristics
Mean
SD
n
Happy irons (group mean)
3.11
1.11
264
Happy iron A
3.07
1.25
88
Happy iron B
2.95
.96
87
Happy iron C
3.30
1.10
88
Cute irons (group mean)
2.41
1.26
264
Cute iron A
2.80
1.32
87
Cute iron B
2.01
1.12
87
Cute iron C
2.43
1.21
87
Tough irons (group mean)
3.76
1.12
264
Tough iron A
4.10
1.05
88
Tough iron B
3.82
1.13
88
Tough iron C
3.35
1.06
88
Test of hypothesis
In order to determine whether the respondents recognized the personality
characteristic that was used as a basis for the design, we conducted a repeated measures
analysis of variance that compares the group means of the happy, cute and tough irons to
each other. Three analyses were conducted. First, we conducted an analysis that
compared the happy irons with the cute and tough irons for “happy” (results shown in
table 3.2). Second, we conducted an analysis that compared the cute irons to the happy
and tough irons for “cute” (results shown in table 3.3). Finally, we conducted a third
analysis that compared the tough irons to the happy and cute irons for “tough” (results
shown in table 3.4).
As formulated in the hypothesis, we expected that the irons designed to be happy
would receive significantly higher scores for happy than the irons designed to be cute or
tough. The results confirm this expectation. The happy irons are rated significantly
happier (M = 3.11) than the cute irons (M = 2.80, F(1,87) = 10.3, p < .05) and the tough
irons (M = 2.63, F(1,87) = 21.3, p < .001). However, the difference between the happy
and cute irons is less than the difference between the happy and the tough irons (see table
3.2).
52
Table 3.2
The respondents’ ratings of the irons for happy
Happy irons
Cute irons
Tough irons
Mean score for “happy”
SD
n
3.11
.71
.77
.75
88
88
88
2.80 *
2.63 *
* Mean differs significantly from the target mean, p < .05.
Next, we expected that the irons designed as cute would be rated cuter than the
happy and tough irons. The cute irons are indeed rated significantly cuter (M = 2.41) than
the tough irons (M = 1.94, F(1,87) = 18.5, p < .001), but the happy irons are rated cuter
than the cute irons (M = 2.70, F(1,87) = 7.6, p < .05). This means that the cute irons are
not perceived as the cutest (see table 3.3).
Table 3.3
The respondents’ ratings of the irons for cute
Cute irons
Happy irons
Tough irons
Mean score for “cute”
SD
n
2.41
.84
.75
.64
88
88
88
2.70 *
1.94 *
* Mean differs significantly from the target mean, p < .05
Third, respondents should rate the irons designed to be tough significantly
tougher than the irons designed to be happy or cute. The results show that this is the case
(see table 3.4). The tough irons are rated as tougher (M = 3.76) than the cute irons (M =
2.57, F(1,87) = 98.5, p < .001) and the happy irons (M = 2.88, F(1,87) = 63.4, p < .001).
Table 3.4
The respondents’ ratings of the irons for tough
Tough irons
Happy irons
Cute irons
Mean score for “tough”
SD
n
3.76
.75
.79
.82
88
88
88
2.88 *
2.57 *
53
We expected that if respondents recognized the product personality that was
used as a basis for the design, they would rate the irons designed to express a certain
characteristic higher for that characteristic than the irons not designed to express this
characteristic. For five of the six comparisons this expectation is confirmed.
The use of appearance characteristics
The design students who created the sketches of the irons, also answered a
questionnaire that asked them to explain the design choices (with respect to product
form) and to indicate which colors and textures they would have preferred for their
design. The answers were analyzed in order to gain some insight into the role of
appearance characteristics in product personality.
The results show that almost all students used associations as the basis of their
design. Despite the fact that there was no specific question about associations, almost all
students (16 out of 18, 89%) mentioned at least one. We also know from the
questionnaires that there is reasonable consensus about the kind of associations that were
used as a base for the designs. Most of the students that designed a happy iron associated
happy with smiling faces (4 out of 6) and/or the sun (3 out of 6). Almost all the students
designing a cute iron referred to small and young creatures (4 out of 6), either human or
animal. The students who designed a tough iron all mentioned masculinity, specified by
some of them in terms of machines, trucks and tools.
The use of associations may have caused the conformity in the appearance of the
irons that expressed the same personality characteristic. Analysis of the questionnaires
showed that happy is visualized using round and open forms and is associated with
bright, fresh colors, like yellow, orange and red. Cute is visualized using round and stocky
forms and associated with soft, warm colors. Some students even wanted to cover their
cute iron with fur. The students who designed a tough iron used big and robust forms
with sharp lines. They associated tough with dark colors, such as blue, green and grey.
3.3.3
Discussion
The purpose of the study reported in section 3.3 was to determine whether or
not people would recognize a personality characteristic in a product that is designed to
express that characteristic. Design students designed a happy, cute or tough iron and 88
respondents rated these irons on these same personality characteristics. The results of a
repeated measure ANOVA have, on the whole, shown a confirmation of our hypothesis
that respondents would rate the irons designed to be happy/cute/tough as
happier/cuter/tougher than irons not intended to be happy/cute/tough. The happy irons
54
are rated happier than the irons not designed to be happy. The cute irons are rated cuter
than the tough irons but not cuter than the happy irons. In fact, the happy irons are rated
cuter than the cute irons. People do recognize “tough” in the irons designed to be tough.
The irons designed to be tough were rated tougher than both the happy and the cute
irons.
The ambiguous result regarding the cuteness of the cute irons may be due to the
fact that happy and cute have related meanings. Happy and cute are more alike than
happy and tough or cute and tough. This semantic relation between the personality
characteristics is also visible in the product appearances. Cute and happy are both
visualized using round forms. The only difference is that happy is visualized using open
forms, whereas cute is visualized using stocky forms. On the other hand, tough is
visualized using completely different forms; there is nothing round in the appearance of
the tough irons.
Limitations
The most important limitation to this study is the minimal quality of the stimuli.
Design students were given only 45 minutes to sketch an iron and they were not allowed
to use color. It may be assumed that with more time the quality of the sketches would
have improved. Moreover, the students that participated were not selected on the basis of
their drawing skills. Some students participated because they liked drawing and were good
at it, others participated because they saw the project as a chance to improve their
(moderate) drawing skills. It may be assumed that experienced professional designers may
out-perform students in designing a particular impression.
Another limitation of the study is that we did not make a selection of the best
designs to serve as stimuli. The results of the cute irons are a good example of this effect.
Especially iron C2 stands out. It has a different product form compared to the other cute
irons (it is not round and stocky) and scores very low on “cute” (M = 2.1, SD = 1.1, n =
87). We took a random selection of nine from the original 18 irons, reasoning that this
would generate the most robust results. If people would recognize a target personality
characteristic in a random selection of product variants designed to reflect that
personality characteristic, then they would certainly recognize it in designs selected as the
best representatives of that characteristic. A stimuli selection that included stimuli that
best expressed the target personality characteristic would have resulted in better
recognition.
The fact that we used black and white sketches might also have had a negative
influence on the recognition of the target characteristics. Colored stimuli would have
probably increased recognition of the personality characteristics both in comparison to
the others and in isolation. The analysis of the questionnaires about the design of the
55
irons revealed a consistent preference for color with respect to each personality
characteristic. Nonetheless, people did recognize the intended personality characteristic
even with black and white stimuli.
Together, the results suggest that consumers recognize the target personality
characteristic based on limited appearance cues.
3.4
Conclusions
This chapter was set out to explore the influence of product appearance on the
perception of product personality. Two studies were conducted and the results show that
people use personality characteristics to describe and discriminate between product
variants that have different appearances, and that they recognize a product’s personality
based on appearance cues.
The personality of a product seems to be determined, at least in part, by the
appearance of a product variant. In both studies we used product variants from the same
product type as stimuli. The product variants in study 1 differed in appearance on many
accounts: color, size, form, etc., including visible functional differences. The product
variants that we used in study 2 were especially created sketches of irons. Each stimulus
offered exactly the same functional benefits, and the differences between the irons were
restricted to product form. Though the appearance of the stimuli in study 1 differed in
more aspects than the stimuli in study 2, we found differences in product personality
between the product variants in both studies.
Another indication that appearance is an important factor in product personality
stems from a comparison of the product variants that are described using the same
personality characteristics. A comparison of the product variants that have the same
product personality to product variants with another product personality reveals that the
differences in appearance between the groups are bigger than the differences within a
group. It even seems that product variants from different product types that are described
using the same personality characteristics have similar visual appearance characteristics.
The watches that are described as tough in the first study are big and robust, and the
irons designed to be tough in the second are big and robust too. Figure 3.6 shows an
example of this similarity in appearance between a tough watch and a tough iron.
56
Figure 3.6: A tough watch and a tough iron
We know now that people describe products using personality characteristics and
that product personality can be designed. At this point, the influence of product
personality on consumer preference becomes relevant. To make it worthwhile using
product personality in product design, it should not only be used to describe products, it
should also influence consumer preference. The next chapter will address this
relationship, by introducing the concept of product-personality congruence.
4
4.1
Product-Personality Congruence
and Consumer Preference
The similarity-attraction relationship
There are two main theories of attraction: one theory states that “opposites
attract” and the other states that “similarities attract”. The first theory states that people
are attracted to others who complement them. The “similarities attract” theory, otherwise
known as “homogamy”, states the contrary: people are attracted to others who are similar
to themselves. Research has mostly found confirming evidence for the similarityattraction relation (Antill, 1983; Nias, 1979). There are some situations in which opposites
attract (Dryer & Horowitz, 1997), but this effect faints in comparison to the influence of
similarity on attraction. Similarity of demographics (Newcomb, 1961), attitudes (Byrne,
1971), physical attractiveness (Feingold, 1988), and personality (Barry, 1970) have been
shown to increase attraction between people.
One of the first studies on similarity as a determinant of attraction (Newcomb,
1961) showed that similarity of demographic characteristics (such as age, college major,
urban versus rural background) increased attraction. In a longitudinal study of friendship
formation Newcomb measured the demographic characteristics of students in a
dormitory before they met, and tested whether those who were similar to each other liked
each other more than those who were dissimilar. The findings confirmed that early
demographic similarity is a good predictor for later interpersonal attraction.
Attitudinal similarity, the sharing of similar opinions, beliefs and values, has also
been proven to be a strong determinant of interpersonal attraction. Byrne (1961)
systematically varied attitudinal similarity. He first administered an attitude scale to
students then and later presented the same students with a scale alleged to be filled out by
58
another student. The students were then asked to indicate how much they liked the other
person and how much they would enjoy working together (measures of attraction).
Comparison of three levels of similarity (0% - 50% - 100%) clearly showed an increase of
attraction with an increase of similarity (Byrne, 1997). Later studies have confirmed and
elaborated upon these findings (e.g. Byrne, Clore & Smeaton, 1986; Byrne & Nelson,
1965; Singh & Ho, 2000).
The effect of similarity in physical attractiveness on attraction is also known as
the matching hypothesis. The matching hypothesis states that people prefer other people
who are similar to themselves in level of physical attractiveness. Similarity in physical
attractiveness has been found among dating couples, engaged couples, cohabitating
couples and married couples (Feingold, 1988).
People who are similar in personality are also more attracted to each other. With
the exception of the traits dominance and submission, where dissimilarity of traits seems
to increase mutual satisfaction, studies in interpersonal attraction have generally shown
that similarity in personality traits increases attraction (Byrne & Griffitt, 1969). College
students prefer roommates with similar personality traits (Carli, Ganley, & Pierce-Otay,
1991). People like strangers more when they share similar personality characteristics
(Byrne & Griffitt, 1969), and people are more willing to share assets or information with
others who are similar to themselves (Byrne, Clore & Worchel, 1966). Furthermore,
similarity in personality is an important factor in long-term relationships. Spouses with
similar personality characteristics report greater marital happiness (Antill, 1983; Barry,
1970; Tharp, 1963), and satisfaction with life (Arrindell & Luteijn, 2000).
Preference for similarity in human interaction is explained through the principle
of positive reinforcement. Similarity between two individuals is gratifying because each
individual validates and reinforces the self-concept of the similar other (Clore & Byrne,
1974). Since products may also allow people to validate and reinforce their self-concept,
we tried to extend the similarity-attraction hypothesis to human-product interaction.
Levy (1959) was one of the first to state that people prefer products that are
similar to themselves. Self-congruity theory (Sirgy, 1982) elaborates upon this idea. Selfcongruity theory suggests that consumers compare their self-concept with the user-image
of a product. This product-user image is the “stereotype of the generalized user”.
Motivated by self-consistency, people are expected to prefer a product with a user image
that is congruent with their self-concept. In other words: people prefer a specific product
because they see themselves as similar to the kind of people that are generally thought to
use this product. Many studies found supporting evidence for this user-image congruence
effect (Dolich, 1969; Ericksen & Sirgy, 1989, 1992; Grubb & Hupp, 1968; Heath & Scott,
1998; Hong & Zinkhan, 1995; Landon, 1974; Malhotra, 1988; Sirgy, 1985). Another
similarity attraction effect that is found with respect to products is the brand-personality
59
congruence effect. Aaker (1999) found that people prefer brands with which they share
personality characteristics.
An important aspect of the similarity-attraction studies in interpersonal attraction
is similarity of personality. We converted this personality aspect to human-product
interaction by means of product personality. The idea is that consumers compare
themselves with the personality of a product variant and will prefer product variants with
a matching product personality. The current chapter reports three studies that test this
idea. The first study investigates whether consumers prefer product variants with a
matching product personality, and whether this influence is independent of Sirgy’s userimage congruence effect (section 4.2). The next two studies extend the similarityattraction principle to a post-purchase situation, and investigate whether consumers are
more satisfied with, and more attached to, product variants with a matching product
personality (section 4.3).
4.2
The influence of product personality in a pre-purchase situation
Research in human interaction showed that the similarity effect is already present
in first encounters with strangers. In terms of consumer behavior, this means that the
similarity effect influences pre-purchase behavior and, in particular, product evaluation.
Products with personality associations similar to the personality of the person allow
him/her to validate and reinforce his/her self-concept and thus will be preferred over
products with a dissimilar personality. In line with the previous literature, we will call this
the product-personality congruence effect.
The comparison that is the basis of product personality congruence includes the
self-concept of the consumer. The self-concept can be regarded either as a single
construct or as consisting of multiple constructs. Some authors (e.g., Bellenger, Steinberg
& Stanton, 1976; Birdwell, 1968; Grubb & Hupp, 1968; Grubb & Stern, 1971; Hughes &
Guerrero, 1971) have discussed “self-concept” as a single construct referred to as the
actual self, real self, basic self, or extant self (Sirgy, 1982). Other authors (Belch, 1978;
Belch & Landon, 1977; Dolich, 1969; Hong & Zinkhan, 1995; Kassarjian, 1971;
Malhotra, 1981, 1988) view the self-concept as consisting of multiple components. Most
of them identify two aspects within the self-concept: the actual and the ideal self. Sirgy et
al. (1997; 2000) distinguish four different self-concepts: (1) actual self (defined as how
people see themselves), (2) ideal self (defined as how people would like to see
themselves), (3) actual-social self (defined as how people believe they are seen by
significant others) and (4) ideal-social self (defined as how people would like to be seen
by significant others). The risk of this division into multiple selves is that the concept
60
looses its meaning (Baumeester, 1998). According to the frequently quoted definition that
the self-concept is “the totality of the individuals thoughts and feelings having reference to himself as
an object” (Rosenberg, 1979, p. 9), the self-concept encompasses all ideas one has about
himself, the “totality”. Therefore, we will refer to the self-concept as a single construct.
The current study investigates whether congruity between the self-concept and
the personality of a product variant (product-personality congruence) results in a more
positive product evaluation. We have seen that this principle is very strong in human
interaction and expect that people prefer products that are like themselves. This is
formulated in the first hypothesis.
Hypothesis 1
Product-personality congruence has a positive influence on product evaluation.
Just like the user-image congruence effect, the effect of product-personality congruence
on product evaluation is based on congruity. Yet the nature of the two effects is different.
Product-personality congruence suggests that a consumer compares him/herself to the
personality that he/she ascribes to a product, whereas user-image congruence proposes
that a consumer compares him/herself to the stereotype of the generalized user.
Therefore, the product-personality congruence effect and the user-image congruence
effect should be independent. This expectation is tested with hypothesis 2.
Hypothesis 2
Product-personality congruence has a positive influence on product evaluation
independent of the influence of user-image congruence.
We conducted a two-phase study to test these hypotheses. In the first phase, productpersonality congruence was measured. The second phase was used to establish user-image
congruence.
4.2.1
Method
Respondents
Respondents were randomly selected from the consumer household panel of the
Faculty of Industrial Design Engineering, Delft University of Technology. The consumer
panel contains an a-select sample of the population of Delft and the surrounding area
(see Tan, 2002). Forty-eight respondents participated in the first phase of the study. Their
age varied from 26 to 69 years with an average of 46. All 48 respondents who participated
61
in the first phase of the study were approached to participate in the second phase. Thirtyseven respondents (77%) reacted positively to the invitation. Their ages varied from 28 to
69 with an average age of 48.
Participation was voluntary in both phases. In the first phase respondents were
rewarded with a small financial compensation. In the second phase, they received a gift
voucher. Both times, they received a written debriefing. The amount of male and female
respondents in each phase was about equal.
Stimuli
Twelve stimuli products were selected, three product variants from four different
product classes: screwdrivers, coffeemakers, soap-dispensers and table wine. The soapdispensers were decorative bathroom equipment, empty and to be filled at home. Pictures
of the stimuli are shown in figure 4.1 (page 64).
Since we know that differences in product appearance will generate differences in
product personality, the selection of product variants within each product class was based
on creating variety in appearance. Furthermore, in order avoid that quality and/or price
differences would affect consumer preference, the product variants within a product class
offered the same standard functionality and were approximately similar in price. All
stimuli were bought in a local department store, with the exception of the wine. The wine
was an inexpensive red table wine bought at a supermarket.
Design
To reduce order effects, three sets containing four product variants – one from
each product class – were created. These three sets were presented in every possible
order, resulting in six different set-sequences. Within the sets, the four products were
presented in four balanced orders, according to the digram-balanced square. This means
that every product variant immediately follows and immediately precedes another product
variant in the set only once (Maxwell & Delaney, 1990). This resulted in a total of 24 (six
times four) different sequences of stimulus presentation that were each represented twice.
In phase 1, every two respondents received the product variants in a unique sequence. In
phase 2, each respondent received a questionnaire picturing the product variants in the
same sequence as they were presented to them in phase 1.
Procedure
The study was split into two phases. The first phase consisted of face-to-face
interviews used to establish the product personalities, the experienced productpersonality congruence and the products’ evaluation. In the second phase, the same
62
respondents received a mail questionnaire assessing user-image congruence. In both
phases, the respondents rated the same 12 stimuli. Table 4.1 gives an overview of the
study’s characteristics.
Table 4.1
Overview of the study
n
Phase 1
48
Phase 2
37
Method
Interview
Self completion questionnaire
Self completion questionnaire
Mail questionnaire
Measures
Product Personality
Product-personality congruence
Product Evaluation
User-Image Congruence
Phase 1
The interview consisted of two parts and lasted approximately one hour.
The first part of the interview started after a test-run in which the respondent had to
describe some famous persons (e.g. the Queen, Madonna, the Prime Minister) using
personality descriptors. The goal of this test-run was to train respondents to answer in
adjectives. Following this task, the respondents were shown the 12 stimuli, one at a time.
A respondent was first asked to describe the personality of a stimulus. Then, the
interviewer read this personality description back to the respondent. Instructed to keep
the personality description in mind, respondents filled out the questionnaire measuring
the experienced product-personality congruence with that stimulus. This procedure was
repeated until for each stimulus and lasted approximately 45 minutes.
In the second part of the interview, the product variants were shown to the
respondents again. They were now asked to fill out a questionnaire, evaluating each
stimulus. This procedure lasted approximately 15 minutes.
Phase 2
The second phase started ten months after phase 1 had finished. Leaving
considerable time between the two phases allowed people’s memory of the product
personality to fade out, thereby reducing the chance of product personality affecting the
product-user image. Respondents received a mail questionnaire of 12 pages. A color
image of one stimulus was pictured on each page. The same 12 stimuli as in phase 1 were
used. On each page, a short text instructed respondents to “take a moment to call this
screwdriver/coffeemaker/soap-dispenser/table wine to mind. Then, try to imagine what kind of people
generally prefer this screwdriver/coffeemaker/soap-dispenser/table wine”. Respondents then filled
out the user-image congruence questions for that stimulus. Answering the complete
questionnaire took approximately 10 minutes.
63
Measures
Product personality is measured in the same way in which Aaker
Product personality
(1997) measured brand personality before she developed a brand personality scale.
Respondents were asked to verbally describe each product variant “as if it were a person”.
The interviewer wrote down their description. To make sure the written description
correctly represented the respondents’ answer, the product personality description was
read back to the respondent. If necessary, the respondent was able to make corrections or
additions. This procedure reassured respondents that their answer was written down
accurately.
In order to measure product-personality congruence, we used a direct measure of
congruence instead of the traditional difference scores. Difference scores involve the
subtraction of one measure from another to create a measure of a distinct construct.
They have been criticized as being poor measures (Cronbach & Furby, 1970; Peter,
Churchill, & Brown, 1993; Wall & Payne, 1973). Difference scores may seem like purer
measures, yet they (1) are less reliable (due to the correlation between the components),
(2) have discriminant validity problems (since the difference-score measure is not distinct
from its components), and (3) show artificial correlations with other variables (caused by
correlation between this other variable and one of the components) (Peter et al., 1993).
Peter et al. therefore suggest to “operationalize constructs more directly so that the subsequent
calculation of difference scores is unnecessary” (p. 661). Sirgy et al. (1997) used a direct measure to
measure user-image congruence and showed that this method had a greater predictive
value than traditional difference scores.
Another problem with difference scores is more specific to the study of product
personality. There is no measure that can be used to validly measure both the personality
of a person and that of a product. Though we use the same personality terms to describe
people and products, the two concepts of human personality and product personality are
too different to be measured with the same scale. Kassarjian (1971) made a strong case
that personality measures developed to assess human personality should not be used in
marketing research and advised researchers to develop their own instruments.
Based on these considerations, we used items analogous to Sirgy et al. (1997) to
measure product-personality congruence, except that we referred to the product
personality description given by the respondents instead of referring to the general user
(see table 4.2). We used 4 items and asked respondents to rate them on 5-point scales.
Reliability analysis indicated that the item “matches me” did not contribute to the internal
consistency of the scale. For this reason, we have chosen to leave it out of the analysis.
64
Based on the remaining items, an average product-personality congruence score of each
respondent for each product was calculated (α = .91).
Product evaluation was measured with four items using 5-point
Product evaluation
scales (see table 4.2). Reliability analysis indicated that the internal consistency of the
evaluation increased if the item “good” was removed. Therefore, this item was left out of
the analysis. Based on the three remaining items, an average product evaluation score of
each respondent for each product was calculated (α = .89).
We used the direct measure (Sirgy et al., 1997) to measure userUser-image congruence
image congruence. In order to create greater predictive validity the items were analogous
to the product-personality congruence questionnaire. Since it appeared from phase 1 that
the item “matches me” did not contribute to the internal consistency of the measure, we
left it out of the user-image congruence measure. This way, three items remained (see
table 4.2). Answers were rated using 5-point scales. An average product-personality
congruence score of each respondent for each product was calculated (α = .95).
Table 4.2
Items used in the measurement scales (in English and Dutch)
• This product is not like me/is like me.
Dit product lijkt niet op mij/lijkt op mij.
• I do not identify/I identify myself with my description of the product.
Ik herken mezelf niet/Ik herken mezelf in de omschrijving van dit product.
• This product does not match me/matches me.
Dit product past niet bij mij/past bij mij.
• If you consider your own personality and compare it to the description you just
provided, to what extend are they dissimilar/similar?
Als u denkt aan uw eigen karakter en dat vergelijkt met de omschrijving die u net heeft
omschreven, in hoeverre komen die twee dan overeen? (komen niet overeen/komen wel
overeen)
Product evaluation
• I think this product is not beautiful/is beautiful.
Ik vind dit product niet mooi/mooi.
• I would not like/I would like to have this product.
Dit product zou ik niet willen hebben/zou ik willen hebben.
• I think this product is not attractive/is attractive.
Ik vind dit product niet aantrekkelijk/aantrekkelijk.
• I think this is not a good product/is a good product.
Ik denk dat dit geen goed product is/een goed product is.
65
Table 4.2 continued
User-image congruence
• If you consider the types of people who prefer this product, do you identify with
these people? (I do not identify/I do identify with this kind of people)
Als u denkt aan het soort mensen dat een voorkeur heeft voor dit product, herkent u zichzelf dan
in dit soort mensen? (Ik herken mezelf niet/ik herken mezelf wel in dit soort mensen)
• If you consider the types of people who prefer this product, are you like these
people? (I am not like /I am like this kind of people)
Als u denkt aan het soort mensen dat een voorkeur heeft voor dit product, lijkt u dan op dit soort
mensen? (Ik lijk niet/ik lijk wel op dit soort mensen)
• If you consider the types of people that prefer this product, are they similar to the
way you see yourself? (The way I see myself is not similar/similar to this kind of
people)
Als u denkt aan het soort mensen dat een voorkeur heeft voor dit product, komt dit dan
overeen met het beeld dat u van uzelf heeft? (Komt niet overeen/komt overeen met dit soort
mensen)
4.2.2
Results
Product personality
The product personality of the stimuli was established by gathering the
personality adjectives per product variant. The results indicate that respondents used an
average of 4.44 (n = 48, SD = 2.02) personality characteristics for each product variant.
Figure 4.1 gives an impression of the product personality of the stimuli. The
characteristics in figure 4.1 are personality descriptors mentioned by Anderson (1968)
and/or Doddema and De Raad (1997).
Validity and uni-dimensionality of measurement scales
Examining the items of the product-personality congruence, product evaluation
and user-image congruence measurement scales in a principal component analysis with
varimax rotation yielded three factors with eigenvalues greater than 1; one factor for each
measurement scale. The items that were removed because they were found not to be
internally consistent were not included in the analysis. The remaining items each have
loadings above .80 (product-personality congruence), .70 (product evaluation) and .90
(user-image congruence) on their factor indicating discriminant validity.
A principal component analysis on each scale individually results in a one-factor
solution for every scale. All three scales thus appear to be unidimensional.
66
Screwdrivers
Reliable
Cold
Straightforward
Honest
Steadfast
Rugged
Aggressive
Distant
Unkind
Sociable
Kind
Warm
Friendly
Dependable
Funny
Childish
Two-faced
Domestic
Silly
Cheerful
Feminine
Charming
Conventional
Gullible
Firm
Coffeemakers
Introvert
Outspoken
Soap-dispensers
Arrogant
Masculine
To-the-point
Table-wines
Self-conscious
Independent
Intelligent
Figure 4.1: Personality characteristics used to describe the stimuli per product variant
Test of the hypotheses
First we consider the expectation that product-personality congruence has a
positive effect on the evaluation of the stimuli (hypothesis 1). We performed a linear
regression in which the dependent variable product evaluation is predicted based on the
independent variable product-personality congruence. The resulting regression weights
and proportion of explained variance are shown in the first row of table 4.3. The results
indicate that product-personality congruence determines a significant part of the product
evaluation (ß = .48; t(442) = 13.08, p < .001), meaning that the product variants with a
high score on product-personality congruence are evaluated more positively than the
67
product variants with a low product-personality congruence score. These findings
confirm hypothesis 1: products are evaluated more positively when they have a
congruous product personality.
Table 4.3
Effects of product-personality congruence and user-image congruence on product
evaluation
R2 adjusted
B ß
R2
h1 Product-personality congruence
.56 .48*
h2
Step 1
.52
.49*
Step 2
.37
.46
.35*
.40*
.23
.23
.24
.24
.38
.38
* p < .001
Hypothesis 2 predicts that product-personality congruence has a positive
influence on product evaluation independent of the influence of user-image congruence.
Using a two-step linear regression, we first entered user-image congruence as a predictor
variable and then added product-personality congruence. The results are shown in the
second and third row of table 4.3. Though the addition of product-personality
congruence as an extra predictor next to user-image congruence results in a decrease of
the user-image congruence effect, both betas remain significant. Furthermore, the
addition of product personality congruence leads to a significant increase of explained
variance (R2change = .14, F(2,441) = 101.46, p < .001). The fact that product-personality
congruence and user-image congruence are independent predictors of product evaluation
is confirmed by the high tolerance value, .89 (Hair et al., 1999). These results confirm
hypothesis 2, user-image congruence and product-personality congruence are
independent determinants of product evaluation.
4.2.3
Discussion
The first purpose of this study was to determine whether product-personality
congruence positively influenced product evaluation. Secondly, we investigated whether
this influence is independent of the influence of user-image congruence. In a two-phase
study, respondents first indicated the product personality of 12 product variants and
indicated the experienced level of product-personality congruence. Later, they evaluated
68
the stimuli. Then, ten months later, the same respondents indicated the level of userimage congruence with the same stimuli. The results show that there is a significant
positive product-personality congruence effect. As with interpersonal attraction, people
prefer a product with a personality that is similar to their own.
In addition to the fact that a self-congruent product personality has a positive
influence on product evaluation, we also found that this influence is independent of the
user-image congruence effect. Both effects result in significant beta-values, though the
influence of user-image congruence decreases somewhat if the product-personality
congruence effect is added. This decrease is not surprising, assuming that productpersonality congruence will eventually result in user-image congruence. If the consumers
who see themselves as similar to a product actually buy the product, the part of the userimage that is determined by the image of the actual users (in addition to, for example, the
influence of advertising) truly is an effect of product-personality congruence.
Product-personality congruence can be seen as an extension of the self-congruity
theory. It positively influences preference according to the same principle as user-image
congruence, yet on a different level. Product personality congruence is more individual
and product specific than user-image congruence. Since it is more difficult for a
stereotypic image of the users to develop when a product is privately consumed, the
effect of user-image congruence may be stronger for publicly consumed products than
for products that are privately consumed (Sirgy, 1982). Due to its more individualistic
nature, product personality might be more relevant for privately consumed products than
user-image congruence. A post hoc analysis supports this argument. For the least visible
product variants (the screwdrivers), product-personality congruence is a much better
predictor of product evaluation than user-image congruence (ß product personality = .48,
t(110) = 5.81, p < .001; ß user-image = .19, t(110) = 2.33, p < .05).
In conclusion, the results of this study confirm that the similarity-attraction
principle that applies to human interaction also applies to human-product interaction.
People prefer products that have a similar personality to products that have a dissimilar
personality. Furthermore, the results confirm that product-personality congruence and
user-image congruence are two different effects. A consumer may prefer a product
variant because of product-personality congruence. This consumer may or may not
experience user-image congruence at the same time. For example, a person may want a
yellow Volkswagen Beetle because it has a funny, feminine, light-hearted personality with
which she identifies (product-personality congruence). At the same time she may identify
with the young, trendy, self-conscious people that generally drive Volkswagen Beetles
(user-image congruence). Both kinds of congruity lead to positive product evaluation but
in different and independent ways.
69
Limitations
A limitation of this research is that the effect that we found of productpersonality congruence on product evaluation may be the result of the same relationship
the other way around. That is, people may have judged themselves to be similar to the
product variants that they liked the most. Even though we measured product evaluation
after we had established product-personality congruence, we only assessed correlations.
Regression analysis does not prove a causal influence and we know from attraction
literature that similarity is not only the cause of attraction, it can also be its consequence
(e.g. Dryer & Horowitz, 1997).
Unfortunately, we cannot rule out this alternative explanation. Had we used a
more indirect measure of similarity, then this alternative explanation may have been less
plausible. A solution may be found in an experimental set-up where similarity is artificially
construed instead of measured. The next section reports a study using this kind of set-up.
The study investigates how robust the product-personality congruence effect is. Similarity
in human personality not only influences first impressions, it also influences long-term
relationships (Barry, 1970; Tharp, 1963). Therefore, we also wanted to investigate the
influence of product-personality congruence in a long-term consumer-product relation.
4.3
The influence of product personality in a post-purchase situation5
After the successful translation of the similarity-attraction hypothesis to prepurchase product evaluation, this section we will try to extend this similarity-attraction
relationship to post-purchase product evaluation. We will concentrate on satisfaction and
product attachment as forms of long-term product evaluation. That is, we will test the
positive influence of the product-personality congruence effect on satisfaction and
product attachment.
The principle of similarity increasing attraction has been shown to influence the
duration and quality of long-term relationships (Barry, 1970). In a longitudinal study of
marital success and failure, Bentler and Newcomb (1979) assessed several personality
traits of newlywed couples and obtained information about their background
(demographic variables such as age, race, education and occupation). They hypothesized
that those marriages that are still intact in a follow-up, had shown more similarity at the
beginning of the marriages. The result showed support for this hypothesis: the couples
that are still married after 4 years had more positive correlations in personality and
demographics than the divorced couples. Other studies concluded: (1) that similarity of
5
The research in this section has been conducted in cooperation with ir. R. Mugge, Delft
University of Technology.
70
physical attractiveness between partners in a romantic relationship is predictive of the
success of the relationship (Critelli & Waid, 1980; Feingold, 1988), (2) that spouses with
similar personality characteristics report greater marital happiness (Tharp, 1963; Antill,
1983), and (3) that similarity between partners’ personalities is a major factor in marital
satisfaction and satisfaction with life (Arrindell & Luteijn, 2000).
Translated to human-product interaction, long-term human relationships
approximate product attachment. Product attachment is defined as the strength of the
emotional bond a person experiences with a product (Schifferstein & Pilgrim, 2003).
Several studies have investigated the construct of product attachment (Ball & Tasaki,
1992; Belk, 1988, 1992; Kleine, Kleine, & Allen, 1995; Schultz, Kleine, & Kernan, 1989;
Sivadas & Venkatesh, 1995; Wallendorf & Arnould, 1988) and concluded that reflection
of the self-concept is an important determinant for product attachment. Products with
personality associations similar to the personality of the person allow him/her to reflect
their self-concept by means of these products. As a consequence, people may become
more attached to product variants with a product personality that is similar to their own
personality. Consequently, we expect product-personality congruence to have a positive
effect on product attachment.
This expectation is tested using an experimental procedure. Respondents are
presented with a scenario in which the personality of a person is portrayed. A scenario is
a written outline of a specific situation or, in this case, a person. Scenarios have been used
extensively in post-purchase consumer research, such as in studies of complaint behavior
and studies of regret (e.g., Inman & Zeelenberg, 2002; Tsiros & Mittal, 2000). Scenarios
have the advantage that the independent variables can be manipulated while at the same
time controlling for intervening effects (e.g. usage situation) that can play a role in the
development of product attachment. Together with the scenario, respondents were
presented a color picture of a product variant with a personality that is either similar or
dissimilar to the personality of the person in the scenario. Respondents were then asked
to indicate how strong they thought the person in the scenario to be attached to this
product variant. The hypothesis we tested is:
Hypothesis 1
People are thought to be more attached to product variants with a product personality
that is similar to their own personality, than to product variants with a product
personality that is dissimilar.
In addition to product attachment, we also studied the effect of productpersonality congruence on satisfaction with the product because research on post-
71
purchase behavior has largely focused on this aspect (e.g., Fournier & Mick, 1999;
LaBarbera & Mazursky, 1983; Richins & Bloch, 1991).
Satisfaction is generally defined as a judgment of the chosen product variant
following the purchase act, with respect to pre-purchase beliefs and expectations. The
confirmation or disconfirmation of pre-purchase product standards is considered to be
the essential determinant of satisfaction. It is suggested that: confirmation of standards
leads to moderate satisfaction, positively disconfirmed (exceeded) standards lead to high
satisfaction and negatively disconfirmed (underachieved) standards lead to dissatisfaction.
However, recent research has shown that it is not only the comparison of standards with
product performance that matters in satisfaction. The symbolic meaning of products is
also relevant (Fournier & Mick, 1999) and, as such, product personality may also be
relevant to post-purchase satisfaction. We expect that this relationship may also work
through the reflection of self-concept; that is, people are more satisfied with products
that have a product personality that matches their own personality. The procedure was
similar to the procedure used for product attachment. The hypothesis tested is:
Hypothesis 2
People are thought to be more satisfied with product variants that have a product
personality similar to their own personality, than with product variants that have a
dissimilar product personality.
In the following sections, two studies are reported. Study 1 tests the two
hypotheses with respect to the personality dimension “conscientiousness”. Study 2 tests
the hypotheses with respect to the personality dimension “extroversion”. The stimuli
personalities in both studies are based on dimensions from the five-factor model of
human personality. The existence of these five dimensions is supported by a large body
of research and literature (e.g. Digman, 1990; Goldberg, 1981, Hendriks, Hofstee, & De
Raad, 1999; McCrae & John, 1992). The dimensions differentiate themselves from each
other and are recognizable. A possible disadvantage connected to the use of higher-order
dimensions like the ones from the five-factor model is that they are very comprehensive.
Therefore, it may be difficult to accurately describe their nature and they may be more
open to various interpretations. However, the use of higher-order dimensions makes the
person’s personality more realistic and makes the interpretation more intuitive (Carver &
Scheier, 1996). Moreover, each dimension consists of numerous traits which makes it
easier to describe a person’s personality than it would have been had we used individual
traits. Since the recognition of both the personality of the person and the personality of
the product are essential to the success of the manipulation of our study, we chose to use
the dimensions of extroversion and conscientiousness. In person perception, the
72
dimensions of extroversion and conscientiousness were found to be the best recognized
by observers (Kenny et al., 1994).
4.3.1
Method study 1
Respondents
Ninety respondents were randomly selected from the consumer household panel
maintained by the Faculty of Industrial Design Engineering. The consumer panel consists
of an a-select sample of the population of Delft and the surrounding area (see Tan, 2002).
Seventy-three respondents (33 males and 40 females) returned their questionnaire (81%).
Ages ranged from 23 to 70 years, with an average age of 50. Respondents were rewarded
with a financial compensation and received a written debriefing afterwards.
Stimuli
The scenario that was used to depict an individual with a conscientious
personality portrayed a 27-year-old woman, named Susan, and described her as
conscientious by outlining her traits (e.g., precise) and hobbies (e.g., reading). To make
the scenario as realistic as possible, the scenario was based on the items from several
human personality tests (FFPI; Hendriks, Hofstee, & De Raad 1995; NEO-PI-R;
Hoekstra, Ormel, & De Fruyt 1996). A pre-test tested the internal validity of the scenario.
The scenario was rated on five traits that are typical for the conscientiousness dimension:
items: (1) “(not) precise”, (2) “(not) neat”, (3) “(not) consistent”, (4) “(not) serious”, and
(5) “(not) trustworthy” (α = .92). All items were measured using five-point scales. We
used a convenience sample of students (n = 19) who each rated the scenario. Results
showed that the personality in the scenario was perceived as intended. The mean score on
the conscientiousness items was 4.46 (SD = .38). A one-sample t-test revealed that this
mean score was significantly higher than the neutral midpoint of the scale (t(18) = 16.64,
p < .001), suggesting internal validity of the scenario. The scenario is reprinted in
appendix D.
A second pre-test was conducted in order to select a conscientious product
variant and a non-conscientious product variant. A stimulus-set of five toasters with
design variety was created, as we expected that this would generate diversity in product
personality. To avoid the possibility of price differences affecting product attachment, the
selected toasters were approximately similar in price. Furthermore, the toasters’ brand
names were removed from the pictures in order to avoid possible brand personality
effects. We used a convenience sample of students (n = 92) who each rated one toaster.
The toasters were assigned randomly. Students that had participated in the first pre-test
73
were excluded. Color pictures of the five toasters were tested on the same traits as used
for the test of the scenarios (α = .77). The toasters and their mean scores on the
conscientious traits are shown in table 4.4. The toaster with the highest mean, toaster 4
(M = 4.16, SD = .60, n = 19), and the toaster with the lowest mean, toaster 3 (M = 3.08,
SD = .52, n = 18), were selected as stimuli for the main study.
Table 4.4
Mean scores of the toasters on conscientiousness
Toaster 1
Toaster 2
Toaster 3
Toaster 4
Toaster 5
M
3.40
3.37
3.08
4.16
3.71
SD
.71
.50
.52
.60
.52
n
18
18
18
19
19
Further tests included a one-sample t-test, that revealed that the mean score of
the conscientious toaster was significantly greater than the neutral midpoint of the scale
(t(18) = 8.38, p < .001). This suggests that the toaster was indeed perceived as
conscientious. In contrast, no significant difference was present between the mean score
of the non-conscientious toaster and the neutral midpoint of the scale (t(17) = .63, p =
.54). Furthermore, the results showed that the selected conscientious toaster was
perceived as significantly more conscientious than the selected non-conscientious toaster
(t(35)= -5.81, p < .001).
Design
Two conditions were generated, one resulting in high product-personality
congruence (conscientious person and conscientious toaster) and one resulting in low
product-personality congruence (conscientious person and non-conscientious toaster).
Each respondent was assigned randomly to one of the two conditions.
Procedure
Respondents were sent a questionnaire including the scenario and a color picture
of a toaster. The toaster was said to be owned by the person in the scenario. Respondents
were instructed to read the scenario of the person carefully and to form an impression of
74
the person described. Subsequently, they were asked to take a look at the picture and to
indicate the expected level of product attachment.
Measures
Product attachment was measured with six items (Schifferstein
Product attachment
& Pilgrim, 2003 – see table 4.5). All items were measured using five-point scales (1 =
“disagree”, 5 = “agree”). An average product attachment score of each respondent for
both conditions was calculated (α = .94).
Satisfaction
The expected level of satisfaction with the product variant was measured
with three items (see table 4.5) using 5-point scales (1 = “disagree”, 2 = “agree”). An
average satisfaction score of each respondent for both conditions was calculated (α =
.94).
Table 4.5
Product Attachment
• This toaster has no special meaning to Susan (-)
Deze broodrooster heeft geen speciale betekenis voor Susan
• This toaster is very dear to Susan
Deze broodrooster is Susan erg dierbaar
• Susan has a bond with this toaster
Susan heeft een band met deze broodrooster
• This toaster does not move Susan (-)
Deze broodrooster doet Susan niets
• Susan is very attached to this toaster
Susan is erg gehecht aan deze broodrooster
• Susan feels emotionally connected to the toaster
Susan voelt zich emotioneel verbonden met deze broodrooster
Satisfaction
• Susan is satisfied with this toaster
Susan is tevreden met deze broodrooster
• Susan is content with this toaster
Susan is content met deze broodrooster
• This toaster pleases Susan
Deze broodrooster bevalt Susan
4.3.1
75
Results study 1
Manipulation Check
A manipulation check was performed in order to check whether the two
experimental conditions resulted in respectively high (high ppc) and low productpersonality congruence (low ppc). To minimize demand characteristics, this manipulation
check was not included in the main study, but was performed as an additional pre-test.
For this pretest, we used a convenience sample of students (n = 46). Students that had
participated in one of the other pre-tests were excluded. The scenario and color pictures
of both the conscientious and the non-conscientious toaster were presented to each
respondent. Subsequently, the measure for product-personality congruence was obtained
for the two toasters. Product-personality congruence was measured with four items using
five-point scales (see table 4.2). Based on these items (α = .94), an average productpersonality congruence score of each respondent for each product was calculated. A t-test
showed that the manipulation was successful: a significant difference was present
between the two conditions (t(44) = 9.10, p < .001). As expected, respondents in the
“conscientious person and conscientious product” condition perceived the productpersonality congruence to be higher (Mhigh ppc = 4.01, SD = .73, n = 23) than those in
the “conscientious person and non-conscientious product” condition (Mlow ppc = 2.23,
SD = .59, n = 23).
Validity and unidimensionality of measurement scales
Examining the items of the product attachment and satisfaction measurement
scales in a principal component analysis with varimax rotation yielded two factors with
eigenvalues greater than 1; one factor for each measurement scale. The items of one scale
each have loadings above .6 (product attachment) and .8 (satisfaction) on their factor
indicating discriminant validity.
A principal component analysis on each scale individually results in a one-factor
solution for the satisfaction scale. Yet, a two-factor solution appears for the product
attachment scale. The product attachment scale thus seems to lack unidimensionality. A
confirmatory factor analysis (CFA) was performed in LISREL 8.50 (Jöreskog & Sörbom,
1993) on the product attachment items to further examine the dimensional structure of
the product attachment scale. Although our sample size seems relatively small to conduct
a CFA, the five-to-one ratio of sample to parameter estimates as recommended by
Bentler & Cho (1988) is not violated. The CFA demonstrated that for the six product
attachment items, the fit of a single-factor solution was poor (χ2 = 26.66, df = 9, p < .01;
GFI = .89; RMSEA = .166; SRMR = .086). To obtain an adequate fit, the modification
indices were inspected. This resulted in the deletion of the item “This toaster does not move
76
Susan (-)”. The remaining five items yielded a single-factor model with a good fit (χ2 =
4.03, df = 5, p = .54; GFI = .98; RMSEA = .000; SRMR = .015). This suggests that for
these five items the construct is unidimensional. Cronbach’s alpha of the five-item
product attachment scale was .89.
First we tested hypothesis 1. People are thought to be more attached to product
variants with a product personality that is similar to their own personality, than to
product variants with a product personality that is dissimilar. A t-test was performed with
product attachment as the dependent variable and product personality congruence (high
versus low) as the independent variable. In this analysis, the mean score on the adjusted
product attachment scale derived from the confirmatory factor analysis was used as the
dependent variable.
The results show a significant difference between the high and low conditions of
product-personality congruence (t(70) = 2.72, p < .01). The respondents who read about
the conscientious person and were presented the conscientious toaster (high productpersonality congruence) predicted a higher degree of product attachment (Mhigh ppc =
2.80, SD = .91, n = 37) than those who were presented the non-conscientious toaster
(low product-personality congruence; Mlow ppc = 2.16, SD = 1.08, n = 35). This means
that our first hypothesis has been confirmed. High product-personality congruence
results in a higher degree of product attachment than low product-personality
congruence.
To test hypothesis 2, a second t-test was performed with satisfaction as the
dependent variable and product personality (high versus low) as the independent variable.
People are thought to be more satisfied with product variants that have a product
personality similar to their own, than they are with product variants with a dissimilar
product personality. The results show that there is a significant difference in satisfaction
between the high and low conditions of product-personality congruence (t(71) = -4.16, p
< .001). The respondents who read about the conscientious person and were presented
the conscientious toaster (high product-personality congruence) predicted a higher degree
of satisfaction (Mhigh ppc = 4.18, SD = .56, n = 37) than those who were presented the
non-conscientious toaster (low product-personality congruence; Mlow ppc = 3.24, SD =
1.26, n = 36). Our second hypothesis has thus also been confirmed. High productpersonality congruence results in a higher degree of satisfaction than low productpersonality congruence.
77
The results of the first study are based primarily on the conscientiousness
dimension of the five-factor model of human personality (McCrae & John, 1992). In
order to validate and generalize our findings, it is interesting to also test the hypothesis on
another personality dimension. Therefore we conducted a second study. The main
objective of this second study is to replicate the findings for another personality
dimension of the five-factor model of human personality, namely extroversion (McCrae
& John, 1992). The procedure and measures used in study 2 are identical to those used in
study 1.
4.3.3
Method study 2
Respondents
Ninety respondents were selected from the same consumer household panel as
used in study 1. Panel members who participated in study 1 were excluded from
participation in this second study. Sixty-two respondents (26 males and 36 females)
returned their questionnaire, a response rate of 69%. Ages ranged from 30 to 70 years,
with an average age of 49. Respondents were rewarded with a financial compensation and
received a written debriefing afterwards.
Stimuli
Similarly to study 1, a scenario was used to depict the personality of the person.
The scenario was similar to the conscientious version in that it depicted a 27-year-old
woman, named Susan. This time she was described as an extroverted person using traits
(e.g., enthusiastic) and hobbies (e.g., singing in a band). The scenario was based on items
from well-established human personality tests (FFPI; Hendriks, Hofstee, & De Raad
1995; NEO-PI-R; Hoekstra, Ormel, & De Fruyt 1996). A pre-test tested the internal
validity of the extrovert scenario. The scenario was rated on five items typical for
extroversion: (1) “(not) cheerful”, (2) “(not) sociable”, (3) “(not) exuberant”, (4) “(not) defiant”,
and (5) “(not) conspicuous” (α = .81). All items were measured using five-point scales. An
average score was calculated for each respondent. We used a convenience sample of
students (n = 18) to rate the scenario. Results showed that the person in the scenario was
indeed perceived as an extrovert. The mean score on the extroversion items was 4.20 (SD
= .53). A one-sample t-test revealed that this mean score was significantly higher than the
neutral midpoint of the scale (t(17) = 9.58, p < .001), suggesting internal validity of the
scenario. Appendix D illustrates the scenario.
A second pre-test was conducted in order to select an extrovert product variant
and a non-extrovert product variant. We used the same set of color pictures of toasters as
78
in study 1 to keep the results of study 1 and 2 comparable. This time we tested the five
toasters on the extroversion traits (α = .88). A convenience sample of students was used
(n = 92) and each student rated one toaster. The toasters were assigned randomly.
Students that had participated in the first pre-test were excluded. The toasters and their
mean scores on the extrovert traits are shown in table 4.6. Analogous to study 1, the
toaster with the highest mean, toaster 3 (M = 4.10, SD = .91, n = 18), and the toaster
with the lowest mean, toaster 4 (M = 2.38, SD = .65, n = 19), were selected as stimuli.
Table 4.6
Mean scores of the toasters on extroversion
Toaster 1
Toaster 2
Toaster 3
Toaster 4
Toaster 5
M
2.96
3.68
4.10
2.38
3.19
SD
.99
.69
.91
.65
.74
n
18
18
19
19
18
A one-sample t-test revealed that the mean score of the extrovert toaster was
significantly greater than the neutral midpoint of the scale (t(17) = 5.12, p < .001). This
suggests that the toaster was indeed perceived as extroverted. In contrast, the mean score
of the non-extrovert toaster was significantly smaller than the neutral midpoint of the
scale (t(18) = -4.17, p < .01). Furthermore, the results showed that the selected extrovert
toaster was perceived as significantly more extroverted than the selected non-extrovert
toaster (t(35) = 6.65, p < .001).
Design
Two conditions were generated, one resulting in high product-personality
congruence (extrovert person and extrovert toaster) and one resulting in low productpersonality congruence (extrovert person and non-extrovert toaster). Each respondent
was assigned randomly to one of the two conditions.
4.3.4
79
Results study 2
Manipulation Check
Similarly to study 1, the manipulation check for the high/low conditions of
product-personality congruence was pre-tested. We used a convenience sample of
students (n = 46). Students that had participated in any of the other pre-tests were
excluded. The scenario and color pictures of both the extrovert and the non-extrovert
toaster were presented to each respondent. Subsequently, the measure for productpersonality congruence (α = .94) was obtained for the two toasters. A t-test showed that
there was a significant difference between the two conditions (t(44) = 6.73, p < .001) and,
thus, the manipulation was successful. When the personalities of the person and toaster
were similar, this resulted in the experience of high product-personality congruence
(Mhigh ppc = 3.61, SD = .78, n = 23). Dissimilarity between the personalities of the
person and the toaster resulted in low product-personality congruence (Mlow ppc = 2.01,
SD = .83, n = 23).
Validity and unidimensionality of the measurement scales
The items of the measurement scales were again examined using a principal
component analysis with varimax rotation. The results again yielded two factors with
eigenvalues greater than 1; one factor for product attachment and one for satisfaction.
The items each have loadings above .60 (product attachment) and .80 (satisfaction) on
their factor indicating discriminant validity.
A strictly confirmatory factor analysis was performed to test the structure from
the five-item measure of product attachment as used in study 1 on an independent
sample (Jöreskog & Sörbom 1993). The results showed the one-factor solution to be
replicable and well fitting (χ2 = 4.14, df = 5, p = .53; GFI = .97; RMSEA = .000; SRMR =
.024). Cronbach’s alpha of the product attachment scale (five items) in study 2 was .90.
Again, a t-test was conducted to test the first hypothesis. Product attachment was
used as the dependent variable. High versus low product personality congruence was used
as the independent variable. The findings did not reveal a significant difference between
the high and low conditions of product-personality congruence (t(60) = 1.75, p = .085).
The respondents in the “extrovert person and extrovert toaster” condition (high productpersonality congruence) predicted slightly more product attachment (Mhigh ppc = 2.56,
SD = .96, n = 31) than those in the “extrovert person and non-extrovert toaster”
condition (low product-personality congruence; Mlow ppc = 2.13, SD = .99, n = 31). Yet,
this difference is not significant. Therefore, hypothesis 1 has been not confirmed.
80
Another t-test was conducted to test hypothesis 2. Satisfaction was included as
the dependent variable and product-personality congruence as the independent variable.
It was expected that high product-personality congruence would result in more
satisfaction than low product-personality congruence. The results show a significant
difference in satisfaction between the two conditions (t(60) = 1.96, p = .05). The
respondents who read about the extrovert person and were presented the extrovert
toaster (high product-personality congruence) predicted a higher degree of satisfaction
(Mhigh ppc = 3.80, SD = .78, n = 31) as compared to those who were presented the nonextrovert toaster (low product-personality congruence; Mlow ppc = 3.28, SD = 1.09, n =
31). This means that hypothesis 2 has been confirmed.
4.3.5
Discussion
The purpose of the studies reported in section 4.3 was to extend the similarityattraction relationship we found in section 4.2 to post-purchase behavior. We did so by
examining the effect of product-personality congruence on product attachment and
satisfaction. The expectation was that product variants with a personality similar to the
personality of the person (high product-personality congruence) better reflect the
person’s self-concept. This thereby caused the owner to experience more satisfaction and
to become more attached to the product. With respect to conscientiousness, we found a
significant difference in product attachment and satisfaction between the condition of
high product-personality congruence and the condition of low product-personality
congruence. Study 2 replicated these findings for the extroversion dimension of human
personality. Yet, this only revealed a significant difference for satisfaction. Based on the
fact that three of the four hypotheses showed significant results that confirm our
hypotheses, we believe product-personality congruence to have a positive influence on
post-purchase behavior.
A noteworthy result of both studies is that, in general, the effect of productpersonality congruence seems to be stronger on satisfaction than for product attachment.
These results are in correspondence with our expectations. A person may experience
satisfaction with a product sooner and more often than (s)he will experience product
attachment. What is more, it is not likely that consumers develop very strong emotional
bonds with ordinary products such as toasters. Yet, consumers do experience high levels
of satisfaction with these kinds of products.
81
Limitations
The fact that the results of this study are solely based on the product class of
toasters can be seen as a limitation. Future research will have to investigate the productpersonality congruence effect on post-purchase behavior for other product classes as
well. However, the results of the study reported in section 4.2 showed that in a prepurchase situation, product-personality congruence is as relevant for utilitarian products
(screwdrivers) as it is for other product classes. The fact that we used a utilitarian product
like a toaster may have influenced the strength of the effect. Yet, this does not influence
the fact that product-personality congruence has a positive effect on post-purchase
consumer behavior.
Another aspect of this research that may be seen as a limitation is the fact that we
used scenarios to manipulate the personality of the owner. We did not measure real
attachment and satisfaction, but asked respondents to indicate the level of attachment
and satisfaction that they thought a fictional other experienced. However, the use of
scenarios was the only way to create an exact (mis)match between a person and a product
variant, while keeping other variables constant. In addition, asking people to indicate
what another person would do or think in a particular situation is a well-known
projection technique in qualitative research, which is used to find out people’s true
motivations (Gordon & Langmaid, 1995). What respondents indicated as the experience
of another may be interpreted as what they would experience if they were in a similar
situation. Nevertheless, future research will have to investigate whether productpersonality congruence has a positive effect on product attachment and satisfaction in
real life situations.
A final limitation is that in each study, the product-personality congruence effect
was examined for only one personality dimension. Yet, human personality consists of a
profile on five, very broad, dimensions (McCrae & John, 1992). We recognize the
potential disadvantages of investing a complex construct like personality in such a
simplified manner. However, we feel these limitations do not take away the value of our
findings, especially if we take into account that our research provides a first attempt to
study the effect of product-personality congruence on post-purchase behavior.
4.4
Conclusions
The research reported in this chapter has shown the relevance of product
personality to consumer behavior. Based on the similarity-attraction relationship from
social psychology, we formulated expectations about human-product interaction and
largely found confirming evidence. People not only recognize product personality
82
(chapter 3), product personality also influences their preference for, satisfaction with and
attachment to a product variant.
The results of the current research suggest that product-personality congruence,
user-image congruence and brand-personality congruence have some similarities. Userimage congruence and brand personality congruence are suggested to increase preference
(Sirgy, 1982; Aaker, 1999). Brand personality is also suggested to increase brand loyalty
and can thus result in a lasting relationship between the consumer and his/her brand
(Fournier, 1998). Similarly, our results suggest that product-personality congruence can
result in preference and in the development of a lasting relationship between the person
and the product variant. All three constructs thus appear to result in consumer
preference.
The fact that product-personality congruence has a positive influence on both
pre- and post-purchase behavior indicates that product personality may be a promising
tool for directing product development. If we are able to design product variants
according to a pre-determined personality, then we can create preference, stimulate
satisfaction and encourage attachment in different consumer segments based on
personality. This is even more relevant because personality is by definition a stable
construct and does not change much over time (Costa & McCrae, 1988). For this to be
possible we need to know what a certain personality characteristic means in terms of
products. What is a tough car? What should it look like? Figure 4.1 shows that there are
several plausible answers to this question.
Figure 4.2: A tough car?
The beginning of answering this question lies in the possibility to measure product
personality. Measurement of the product personality of existing product variants enables
us to better understand the precise meaning of product-personality characteristics as they
are used by the consumer. The next two chapters will report the development of such a
product personality measurement scale. The first step is to select the personality
characteristics that people use to describe both people and products. This process is
described in chapter 5.
5
Example of inkblot from Rorschach’s Inkblot Test
5.1
Assessment of human personality
Basic to all theories of human personality is the idea that the individual
differences described by personality can be measured through personality assessment
(Pervin, 1975). Personality assessment is something that, in an informal way, people do
everyday. People form an impression of the personality of other people in order to know
what they can expect of them (Carver & Scheier, 1996). The difference between
impression formation in daily life and personality assessment in personality psychology is
that personality psychology is concerned with systematic personality assessment (Carver
& Scheier, 1996; Pervin, 1975).
Systematic personality assessment involves the issues of validity and reliability.
Validity relates to the question of whether a method assesses what it is supposed to
measure. It refers to the degree to which a scale correctly represents the concept of
interest. Validity starts with defining exactly what is to be measured and making the
assessment method as accurate as possible. If a personality assessment method is not
valid, then the results are meaningless because it was not personality that was measured
but something else (Carver & Scheier, 1996; Murphy & Davidshofer, 1994). Reliability
differs from validity in that it relates not to what should be measured, but to how
something is measured (Hair et al., 1998). Reliability of an assessment involves the degree
of consistency between multiple assessments; high consistency means high reliability. One
form of reliability is test-retest reliability. Test-retest reliability involves the consistency
between the assessments of one individual at two points in time. Another form of
reliability is called inter-judge or inter-rater reliability and involves the consistency of
84
assessments conducted by different researchers (or rather “judges”). A third form of
reliability, internal consistency, is especially relevant to summated measurement scales,
and concerns the consistency between the items in the scale. The rationale for internal
consistency is that individual items in a scale should be measuring the same construct and
thus should provide consistent results (Hair et al., 1998).
Both validity and reliability are important in personality assessment. The
reliability or consistency of a scale ensures that variation in the results of the assessment
reflects true variation between the subjects (Murphy & Davidshofer, 1994; Pervin, 1975).
The validity of a scale ensures that these differences reflect differences concerning only
the characteristic one is attempting to measure (Churchill, 1979).
Different methods of human personality assessment
Literature distinguishes between different methods of human personality
assessment (Carver & Scheier, 1996; Murphy & Davidshofer, 1994; Pervin, 1975). There
are projective methods, subjective methods and psychometric methods of human
personality assessment. Projective and subjective methods are used mostly in clinical
settings. Psychometric methods are used mostly in empirical research.
In projective methods, people are asked to interpret an abstract or unstructured
stimulus. The assumption is that the person’s personality is revealed by the way in which
(s)he interprets this stimulus. Examples of projective methods are Rorschach’s Inkblot
test (see example at the beginning of this chapter) and the sentence completion technique
(Murphy & Davidshofer, 1994; Pervin, 1975).
Subjective methods are called subjective because a person’s personality is
assessed indirectly. During interviews or therapeutic sessions, a person is asked about
his/her perception of the world. This subjective perception will then provide information
about the personality of that person. Examples of subjective methods include the
(unstructured) interview as used by Carl Rogers and in psychoanalysis (Carver & Scheier,
1996; Pervin 1975).
Both projective and subjective methods of personality assessment have been
criticized for being intuitive and having difficulties regarding validity and reliability
(Pervin, 1975). Psychometric methods of personality assessment are especially concerned
with the issues of reliability and validity. Psychometric personality assessment involves
the use of standard questionnaires, which often are self-report scales. It is sometimes
referred to as objective personality assessment because the person’s responses are
recorded directly without interpretation taking place by an observer (Carver & Scheier,
1996). The best known examples of psychometric assessment are scales for measuring
human personality as defined by the five-factor model, such as the NEO Personality
Inventory (NEO PI, Costa & McCrae, 1992a), or (in the Netherlands) the Five Factor
85
Personality Inventory (FFPI, Hendriks, Hofstee, & De Raad, 1995) and NEO Personality
Inventory-Revised (NEO-PI-R, Hoekstra, Ormel, & De Fruyt, 1996). Psychometric
methods are preferred in empirical research because they out-perform the other
assessment methods on both validity and reliability.
5.2
The idea that individual differences in personality can be measured through
personality assessment is also relevant to product personality. In chapters three and four
we measured product personality partly by means of free description (sections 3.2 and
4.2). A respondent was presented with a product variant and asked to describe it “as if it
were a person”. This method resembles the projective method of human personality
assessment and has similar limitations. Systematic comparisons between product variants
were difficult because different people use different descriptions when referring to the
same aspect of a product’s personality. For example, some people referred to “aloof” as
“distant” and others as “cold”. Content analysis can solve this problem, yet it is very time
consuming and also provides an extra source of error.
In addition to free descriptions, we also used self-report scales asking
respondents to indicate how they perceived the personality of different product variants
(sections 3.3 and 4.3). This method was similar to the psychometric method of human
personality assessment. It provided results that allowed us to compare product variants
for a specific personality characteristic. An additional advantage of the self-report scales is
that they can be used in large scale empirical investigations without causing the analysis
procedure to become too time-consuming. Therefore, in future research, we would like to
be able to assess product personality with self-report scales. We cannot use the scales
used in sections 3.3 and 4.3 because these scales were only developed to assess product
variants for a specific range of personality characteristics. For use in future research we
need a valid and reliable scale that allows us both to assess the complete product
personality of product variants and to compare the personality of different product
variants. For this kind of scale to be valid it needs to comprise the full scope of the
relevant personality characteristics. This is not the case with the questionnaires used
earlier in this thesis.
A valid scale of product personality should be compatible with the definition of
product personality (Hair et al., 1998; Pervin, 1975). The definition of product personality
states that product personality involves the use of human personality characteristics to
describe and differentiate between product variants. It can be concluded from this
definition that the product personality scale should contain human personality
86
characteristics as defined in chapter 1. These human personality characteristics should not
only describe “stable, non-physical qualities of a person on which (s)he discriminates him/herself from
others” (Doddema & De Raad, 1997, p. xiv) but should also be used to describe and
discriminate between product variants. In describing product variants these personality
characteristics should refer to the same concept as when describing humans. This is
important because by transferring the personality characteristics from one context to the
other they could be interpreted differently. For example “coarse” is primarily used to
mean “rude and offensive” when used to describe a person, yet they could turn to mean
“rough and unpolished” when used to describe a product. In order to ensure that the
product personality scale is valid, the primary interpretation of the personality
characteristics should be similar in both situations.
Another defining component of product personality is the fact that personality
characteristics should be used to describe and differentiate between all kinds of products.
In chapter 1, we delineated the scope of our study to durable consumer products.
Therefore, a scale that assesses product personality should be fit to measure the
personality of different classes of durable consumer products. This includes highpersonalizing products, such as cars, and low-personalizing products, such as
screwdrivers.
In short, in order for a product personality scale to be valid it must contain items
that describe human personality characteristics, which then can be used to describe both
people and all kinds of durable consumer products. The next section will provide an
overview of existing scales used to assess the personality of products. Each scale is
evaluated for its validity as a measure of product personality.
5.3
Scales used to assess the personality of products
The scales used to assess the personality of products can be classified into two
categories. The first category includes scales originally developed to assess human
personality (for an overview see: Kassarjian, 1971; Malhotra, 1988 and Sirgy, 1982). Some
researchers have used standardized personality scales to assess the personality of products
and brands. This use has been criticized because these tests have been developed and
validated for measuring the personality of humans and are often especially attuned to a
specific population. In order to be appropriate for product variants and consumers,
researchers have made changes in the measures. Items have been taken out of context of
the total instrument, words have been changed, and tests have often been shortened
(Kassarjian, 1971). Kassarjian states that human personality scales are not valid for
assessing the personality of products and brands. He concludes that researchers in
87
marketing and consumer research must develop their own instruments, rather than using
tools designed for different purposes.
The second category contains scales especially developed for use in marketing or
consumer research. For example, Wells et al. (1957) developed “an adjective checklist for the
study of product personality”. The goal was to design an adjective checklist that was easy to
understand by lay people and short enough to be used in door-to-door surveys. A
preliminary selection was made of trait adjectives that are “taught for permanent knowledge in
the first, second and third grade” (p. 317). This selection was then edited by removing words
that (1) seemed unlikely to describe users of a particular product, (2) usually refer to
objects rather than people, and (3) are highly ambiguous when used out of context. Items
that describe social class and gender were added to the pool. The final measure consisted
of 108 trait adjectives to be rated on a 7-point scale (see appendix E).
This measure is not appropriate for measuring product personality in the present
research because it is developed to measure product personality defined as “the tendency to
associate well-known product variants with particular kinds of people” (Wells et al., 1957, p. 317).
The personality associations as assessed by this scale concern the users of a product, and
not the product variant itself. Moreover, the concept of “product” is interpreted by Wells
et al. in a broad sense, and seen as encompassing brands as well as products. Finally,
product personality as defined by Wells et al. is mostly concerned with fast-moving
products (e.g. soft drinks, cigarettes, tea) instead of durable consumer products. The scale
is thus not an appropriate measure of our definition of product personality.
Malhotra (1981) developed a scale for measuring self-concepts, person concepts
and product concepts. A beginning pool of 70 adjectives was generated and subsequently
reduced by four judges who independently evaluated the items. Agreement among all the
judges resulted in a pool of 27 items that they considered applicable for measuring selfconcepts, person concepts and product concepts. After statistical analysis based on two
samples, a final pool of 15 semantic differential adjective pairs resulted (see appendix E).
Test-retest correlations over a 4-week delay provided significant correlations. Coefficient
alpha estimates for sub-scales ranged from .50 to .70. Evidence of convergent and
discriminant validity is also reported.
The scale developed by Malhotra (1981) is not a valid scale for product
personality assessment either. The development of this scale is too product specific. One
of the defining components of product personality was that the personality characteristics
should be used with respect to all kinds of durable consumer products. The fact that
Malhotra only used one product class (automobiles) as stimuli during scale development
may have resulted in less diversity of personality characteristics than necessary for
describing all kinds of durable consumer products.
88
Jordan (2000) developed a scale for studying whether product personality is a
concept that is meaningful for designers. He collected 209 personality characteristics
during a brainstorm session with 4 designers, each describing 6 persons: a personal
friend/enemy, a (dis)liked film character (e.g. Uncle Scrooge from Donald Duck, Phoebe
from Friends), and a (dis)liked public figure (e.g. Margaret Thatcher or Nelson Mandela).
The personality characteristics that resulted from describing these persons were grouped
together on the basis of referring to related aspects, or of referring to opposite aspects. In
this way, 17 dimensions were identified and named (see appendix E). The same
respondents subsequently used these dimensions to describe the personality of 14
product variants. No validity or reliability tests are reported because the goal of this study
was not scale development per se. As a consequence, the scale is not a standard scale that
can be used in other contexts.
A scale that is developed for general use is the brand personality scale of Aaker
(1997). Combining personality characteristics from existing measures and adding
personality characteristics from a free association elicitation procedure generated an initial
pool of 309 personality characteristics. After an empirically based selection of the most
relevant personality characteristics, a set of 114 characteristics remained. Via factor, item
and reliability analyses the final scale was formed, consisting of 42 personality
characteristics (see appendix E). These 42 characteristics can be represented by 5
dimensions, three of which resemble three of the Big Five dimensions of human
personality (extroversion, agreeableness and conscientiousness). Throughout the
development of the scale, several reliability and validity tests were performed. These tests
consistently resulted in high scores. The brand personality scale has been explicitly
developed for measuring brand personality. Brand personality is different from product
personality. The scale does exist of human personality characteristics that are used to
describe different kind of product classes, yet these product classes also include fastmoving products. In addition, the personality characteristics in the scale are not selected
because people use them to describe and discriminate between product variants.
Personality characteristics that are used to describe and discriminate between brands may
not be relevant with respect to product variants. For example, “up-to-date” may
discriminate between Nokia and Siemens but may not discriminate between the mobile
telephones in the current collections because they are all “up-to-date”. As a consequence,
Aaker’s brand personality scale is not regarded as a valid scale for assess product
personality.
5.4
89
Conclusions
It may be concluded from the previous sections that there is no scale available
that can validly be used for the assessment of product personality. A scale of product
personality should include personality characteristics that are used to describe the product
variant itself. The scale should represent the range of personality characteristics used to
describe and discriminate between product variants from all kinds of product classes.
However, the existing scales have either been developed for human personality
assessment or for assessing the personality of the typical user of a product or a brand.
In addition, all of these measures are in English. We want to conduct our studies
in The Netherlands, with Dutch subjects, and product variants available at the Dutch
market. This means that the measure must be in Dutch. A lot of people in the
Netherlands understand English. Yet, comprehension of the English language generally is
insufficient to ensure validity of studies in English. Though scales can be translated, a
scale that is translated from another language is not necessarily equivalent to the original
(Mullen, 1995; Wong, Rindfleisch & Burroughs, 2003). Translation of personality
adjectives often leads to a change of connotation (Hofstee, 1990) and decreases the
validity of a scale.
Kassarjian (1971) concluded from his overview that consumer researchers must
develop their own instruments. In accordance with this recommendation, we will develop
a scale for the assessment of product personality. This scale may then provide
information that raises the actionability of product personality characteristics by
providing information about which product variants are perceived to have the same
personality characteristics. For example, if we put together product variants that are
perceived as cute (see figure 5.1), we can see that the product variants share some product
characteristics. It might be inferred from this example that cute product variants have
stocky, rounded forms, reddish color and make a stroke-able impression.
Figure 5.1: Cute product variants
90
It was shown in chapter four that product-personality congruence had a positive
influence on both pre- and post-purchase behavior. If we are able to design product
variants according to a pre-determined product personality, then companies can create
preference, stimulate satisfaction and encourage attachment in segments based on
personality. This is an attractive strategy because personality is, by definition, a stable
concept (Costa & McCrae, 1988). Chapters six and seven will describe the process of
developing a product personality scale that meets our requirements. The scales that have
been described in this chapter will serve as a starting point of this process.
6
Development of a
Product Personality Scale
Results of ‘Design Umfrage’ about Audi A8, BMW Z4 and Renault Megane
(Auto Motor und Sport, 8/2003)
6.1
General steps in scale development
The process of scale development is generally seen as consisting of consecutive
steps or stages (Churchill, 1979; Devellis, 1991; Spector, 1992; see table 6.1). Although
the number of steps differs from author to author, the content of the process is very
similar. Scale development starts with defining the construct. Subsequently, the format of
measurement is determined and an initial item pool is generated. Next, this item pool is
reviewed so that clear and unambiguous items remain that measure the construct as
defined in the first step. Finally, the selection of items is administered and tested for its
validity and reliability. If necessary, in the final stages of scale development the scale
length can be optimized and norms can be established.
The first step in scale development is consistently recognized as defining what
exactly is to be measured. A scale cannot be developed until it is clear what it intends to
measure. The construct of interest must be clearly and precisely defined (Devellis, 1991;
Spector, 1992). This step is considered to be very important. Spector states that it is “one
of the most vital steps in the development of a scale” and that “may seem to be a simpleminded
requirement but it is at this step that many scale development efforts go astray” (p. 7). Churchill
(1979) also stresses the importance of this step. According to him, “the researcher must be
exacting in delineating what is included in the definition and what is excluded” (p. 67).
92
Table 6.1
Steps in scale development according to different authors
Author
Churchill (1979)
Steps
1. Specify domain of construct
2. Generate sample of items
3. Collect data
4. Purify measure
5. Collect data
6. Assess reliability
7. Assess validity
8. Develop norms
Devellis (1991)
1.
2.
3.
4.
5.
6.
7.
8.
Determine what to measure
Generate item pool
Determine format of measurement
Have initial pool reviewed by experts
Consider inclusion of validation items
Administer items to development sample
Evaluate the items
Optimize scale length
Spector (1992)
1.
2.
3.
4.
5.
Define construct
Design scale
Pilot test
Administration and item analysis
Validate and norm
After defining the construct, subsequent steps involve determining the format of
measurement and generating items to form an initial item pool. Devellis (1991) is the only
author who lists these steps separately. He also emphasizes that determining the format
of measurement should occur together with the generation of items so that the two are
compatible. The second step of Spector’s (1992) procedure, “designing the scale”, also
includes both the determination of measurement format and the generation of an initial
item pool. Churchill does not describe the format of measurement as a separate step. The
second step in his procedure is the generation of items. Once it is known what kind of
items the scale will include, one can start generating an initial pool of items. These items
should be created or gathered with the definition of the construct in mind. An initial item
pool consists of items that could be included in the scale. (Churchill, 1979; Devellis, 1991;
Spector, 1992).
After an initial item pool has been generated, then the items in the pool should
be reviewed. Churchill’s stage “generate sample of items” includes the careful editing of
the items. Devellis (1991) and Spector (1992) describe the reviewing of items as a separate
93
step. Devellis (1991) advises that the initial item pool be reviewed by experts. Spector
(1992) recommends a pilot test with a small number of respondents who are asked to
critique the scale. Both argue that the initial pool of items should be reviewed to measure
what it is supposed to measure (Devellis, 1991; Spector, 1992).
The final step in all procedures includes the collection of data and testing this
selection of items with respect to validity and reliability. Only Devellis (1991) mentions
the optimization of scale length. Churchill (1979) and Spector (1992) both include a step
for developing norms.
6.2
Steps in developing a product personality scale
The development of the product personality scale largely followed the steps
described above. The definition of product personality has been discussed in chapter one
and the defining components of the concept are delineated in chapter five. The format of
measurement will be discussed in the next section. Generating and reviewing the initial
item pool involved several small steps, which are described in more detail in sections 6.4
to 6.9. Figure 6.1 provides a graphic overview of this process. The item pools that remain
after each step are named with a capital. These capitals, used to refer to the different
pools in table 6.1, will also be used throughout the text.
We started generating an initial item pool by gathering scales, including those
discussed in section 5.3, that measure the personality of people or products in different
fields of study (product design, marketing, consumer research, human personality and
person perception). We then translated the items into Dutch (section 6.4 – pool A). Next,
items from qualitative research were added (section 6.5 – pool B) to form an initial item
pool of product personality (pool C). This pool was subsequently reviewed and reduced
with respect to the defining components delineated in section 5.2.
A defining component of product personality is that the scale should only
include items that describe human personality characteristics. The first step in reviewing
the initial item pool thus involved reviewing the items in relation to the definition of
human personality characteristics (section 6.6). Items that did not fit this definition were
eliminated. The remaining items (pool X) were reviewed with respect to similarity of
meaning (section 6.7). That is to say, we grouped together items that describe the same
personality characteristic and selected an adjective to represent this personality
characteristic. This procedure significantly reduced the item pool to include 458 adjectives
(pool Y).
The next review step involved another defining component of product
personality. The items should describe personality characteristics used with respect to
94
both people and products. By selecting only these personality characteristics that are used
to describe both people and products, a pool of 78 characteristics remained (section 6.8,
pool Z).
A final item selection procedure is conducted (section 6.9) to optimize scale
length. Respondents should be able to rate different stimuli in one study without loosing
interest. A scale of 78 personality characteristics is considered to be too long. We prefer a
scale consisting of 15-25 items because a scale of this length can be filled out several
times without loosing interest.
General steps
Generating item pool
Steps in generating and reviewing the item pool
Literature review
and translation
Qualitative research
Section 6.4
Pool A (# 348)
Section 6.5
Pool B (# 953)
Pool C (# 1142)
Reviewing item pool
Reviewing by definition
Section 6.6
Pool X (# 988)
Grouping together and
selecting adjectives
Section 6.7
Pool Y (# 458)
Selecting items relevant to
both people and products
Section 6.8
Pool Z (# 78)
Concept scale
Section 6.9
Pool O (# 20)
Figure 6.1: Overview of the steps taken in developing a product personality scale
6.3
95
Format of measurement
There are many different possibilities with respect to format of measurement.
No one possible format is superior to another. Choosing the right format depends on the
goal of measurement (Devellis, 1991). Deciding on format of measurement is thus a
process of substantiated decisions made in light of the purpose of the measurement. The
purpose of the product personality scale is assessing the personality of product variants
from different product classes in large scale studies.
We chose to use adjectives as items because adjectives are traditionally seen as
the primary terms of personality and frequently used as items in personality measures
(Hofstee, 1990; De Raad et al., 1988). There has been some discussion about using
personality descriptive sentences in personality assessment, instead of adjectives, because
adjectives are said to be too abstract. If sentences are used as items, then the items can be
more specific. Sentences can describe a statement (“The death penalty should be reintroduced into the Dutch justice system”), a behavior (“You donate to the anti-AIDS
fund”), an event (“Traffic controls in your neighborhood”) or a thing (“The shampoo
you are currently using”) (Spector, 1992). However, in order to measure product
personality, the relative abstractness of adjectives is an advantage rather than a
disadvantage. The level of abstraction makes it possible to use the same terms in
describing the personality of people and the personality of products. Sentences like
“..always controls the conversation” and “..not open to other peoples’ ideas” are too specific and only
apply to people. Yet, the adjective “dominant” is descriptive of both people and product
variants. Moreover, “dominant” may not only be descriptive of a car, but also of a
coffeemaker.
We decided to use single adjectives as items instead of adjective pairs
representing two opposites (e.g. “friendly” versus “hostile” or “happy” versus “sad”).
When using adjective pairs, the one end of the item scale gives meaning to the other end.
As a consequence, the meaning of the continuum may differ depending on the
combination of the adjectives. For example, the meaning of an item scale ranging from
“friendly” to “reserved” differs from the meaning of an item scale ranging from
“friendly” to “hostile”. This issue may become a problem with respect to personality
characteristics because not every personality characteristic has a clear negative counterpart
(Hofstee, 1990).
The problem of unclear opposites can be countered by using either unipolar
adjective pairs representing the absence versus presence of a single characteristic
(“friendly” versus “not friendly”) (Devellis, 1991) or single adjectives (“friendly”). The
problem with unipolar adjectives is that the neutral point never really is neutral with
personality characteristics (Hofstee, 1990). Although the assumption is that the items on
96
either sides of the scale represent the absence and presence of one personality
characteristic, “not friendly” does not represent the absence of “friendly”. In fact, it
represents its negative counterpart: the presence of “unfriendliness”. Therefore, we chose
to use single adjectives that can be rated as either “descriptive” or “not descriptive” of a
target product (see for an example figure 6.2). This way of measuring seems least open to
bias. Rating “friendly” as “not descriptive” of a target product appears more objective
than rating it as “not friendly”.
Can you please indicate how descriptive you consider the following personality
characteristics to be of the depicted toaster?
(Kunt u alstublieft aangeven in hoeverre u de onderstaande persoonlijkheidseigenschappen van
toepassing vindt op het getoonde broodrooster?)
Cute
Not descriptive
1
2
3
4
5
Descriptive
Not descriptive
1
2
3
4
5
Descriptive
Serious
Not descriptive
1
2
3
4
5
Descriptive
Tough
Figure 6.2: Example items of the product personality scale
6.4
After the concept has been defined and the format of measurement is
determined, the next step is to generate an item pool (Devellis, 1991). The item pool in
this research has to contain an exhaustive set of adjectives that describe personality
characteristics that are used to describe both people and products. As described in section
5.3, there are many existing scales that measure personality characteristics associated with
product variants. Although these scales are not valid measures of product personality, the
items that they contain may be relevant. The items of the scales that were discussed in
97
section 5.3 formed the beginning item pool (see appendix E). These scales all measure
constructs related to product personality, and originate from marketing and consumer
research. Since the item pool should comprehend the scope of personality characteristics
used with respect to products and people we expanded the pool with items from
literature dealing with human personality.
Literature dealing with human personality can be divided into the categories of
personality psychology and person perception. One of the goals of personality
psychology is to develop standard personality scales used to describe human personality.
The items that are included from personality psychology are gathered from these standard
personality scales and mostly stem from the stream of research that finally resulted in the
Five-Factor model (see section 2.2). The 16 primary factors of Cattell et al. (1977) are
included together with unipolar and bipolar adjectives that describe the five factor
dimensions (adopted from Carver & Scheier, 1996; Norman, 1963). In the case of bipolar
item scales, both items are included in the pool. The personality characteristics that
describe Eysencks’ (1975) two dimensions of personality have also been added to our
pool. Eysencks’ two-dimensional personality structure is not developed in the line of the
five-factor model, yet its two factors are highly similar to the first two dimensions of the
five-factor model (Costa & McCrae, 1992b). All these scales are also presented in
appendix E.
Research in the area of person perception is not concerned with finding a general
model of human personality. Instead, it investigates the process and content of person
perception (Zebrowitz, 1990). Asch (1946), for example, studied the influence of
changing one adjective in a short person description on the total impression of that
person. The adjectives he used to evaluate the persons are included in our pool.
Another way to study the content of person perception is to use free-response
methods. Presenting respondents with a stimulus person and asking them to describe
their impression of this person. The pool does not yet contain personality characteristics
from free-response data and, therefore, some free-response adjectives have also been
added to our pool (Anderson & Klatzky, 1987; Rosenberg & Sedlak, 1972). All the items
that are included in the item pool are shown in appendix E together with their references.
Translation of the item pool
The items gathered from literature are all in English and thus had to be translated
into Dutch. In order to make sure that the Dutch translation had the same meaning as
the English original, we followed a two-way translation procedure suggested by Mullen
(1995). According to this procedure, the original words should be translated into the
target language by a bilingual person, and be translated back again into the original
language by a second bilingual person. The bilingual person should be strongest in the
98
target language. A Dutch person who has good comprehension of the English language
translated the original English items into Dutch. Then, a native English speaker with
good comprehension of Dutch translated the resulting Dutch items back into English.
Any discrepancies between the original item and the back translation (e.g. jacket – jas –
coat, or sweet – lief – dear) indicate a problem. The differences between the two English
versions were discussed until a consensual translation was agreed upon. When no
satisfactory Dutch translation was available, then the personality characteristic was
omitted from the pool. The final Dutch pool contained 348 personality characteristics
(pool A) and is shown in appendix F.
6.5
Personality characteristics from qualitative research
A good set of items represents the complete range of items relating to the
construct. In order to get a complete overview, Devellis (1991) advises including
considerably more items in the initial pool than are planned to be included in the final
scale. The final pool, (pool A), as it resulted from literature, is quite comprehensive and
exhaustive. However, it does not yet include personality characteristics that people use to
describe and discriminate between product variants. In fact, none of the scales studied
product personality at this level. The personality characteristics in the product personality
scale must differentiate between different product variants. Therefore, we set out to
expand pool A by adding personality characteristics explicitly used to describe product
variants.
In order to include personality characteristics explicitly used to describe product
variants, we added the personality characteristics that originate from the first study
reported in chapter 4 (the method is discussed in detail in section 4.2.1). The product
personality of the stimuli in that study was established by asking 48 respondents to
describe them “as if they were a person”. The interviewer wrote down the personality
descriptions of each product variant for each respondent. In order to guarantee the
generalizability of the product personality measure, we used a diversity of product
variants from different product classes as stimuli: 3 coffeemakers, 3 screwdrivers, 3 soapdispensers and 3 bottles of wine (see figure 4.1, page 66).
The number of words and descriptions respondents used to describe the
personality of the stimuli amounted to 2732. This amount included descriptions that were
used several times, various conjugations of one verb and quantity descriptions. The
indicators “little”, “less”, “very”, “maybe”, “not”, “somewhat”, “a lot”, “reasonably”,
“sometimes”, “super”, “not always”, “a touch”, “a bit”, and “none” were deleted and
every word or description was only included once. In case of redundant conjugations,
99
such as “attracting attention” and “attracts attention”, only one was included. The nonredundant pool, which resulted after this review process, contained 953 words and
descriptions that could serve as items of product personality (pool B, appendix G).
The items that resulted from the qualitative study (pool B) encompassed a
broader vocabulary than the results from literature and are colloquial (e.g. “horrible” and
“repulsive”). The items resulting from literature (pool A) seem more formal (e.g.
“dismissive” and “hypochondriacal”) and more specifically related to people (e.g.
“concerned”). Both pools of items were combined to form the initial item pool (pool C).
Yet, pool C is not a simple summation of pool A (#348) and pool B (#953), there was
some overlap. Many of the items from literature also showed up in the qualitative study.
Items that were present in both pools were only included once in pool C, so that pool C
resulted to contain 1142 items. Since this a substantial number of items, we assume that it
represents the full spectrum of personality characteristics. Further elaborations would
probably only provide more synonyms of words and descriptions that are already
included.
6.6
Reviewing the items with respect to definition
The goal of the previous steps was to create an exhaustive item pool containing
human personality characteristics. However, pool C contained items that, according to
the definition of human personality characteristics, should not be included. Not every
scale from which the items were included based their selection of items on the definition
of human personality characteristics. Furthermore, the respondents in the qualitative
study did not solely use human personality characteristics to describe the stimuli “as if
they were a person”. The first step in reviewing the initial item pool (pool C) was to
review the items with regard to the definition of human personality characteristics. The
items that did not describe a human personality characteristic were removed from the
pool.
According to Hofstee (1990), two general selection methods can be used in
delineating personality characteristics: (1) expert judgment and (2) lay judgment. Expert
judgment is often used in the early stages to exclude certain classes of personality
descriptions. Lay judgment is used more often in the final stages to establish the difficulty
of items and to make a final selection. In order to select items that describe human
personality characteristics, we used expert judgment. Five experts were asked to review all
of the items of pool C and to categorize them into four groups. Only one of the four
groups was to be comprised of personality characteristics.
100
Reviewers
Two female and three male employees of the school of Industrial Design
Engineering at Delft University of Technology participated voluntarily. One of them had
an academic background in industrial design engineering, one in psychology, one in
psychology and industrial design engineering, one in communication, and one in
marketing. Their ages ranged from 24 to 32, with a mean age of 26.
Stimuli
Each item was printed on a card of 7.5 by 10 centimeters. These cards were put
in a box from which the reviewer randomly selected them one by one.
Procedure
The reviewers sat at a large table and were provided a written instruction. The
instruction asked them to divide the items into four groups: (1) “personality
characteristics” (defined as “a quality of a person that discriminates him/her from others”), (2)
“physical qualities” (defined as “visible, physical, appearance qualities of a person”), (3) “other”
(items not describing a physical or a personality characteristic), and (4) “I do not know”
(for nonsense words). They were told to use the first three categories as much as possible
and only to use the fourth category if they really had no idea about the meaning of an
item. Further instructions included an assurance that there were no right or wrong
answers and an explanation that the set contained several synonyms, which might create
the impression that they had to categorize the same words several times. Completing the
categorization task took the respondents approximately two hours. They were allowed to
take a break whenever they so desired.
Results
The five reviewers consensually categorized 257 of the 1142 items as a
personality characteristic. Hundred-and-fifty-nine items were not categorized as
personality characteristics by any of the reviewers. Since redundancy was considered to be
an asset rather than a liability in the current stage (DeVellis, 1991), it was decided to use
the most conservative elimination strategy possible. We only eliminated those 159 items
that none of the reviewers categorized as a personality characteristic. This way items that
were considered to be a personality characteristic by at least one of the reviewers
remained in the pool.
In order to reduce the chance that valuable items were wrongfully eliminated, an
extra reviewer categorized the 159 potentially discarded items. This reviewer has a
background in industrial design engineering and psychology. The items that were
categorized as personality characteristics by this reviewer (“desolate”, “pale”, “colorful”,
101
“sticky” and “sexy”) were included in the pool after all. After this procedure, a pool of
988 items remained (pool X). These items were each considered to be a personality
characteristic by at least one of the (six) reviewers.
6.7
Grouping together items with the same meaning
The second step in reviewing the items was putting together those items that
describe or relate to the same personality characteristic. Pool X contained a lot of
synonyms and manifestations of more abstract personality characteristics. As a result of
this step, a pool of distinct personality characteristics should remain. Another
categorization task was carried out in order to reach this goal. The nature of this
categorization was natural grouping; putting together items that, according to the
respondent, belong together. However, in order for a natural grouping procedure to
work, respondents need to comprehend the complete set of stimuli, in this case 988
items. Since this is impossible, we designed a group categorization task.
Reviewers
The five reviewers who had already participated in the first categorization task
(described in the section 6.6) were invited to participate again. These reviewers were
familiar with the personality characteristics and could be expected to have a better
overview than people who would have to familiarize themselves with the personality
characteristics for the first time. They received a gift after the completion of the task.
Stimuli
The items had already been written on cards for use in the categorization task
described in section 6.6. The cards of the remaining 988 items were used again in this
task.
Procedure
The group categorization procedure took place during working hours and lasted
two-and-a-half days. On the first day of the session, all five reviewers participated. For
the second day-and-a-half, only three reviewers participated (two males and one female).
The author acted as discussion leader on both occasions.
Approximately a week before the session started all reviewers were handed an
equal amount (+/- 250) of randomly selected cards. They were asked to familiarize
themselves with the items and to group them according to meaning. This allowed them to
have a clear overview of their cards at the start of the session.
102
The actual categorization session took place around a large table in a conference
room. Reviewers were given the space to lay down their cards on the table. Once all the
participants were ready, one of them stated a personality characteristic and each of the
other participants could put in related descriptors from their portion. They were
instructed to group together: (1) synonyms, (2) manifestations, behaviors, and examples
belonging to the same personality characteristic and (3) descriptors/words that hardly add
meaning to each other. Once a category was constructed, a word to represent the group
was chosen. The decision to in- or exclude a word in a category and the choice of the
representing word were taken by a majority of votes.
Results
Pool X, which contained 988 items, was reviewed with respect to meaning. Items
that described the same personality characteristics were combined into one category and
represented by one item (see table 6.2 for some examples). These representing items were
selected to form the next item pool (pool Y). However, the words chosen to represent a
personality characteristic were not always adjectives. Since we had decided to use
adjectives as items (see section 6.3), a post-hoc selection procedure followed. Two other
reviewers (one male and one female) evaluated the categorization. The non-adjectives
were either rewritten as an adjective or replaced by an adjective from the same category.
For example, “has good taste” was replaced by “tasteful”, and “to no use” was rewritten
as “useless”. Categories that could not be described by an adjective were removed from
the pool. From this final selection procedure, an item pool of 458 items resulted (pool Y).
Table 6.2
Examples of categories of personality characteristics (in English and Dutch)
•
Unusual
Different, odd, peculiar, dissimilar
Apart
Anders, vreemd, bijzonder, verschillend
•
Intelligent
Clever, intellectual, smart, bright
Intelligent
Knap, intelluctueel, slim, snugger
•
Honest
Sincere, straightforward
Eerlijk
Oprecht, recht door zee
6.8
103
Selection of items used to describe both people and products
This section describes the selection of items that are actually used with respect to
people and products, and refer to personality characteristics in both situations. Pool Y
has not been reviewed in this respect. It may include items that are mainly used to
describe people but are hardly ever used to describe a product (e.g. “loyal”, “deceitful”).
On the other hand, it may include items that are used very often to describe products but
occur less frequently in descriptions of people (e.g. “refined”, “classic”). The definition of
product personality requires that the items in the product personality scale should be
relevant to both people and product descriptions. Therefore the next review step involves
the selection of personality characteristics used to describe both people and products.
The selection procedure involved two groups of respondents who indicated
whether they used each item with respect to people (group I) or with respect to product
variants (group II). By combining the results of these two groups, we could identify the
items used in both situations. An advantage of this procedure was that it minimized the
chance of including jargon (Spector, 1992).
Respondents
Two questionnaire studies were conducted. One of the questionnaires was used
to establish which personality characteristics are used for describing people. The goal of
the second questionnaire was to find out which personality characteristics are used for
describing product variants. The questionnaires were filled out by two (different) groups
of 78 students from Industrial Design Engineering. Group I filled out the questionnaires
regarding the use with respect to people. Group II filled out the questionnaires regarding
the use with respect to product variants. The ages of the respondents in group I varied
from 19 to 29 with an average age of 21. Forty (51%) of them were male and 38 (49%) of
them were female. The ages of the respondents in group II also ranged from 19 to 29
with an average age of 21. In this group forty-nine (63%) respondents were male and 29
(37%) were female.
We explicitly chose to use students from Industrial Design Engineering as
respondents because we wanted to reduce the chance of falsely excluding personality
characteristics that are used to describe products. Students from Industrial Design
Engineering are trained to put their impression of products into words and have
developed a broad vocabulary for describing product variants. It can be assumed that
they use more personality characteristics in their product descriptions than a layperson
would. The chance that a personality characteristic would wrongfully be identified as
irrelevant for describing products was thus minimized. Since the results of group I and
group II were to be combined to select the relevant items, we preferred comparable
104
samples. The respondents in group II (answering the “people” questionnaire) were also
students from Industrial Design Engineering. This is thought not to affect the results
because Industrial Design students are just as familiar with describing other people as any
other layperson.
Design
It was impossible for a single respondent to fill out a questionnaire containing all
of the 458 items in pool Y. Therefore, the pool was randomized 13 times. This resulted in
13 different sequences. Each sequence was then divided into six short lists of
approximately 76 items. For each sequence, the first list of 76 items formed a
questionnaire. The next 76 items formed a second questionnaire, and so on. This method
resulted in 78 unique questionnaires (13 * 6) that were easy to fill out. This method also
leveled out possible order and set effects. These 78 questionnaires were used in both
groups. The only difference between the questionnaires about people and the
questionnaires about products were the instructions.
Procedure
Both groups of respondents filled out the questionnaires during class hours. In
both cases the questionnaire was preceded by an introduction about personality
characteristics and how people use them to think and talk about other people. For group
II a passage was added about how people tend to do the same with products. It was
emphasized that when personality characteristics are used to describe products then they
should be used to describe the product itself and not the designer or user. It was also
stressed that the items in the questionnaire should be interpreted to mean the same thing
when used to describe products as when they were used to describe people. After this
introduction, the questionnaires were handed out to the respondents. These instructions
were printed on the front page of each questionnaire.
The questionnaires each contained approximately 76 items. Respondents were
asked to indicate whether they used an item to describe the personality of a person or
product by marking a box representing “yes” or “no”.
Results
Two different groups of 78 respondents indicated whether they used the
remaining 458 items (pool Y) to describe people or products. Each item was judged 26
times, 13 times with regard to persons and 13 times with regard to products. The
frequency with which an item was said to be used in each situation was calculated and a
cut-off score was determined. This cut-off score fulfils two criteria. The first criterion is
105
the minimal frequency with which an item was used to describe persons. The second
criterion is the minimal frequency of usage with respect to products.
The two criteria are not equally important. The fact that the items in the product
personality measure are used to describe the personality of people is a necessity. Including
items that are not used to describe the personality of people would undermine the validity
of the product personality measure. The measure would not measure what was intended:
the personality of products described with personality characteristics that could also be
used to describe the personality of people. The inclusion of items that are not used to
describe products does not affect the validity of the measure because the personality
characteristics in the final measure can be rated from “not descriptive” to “descriptive”.
Irrelevant characteristics are thus always rated as “not descriptive”. It is however not
desirable to include items that nobody uses to describe products. It would make the
product personality measure needlessly long. The criterion for including items that are
used to describe people is thus set higher than the criterion for the items that are used in
reference to products.
The cut-off score is set at 85% and 60%, respectively. Items that at least 60% of
the respondents said to use to describe the personality of products are included, if they
are also used by at least 85% of the respondents to describe the personality of persons.
This means that an item was selected if at least 11 of the 13 respondents used it with
respect to people and more than 8 of the 13 used it with respect to products.
The resulting selection of items meets the defining component of product
personality in that they are used with respect to people and products. However, the
criterion does not rule out the possibility that items, though used with respect to both
products and people, do not measure the personality of a product. Some items are
interpreted literally when used with respect to products. For example, “handig” is a
personality characteristic when it is interpreted as “skillful”. Yet, its primary interpretation
becomes “convenient” when it is applied to a product. Three reviewers (two males and
one female) identified seven items (reported in table 6.3) that could be interpreted literally
when used to describe a product and eliminated them from the selection. After this
review procedure a pool of 78 adjectives remained (pool Z, see also table 6.4).
106
Table 6.3
Items discarded because of literal interpretation if used to describe products (in Dutch)
Meaning with respect to
Meaning with respect to
Dutch item
people
products
Fel
Showing strong feelings about
Brightly colored
something
Grof
Rude and offensive
Rough fabric or consisting of large
pieces
Handig
Skillful, clever
Convenient, useful
Makkelijk
Easy-going, relaxed
Simple, uncomplicated
Onhandig
Clumsy
Unpractical
Sterk
Not easily upset
Robust, not easily broken
Stil
Introverted
Silent, making very little noise
Table 6.4
The selected items of product personality (pool Z, in English)
Sensible
Lively
Decent
Aggressive
Sensitive
Masculine
Dominant
Aloof
Serious
Mature
Easy-going
Annoying
Showy
Modest
Eccentric
Attractive
Nice
Silly
Boring
Energetic
Strict
Excessive
Obtrusive
Bourgeois
Odd
Sweet
Businesslike
Exuberant
Terrific
Old-fashioned
Feminine
Calm
Tough
Open
Flexible
Careless
Unattractive
Pathetic
Friendly
Casual
Uninteresting
Peculiar
Funny
Chaotic
Unreliable
Pleasant
Happy
Charming
Untidy
Popular
Honest
Cheerful
Vulnerable
Precise
Cheerless
Idiosyncratic
Well-groomed
Predictable
Immature
Childish
Wild
Inconspicuous
Pretty
Consistent
Provocative
Young
Conspicuous
Informal
Relaxed
Insular
Corny
Reliable
Intelligent
Creepy
Reticent
Interesting
Cute
Romantic
6.9
107
Selection of items to form the concept scale
Pool Z consists of 78 items that are all valid items of product personality; they
are personality characteristics used to describe the personality of both people and
products. However, the amount of 78 is not considered manageable. The scale will be
used to rate different product variants in one study. Therefore, respondents always have
to rate the amount of items times the number of different product variants. For this
reason, we preferred an amount of 15-25 items for measuring product personality (see
section 6.2).
We believe that is should be possible to reduce the set of 78 items without
loosing the range of meaning that is represented by these items. It is known that
judgments of the presence of some human personality characteristics almost always go
together with the perceived presence of other personality characteristics (Zebrowitz,
1990). For example, a person that is perceived as “intelligent” is often also perceived as
“successful”. If these two personality characteristics are always perceived together, then
“successful” does not provide extra differential information to “intelligent”. In a sense,
the two characteristics describe the same thing.
To determine the final set of product-personality characteristics, we explored
whether or not there were groups of items that could go together. In that case, the
product personality scale could suffice with less than 78 items.
6.9.1
Method
Respondents
A convenience sample6 of 147 students was asked to fill out the questionnaire
and 125 of them (85%) responded. Forty-two of them were male (34%) and 82 were
female (66%). One respondent did not indicate his/her gender. The ages of the
respondents ranged from 19 to 36, with an average of 24. The respondents came from 38
different university schools. Since product designers have been shown to perceive
products differently (Hsu, Chuang & Chuang, 2000), students of Industrial Design
Engineering were excluded from participation.
6
Due to circumstances, we could not use respondents from the consumer household panel of
Delft University of Technology. Although using a student sample was not our primary choice, it
is assumed not to impair our results. The product-personality characteristics mean the same to
students as they do to other people.
108
Stimuli
The stimuli in this study were selected from two different product classes. These
product classes were chosen to represent one of the most frequently used distinctions in
studies about personality associations with products (Aaker, 1997; Sirgy, 1982): high
personalizing products (products with high symbolic value) and low personalizing
products (products with low symbolic value). The extent to which a product has symbolic
value is partly determined by the type of consumption: public or private (Holman, 1981).
Publicly consumed products generally have more symbolic value than privately consumed
products. In order to select two product classes that are typically seen as high
personalizing or low personalizing products, we conducted a pilot study. Fourteen
students were asked to indicate a publicly consumed product class that they thought of as
highly symbolic. Cars were mentioned most frequently; 93% of the respondents indicated
cars as a good representative of a highly symbolic, publicly consumed product. The same
respondents were also asked to indicate a privately consumed product class that they
thought of as low symbolic. Staplers and scissors were mentioned most often (43%).
Unfortunately, these product classes show very little variety in appearance. Lack of
diversity in appearance could cause items in the main study to be paired together for the
wrong reasons because they are consistently rated as “not descriptive”. Vacuum cleaners
were the next best mentioned product class and come in a great variety of appearances.
Of the respondents, 29% indicated vacuum cleaners to be a low symbolic, privately
consumed product. Therefore, cars and vacuum cleaners were chosen as product classes
from which the stimuli were selected.
Sixteen stimuli, eight cars and eight vacuum cleaners, were selected from the
current (2001) assortment of product variants within a pre-set price range. The price
range was set at midrange (vacuum cleaners: € 136,- to € 409,-; cars: € 9.091,- to €
20.455,-7).
Brochures of the main suppliers were gathered from car dealers and
manufacturers of vacuum cleaners. A set of good quality color pictures was selected from
these brochures. The pictures that served as stimuli were selected to represent a variety in
appearances and to provide a realistic picture of alternatives within the pre-set price
range. The stimulus pictures are presented in appendix H.
The stimulus pictures all provide the respondent with the same information. The
pictures of the vacuum cleaners show the tube and the piece. The pictures of the cars
show the complete front and side of the cars. The possibility of respondents recognizing
the brand could not be ruled out. To minimize possible brand effects, the brand names
and logos were removed from the pictures.
7
The prices are adapted from 2001 when pricing was in Dutch guilders. The price range in
Dutch guilders was fl.300,- to fl.900,- for vacuum cleaners and fl.20.000,- to fl.45.000,- for cars.
109
Design
Rating 16 stimuli on 78 items would take respondents more than one hour. We
did not expect respondents to maintain their concentration for such a long time.
Therefore, the stimuli were divided into four sets, each set containing four stimuli. This
resulted in two sets of vacuum cleaners and two sets of cars. The sets were formed so
that each set represented as much variation in appearance as possible. The pictures in
appendix H are arranged according to the sets. The sets were randomly assigned to
respondents. Within each set the stimulus pictures were presented in four balanced
orders, according to the digram-balanced square method (Maxwell & Delaney, 1990).
Procedure
Respondents received an envelope containing a questionnaire and a sheet with
four numbered stimulus pictures. The first page of the questionnaire instructed
respondents to start rating the product variant in picture 1. When finished rating one
picture, the questionnaire indicated that the respondent could proceed on to the next
picture. The questionnaire thus presented the 78 items from pool Z four times. The items
were randomized and depicted in the same order all four times. All items were rated using
5-point scales (1 = “not descriptive” and 5 = “descriptive”). Filling out the complete
questionnaire took approximately 30 minutes. Upon returning the questionnaire,
respondents received a lottery ticket.
6.9.2
Results
A sample of 125 respondents rated the product personality of 16 different
product variants, divided into four stimulus sets of four product variants. The number of
respondents that rated each set is shown in table 6.5. Gender, stimulus set and product
class are all specified. The number of respondents and the distribution of males and
females were approximately similar in all sets. Each set included about 30 respondents, of
which approximately two-thirds were female, and one-third was male.
Table 6.5
Number of males and females specified per stimulus set and product class
Males
Females
Total
11
22
33
Set 1 (vacuum cleaners)
12
18
30 (+ 1 missing)
Set 2 (vacuum cleaners)
Set 3 (cars)
9
19
28
Set 4 (cars
10
23
33
Total
42
82
124 (+ 1 missing)
110
The aim of this study was to explore whether there are groups of items that cooccur in the perception of product personality. We analyzed the data using Ward’s
hierarchical clustering method. Cluster analysis classifies objects into clusters so that each
object is very similar to the other objects in that cluster. Cluster analysis can be
performed using different algorithms. Punj and Stewart (1983) reviewed the different
clustering algorithms and suggested that Ward’s algorithm outperforms other clustering
algorithms. Ward’s clustering algorithm clusters objects by optimizing the minimum
variance within clusters (Hair et al., 1998).
The objects to be clustered in this study were the items. The analysis of the items
was guided by the argument that items co-occurring in the descriptions of different
products and by different respondents could reliably be replaced by one item. The
analysis consisted of two stages. The first stage in the analysis explored whether or not
there were large clusters of items. In the second stage, these clusters served as a base for
reducing the item pool.
Stage 1: Looking for large clusters
In the first stage, the dataset was split up into different subsets. The different
subsets were cluster analyzed separately. The results of the separate analyses were then
compared. This procedure is common as a way of ensuring that the cluster solution is
valid and generalizable for all objects (Hair et al., 1998). Our solution had to be
generalizable for all respondents and for different product classes. Therefore, we first
conducted two cluster analyses after dividing the data set into a subset containing the
male respondents’ ratings and a subset containing the female respondents’ ratings. The
dataset was reunited and split up again, but this time into a subset containing the ratings
of the vacuum cleaners and a subset containing the ratings of the cars. Again, two cluster
analyses were conducted on these two subsets.
After this procedure, four cluster solutions remained: one based on the male
ratings, one based on the female ratings, one based on ratings of the vacuum cleaners and
one based on the ratings of the cars. In order to see if similar groups of items would
appear in all four solutions, we inspected the dendograms and compared them to each
other. The dendograms are added in appendix J. It seemed that in every solution there
were three large clusters, each made up of six groups of items that always appeared
together in one cluster. The alphas of these groups all exceed .80, and all the intraclass
correlation coefficients8 (ri) are significant (p < .001). This implies internal consistency of
the groups. The alphas and ri’s of the groups are shown in table 6.6. In each of the
dendograms in appendix J, the six groups are depicted in different colors.
8
The ri (intraclass correlation coefficient) is a measure of consistency or agreement of values
within cases (SPSS Help).
111
Table 6.6
Statistics of the six groups of items in the concept product personality scale
Careless
Conspicuous Aloof
Attractive Aggressive Bourgeois
Dominant Businesslike
Chaotic
Eccentric
Boring
Casual
Childish
Exuberant
Cheerless
Charming Excessive Calm
Obtrusive Consistent
Corny
Funny
Insular
Cheerful
Showy
Decent
Creepy
Idiosyncratic Masculine
Cute
Honest
Immature Interesting
Old-fashioned
Easy-going
Inconspicuous Odd
Lively
Reticent
Energetic
Pathetic
Mature
Provocative Strict
Feminine
Modest
Silly
Tough
Unattractive
Flexible
Unreliable Wild
Precise
Uninteresting
Friendly
Untidy
Predictable
Happy
Reliable
Informal
Sensible
Nice
Serious
Open
Well-groomed
Pleasant
Popular
Pretty
Relaxed
Romantic
Sensitive
Sweet
Terrific
Young
α = .96
α = .83
α = .88
α = .84
α = .91
α = .89
ri = .51**
ri = .53**
ri = .32**
ri = .34**
ri = .50**
ri = .45**
** Correlation coefficient significantly differs from zero, p < .001.
Three items fall outside of these groups: vulnerable (“kwetsbaar”), intelligent
(“intelligent”), and annoying (“vervelend”). The position of these characteristics changes
from solution to solution. They do not consistently appear together with the same items.
This means that the way they are used to describe products is not similar for all product
classes and respondents. Since we cannot be sure what we would be measuring if we
included these three items in the scale, we chose to delete them from the set.
Stage 2: Reducing the item pool
Six large groups of items, which seem internally consistent, appear in the data.
We could reduce the amount of 78 items by selecting one item from each of these
groups, leaving only 6 items. However, this would not do justice to the broad scope of
meaning represented by the items in each group. The diversity of the set of items caused
the cluster analysis to format clusters at a very general level. We thus set out to cluster
analyze the six groups again, individually. Ward’s algorithm was used in all six analyses.
112
The dendograms that resulted from the six individual cluster analyses are
depicted in appendix K. It appeared that the 6 groups of items could be divided into 20
sub-groups containing two to seven items each. The items in the sub-groups have clearly
related meanings. For example, “provocative” (“uitdagend”), “wild” (“wild”), “exuberant”
(“uitbundig”) and “eccentric” (“excentriek”) form one sub-group. “Dominant” (“dominant”)
and “aggressive” (“agressief”) form another. From each of these 20 sub-groups an item is
chosen to represent the sub-group. The amount of 20 items seems manageable and falls
within the pre-set range of 15-25 items.
The selection of items from the groups should be based on a criterion, but all 78
items are good, valid items. No unambiguous selection criterion is available to select the
items. Therefore, the items were selected based on a combination of criteria. The selected
item preferably had an item-to-total correlation that exceeded .50 (Hair et al., 1998). The
item with the highest item-to-total correlation in the sub-group was preferred. Yet, if the
sub group only contained two items, then the item-to-total correlations could not be
computed. Moreover, the selected items had to cover the meaning of the other items in
the sub-group.
The items that were selected are presented in table 6.7 (in appendix K they are
shown in bold print). Table 6.7 also reports the alphas and ri’s per selected group.
Together, these 20 items form the concept scale of product personality. All but two items
had high item-to-total correlations (ranging from .55 to .81), which indicates internal
consistency (Hair et al., 1998). Modest and honest had low item-to-total correlations,
respectively .39 and .44. As a consequence, the internal consistency of the groups as
measured by the alphas and ri’s is also low for the group that contains these items. In
general, the selection of items can be considered internally consistent. Though the alphas
decreased, the ri’s stayed the same and some even increased. This may be due to the fact
that Cronbach’s alpha increases as the number of items increases (Hair et al., 1998). The
groups with the lower alphas are the groups that contain the least items.
Table 6.7
Statistics of the six groups of items in the concept product personality scale
Selected Cheerful
Dominant Modest
Silly
Idiosyncratic Aloof
items Open
Obtrusive Honest
Childish Interesting
Boring
Relaxed
Serious
Untidy
Lively
Pretty
Provocative
Easy-going
Cute
α .86
ri .53**
.71
.50
.76
.82
.68
.55**
.25**
.41**
.54**
.52**
6.9.3
113
Discussion
The goal of this study was to reduce the set of 78 product-personality
characteristics to a more manageable number. It was argued that if there were groups of
items that co-occurred, then only one item of each had to be included. The other items
were said not to provide much additional information. From cluster analysis of the
results, it appeared that there are six large groups of items that co-occur in the ratings of
the vacuum cleaners, the ratings of the cars, the ratings by females and the ratings by
males. These six groups were cluster analyzed again and further divided into 20 subgroups that are represented by a selection of 20 items. This final set of 20 items because it
is manageable as it falls within the preferred amount of items (15-25) and it has face
validity as to representing the scope of meaning of the whole set.
The selected items of product personality were compared to the classification of
Dutch personality vocabulary into the Big Five dimensions (Doddema & De Raad, 1997).
It appeared that the 20 items described the extremes of the first three dimensions of
human personality: extroversion, agreeableness and conscientiousness. These same three
dimensions were found to be comparable to three of the brand personality dimensions
(Aaker, 1997). If categorized into the six groups that formed the basis of the selection,
then the groups could subsequently be labeled as describing “agreeableness”,
“disagreeableness”, “conscientiousness”, “carelessness” (antonym of conscientiousness),
“extroversion”, and “introversion” (see table 6.7). The results of our studies give no
indication of the existence of two more dimensions that resemble the remaining two
dimensions of human personality, neuroticism and openness to experience. The absence
of neuroticism and openness to experience in product personality could be caused by the
fact that neuroticism and openness are parts of personality that only become apparent in
interaction (Kenny et al., 1994). Since we studied still images of products, we excluded
interaction dimensions. By doing so, we probably ruled out the possibility of finding
aspects of neuroticism and openness to experience in product personality. However,
finding three of the five dimensions, the same three that apply to brand personality
(Aaker, 1997), strengthens the idea that we have selected items that are a valid
representation of the scope of personality characteristics used to describe products.
Limitations
A limitation of this study concerns the analysis of the data. We performed a
cluster analysis on different subsets of the data and compared the resulting dendograms
to find groups of items that co-occur in the same cluster. It would have been better if all
four cluster analyses had resulted in a six cluster solution containing the same set of
items. We did find the same groups of items reappearing in a cluster together, but the
114
groups were not so robust that they appeared as a single cluster in every solution.
Therefore, we did find consistency, though it was not as clear-cut as it ideally would have
been. To find out whether this solution really is consistent, the study should be repeated
with a different set of products and on different samples of respondents.
A second limitation concerns the selection of the 20 items. Although the
selection was done with care, it may not be concluded that this is the only right selection.
All 78 items are valid items of product personality and we selected a manageable number
to be used in a scale for practical reasons. This does not mean, however, that any other
selection will suffice. The results indicate a structure of groups of items. Each of these
groups should be represented in a selection, as it is represented in the current selection.
A final limitation of the study is the fact that the items are only reviewed and
selected based on validity requirements. The items all meet the defining components of
product personality. However, in order to serve as a product personality scale the
selection should also be reliable. To be reliable, the scale needs to provide a personality
profile of a product variant that is consistent for different respondents. The items in the
scale also need to be consistent. That is, respondents need to agree about the applicability
of a personality characteristic with respect to different product variants. If one
respondent perceives product variant A as “serious” and product variant B as “less
serious”, then the opinion of another respondent should be comparable. Therefore, we
calculated Cronbach’s alpha and the ri for every product (consistency of product
personality profile), and for every selected item (consistency of item). The results are
depicted in table 6.8 and 6.9. The numbers in table 6.8 correspond to the numbers in
appendix H.
Table 6.8
Reliability of the scale calculated for each product variant
Stimuli
α
ri
n
Vacuum cleaner 1
.96
.40**
33
Vacuum cleaner 2
.96
.42**
33
Vacuum cleaner 3
.97
.46**
33
Vacuum cleaner 4
.95
.36**
33
Vacuum cleaner 5
.91
.24**
31
Vacuum cleaner 6
.67
.06**
31
Vacuum cleaner 7
.84
.14**
31
Vacuum cleaner 8
.93
.30**
31
Car 1
.96
.46**
28
Car 2
.76
.11**
28
Car 3
.72
.08**
28
Car 4
.93
.32**
28
Car 5
.90
.21**
33
Car 6
.89
.19**
33
Car 7
.97
.52**
33
Car 8
.96
.43**
33
* Correlation coefficient differs significantly from zero, p < .05.
** Correlation coefficient differs significantly from zero, p < .001.
Selected items
Cheerful
Open
Relaxed
Pretty
Easy-going
Cute
Dominant
Obtrusive
Silly
Childish
Untidy
Idiosyncratic
Interesting
Lively
Provocative
Modest
Honest
Serious
Aloof
Boring
Table 6.9
Reliability of the items (n = 33)
α
.96
.93
.89
.93
.94
.95
.89
.90
.69
.91
.78
.94
.88
.91
.92
.87
.67
.95
.91
.95
115
ri
.46**
.31**
.22**
.34**
.37**
.41**
.22**
.24**
.07**
.26**
.11**
.36**
.21**
.26**
.29**
.20**
.07**
.39**
.27**
.39**
It appears that most product variants, as well as most items, have high reliability scores (α
>.8 and ri significant at p < .001). One vacuum cleaner (vacuum cleaner 6), and two items
(“honest” and “silly”) have relatively low reliability scores (vacuum cleaner 6, α = .69, and
ri = .06, p < .05; silly, α = .67 and ri = .08, p < .001; honest, α =. 69 and ri = .06, p < .001).
However, even these alphas exceed .60, which is considered to be the lower limit (Hair et
al., 1998), and the ri’s are significant. Consequently, it may be concluded that the selection
of items provides reasonably reliable results.
6.10
Conclusions
The goal of this chapter was to develop a scale for assessing product personality.
The development of this product personality scale encompassed different steps, which
were outlined at the beginning of the chapter. We started to generate an exhaustive item
pool by gathering items from literature and by adding items that resulted from a
qualitative study into product personality. This resulted in an item pool containing 1142,
which was subsequently reduced to a collection of 78 items. These 78 items can be
116
considered valid measures of product personality. From these 78 items, a manageable set
of 20 items was selected to serve as a concept product personality scale.
The items in this concept scale are especially selected for measuring product
personality, defined as the human personality characteristics used to describe and
differentiate between durable consumer product variants. This selection procedure
ensures that the selected items are valid and reasonably reliable.
Since the scale is has been developed for measuring the personality of durable
consumer products, it cannot be used with respect to fast-moving consumer products. It
is also not meant for measuring the personality of people. Human personality scales are
not valid when used with respect to products (Kassarjian, 1971) and a product personality
scale is not valid when used with respect to people.
Another important issue that needs to be noted is that the product personality
scale is neither reflective nor formative. A reflective scale is used to measure a latent
variable that can be present or absent to a certain degree (e.g. IQ, need for achievement).
A formative scale measures the components of a construct, which together determine the
construct (e.g. education, income, occupation and residence determine social economic
status) (Diamantopulous & Winklhofer, 2001). Product personality is not reflective
because it is not something that one product has more or less than another. All products
have personality, only the content differs. Product personality is also not formative. A
change in the score for one of the items of product personality leads to a change in the
product personality. Yet, a change in one of the components of a formative construct
does not necessarily lead to a change in the construct itself. For example, a decrease in
income does not necessarily lead to a decrease of social status. All of the items of product
personality are equally important in determining a product’s personality. Product
personality is a profile of personality characteristics. In order to see what a product
personality profile looks like, we calculated the profile of two vacuum cleaners, and two
cars, based on the data of the study reported in section 6.9. The profiles of the vacuum
cleaners are shown in figure 6.3 and the profiles of the cars are depicted in figure 6.4.
low
high
cheerful
open
relaxed
pretty
easy-going
cute
idiosyncratic
provocative
interesting
lively
dominant
obtrusive
untidy
childish
silly
boring
aloof
serious
honest
modest
●
W
●
W
●
W
W
●
●
W
●
W
●
W
●
W
●
W
●
W
●
W
W
●
●
W
●
W
●
W
●
W
●
W
●
W
●
●
W
W
Figure 6.3: Personality profiles of two vacuum cleaners
low
cheerful
open
relaxed
pretty
easy-going
cute
idiosyncratic
provocative
interesting
lively
dominant
obtrusive
untidy
childish
silly
boring
aloof
serious
honest
modest
high
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
Figure 6.4: Personality profiles of two cars.
117
118
The underlying goal of developing a product personality scale was to be able to determine
the personality profile of different product variants. These product personality profiles
help us to gain insight into the product characteristics associated with a personality
characteristic. The assumption is that product variants that are described with the same
personality characteristics share certain appearance characteristics. The last study reported
in this chapter (section 6.9) showed some evidence that this might indeed be the case.
Vacuum cleaners and cars with the highest ratings on an item have some shared
characteristics. An example is seen figure 6.5. Figure 6.5 shows the vacuum cleaner and
the car that were rated high for “serious” (Nilfisk metallic vacuum cleaner: M = 4.19, SD
= 1.05, n = 31; grey Peugot 306 sedan: M = 4.48, SD = .71, n = 33). Both are grey and
have basic, rather robust forms. In order to be able to use the product personality scale
for this goal in future research, a final step in scale development must be taken. The
concept scale developed in this chapter needs to be tested. Chapter seven will describe
this final step.
Figure 6.5: Vacuum cleaner and car rated as serious
7
Testing the Product Personality
Scale
Image that visualizes the item “cheerful”
7.1
The final step in scale development
The final step in scale development is testing the concept scale with respect to
validity and reliability. The validity of the concept product personality scale is has been
met by the selection procedure of the items. The items were reviewed and selected based
on the critical components of product personality (see chapter six). Yet, in order to serve
in future research, the scale should also be reliable. A primary test of reliability was
reported in section 6.9.3. The results of this test show that the selected items provided
reasonably reliable results. The product personality profile of most product variants is
consistent for all respondents. This also appears to be true for most of the items. All
items show alpha’s that exceed .60, and most items have alphas exceeding .80. The ri’s are
all significant. However, these reliability measures were not based on data collection with
the concept scale. A final test of a concept scale should be based on data collection with
that scale (Churchill, 1979; Devellis, 1991; Spector, 1992). The current chapter reports
two studies that used the concept product personality scale to collect data. However,
before we tested the concept scale, we conducted some extra developmental steps to
improve the unequivocalness of the items and, as such, the reliability. These steps also
helped to make the scale more attractive.
120
7.2
Differences in the interpretation of the items
The product personality scale consists of a collection of adjectives that in terms
of semantics, can be considered a set of symbols. Since symbols can be interpreted to
have various meanings, different people may also have interpreted the product
personality items differently. For example, some respondents may have interpreted
“modest” as “not very large or important”, whilst others may have interpreted it as “selfeffacing”. This difference in interpretation may make it uncertain as to what we have
measured when we find that a product is perceived as modest. Therefore, we conducted
some extra developmental steps to improve the unequivocalness of the items.
The method we used to improve the unequivocalness of the items is based on
the definition of meaning as described by in the semiotic triangle of Ogden and Richards
(1923). The semiotic triangle represents meaning by connecting symbols, referents and
thought (see figure 7.1). A symbol is something that is used to refer to something else.
Referent stands for “whatever we may be thinking of or referring to” (p. 13) when using a
symbol. The two concepts “symbols” and “referents” are linked via “thought or
reference”. The meaning of any word, in any language, can be represented by a
combination of these three concepts. For example: the meaning of the word torch is
represented by combining the symbol “torch” with the referent “
” and the
thought “a small electric lamp that uses batteries and that you can hold in our hand” (see
figure 7.1). Analogous, the meaning of modest is represented by combining the word
“modest”, with the referent “person that does not dominate the conversation” and the
thought “unassuming, unpretentious, self-effacing”. The dotted line in figure 7.1 indicates
that there is no direct relation between the symbol and the referent. The word “torch” is
used to refer to:
(in fact this picture is also a symbol, but let us assume it is a real
torch), because this symbol and this referent both relate to “a small electric lamp that uses
batteries and that you can hold in your hand”.
Thought or reference
“a small electric lamp that uses batteries and that you can hold in your hand”
Symbol
“torch”
Referent
Figure 7.1: Triangle of meaning (Ogden & Richards, 1923)
121
The fact that there is no direct connection between the symbol and referent is
the source of many difficulties in language (Ogden & Richards, 1923). One of these
difficulties is that a symbol can refer to different referents. The word “torch”, for
example, can also refer to , symbolizing the thought of “a long piece of wood that has
material at one end that is set on fire and that people carry to give light”.
The example of “torch” having two meanings demonstrates the problem of
unequivocalness. In the case of the product personality scale, the adjectives are the
symbols that may be connected to different thoughts. We assume that the
unequivocalness of the items can be improved by providing respondents with the
remaining concepts in the triangle. For this reason, we developed written descriptions
that describe the lexical meaning of the items (the thought), and developed a set of visuals
that depict examples of each item (a combination of referents). The use of visuals also
makes the scale more attractive. People like working with visual stimuli, and as such it
increases the willingness to participate in research (Van de Ven, 2003).
The development of the visuals is described in section 7.4.1, and the validation of
the visuals is described in section 7.4.2. Subsequently, the complete scale is tested (section
7.5).
7.3
Describing the lexical meaning of the items
The first step in improving the unequivocalnessof the items is developing a
written description of their lexical meaning. The lexical meaning of a word describes “the
conventional and arbitrary relation between a word and its referent” (Szalay & Deese, 1978, p. 1).
Although the relation between a word and its referent is arbitrary, or indirect as Ogden &
Richards (1923) called it, there is a conventional relation between the two. The
combination of a word and its referent exists because individuals within a group
consensually use the same word to refer to a certain object or idea. For example, we agree
to use the word “torch” is used to refer to
. The relation is arbitrary (or indirect) in
the sense that we might just as well use the letter combination “sgug” to refer to
, as
long as we all agree that “sgug” means “a small electric lamp that uses batteries and that
you can hold in your hand”.
The lexical meaning of words is described in dictionaries. A dictionary describes
the collection of symbols used in a certain language and explains their meanings. These
meanings reflect the collective agreement about how and when to use the words.
Describing the lexical meanings of the items in the product personality scale should
reduce the individual differences in interpretation because the interpretation is provided
122
with the items. The chance that “modest” is interpreted as “not very large or important”
is minimized when the item “modest” is presented with the description “self-effacing”.
Method
The lexical descriptions were developed by combining the meanings provided by
three different dictionaries: (1) Van Dale, groot woordenboek der Nederlandse taal
(1999), which is the most exhaustive dictionary of Dutch language; (2) Van Dale
basiswoordenboek van de Nederlandse taal (Huijgen & Verburg, 1996), a dictionary
containing the most frequently used words with their most common meanings; and (3)
Idioticon van de Persoonlijkheid (Doddema & De Raad, 1997), a collection of words and
synonyms used to describe human personality.
The meaning of each of the 20 items was looked up in each dictionary, and
together they formed a collection of definitions (see appendix L). Per item, the dictionary
definitions were then compared to the cluster of items from which the item originated
(see appendix K). These clusters consist of items that co-occur in personality descriptions
of products. Therefore, the items in a cluster all have similar meanings. As a consequence,
the other items in the clusters provide information about the interpretation of the
selected items. The lexical description of an item was formulated based on the list of
dictionary definitions as compared to the cluster of an item. For example, according to
the dictionaries the item “modest” had approximately four meanings. It could be
interpreted as: (1) “self-effacing” (“ingetogen”), (2) “discreet” (“discreet”), (3) “unobtrusive”
(“zich niet opdringen”), and (4) “not very large or important” (“niet groot of belangrijk”). In the
product personality scale, “modest” was selected (from the cluster “quiet”, “modest”,
“civil”, and “inconspicuous”) to represent the meaning of the whole cluster. In this case,
comparison of the cluster of personality characteristics with the dictionary definitions led
us to conclude that the lexical meaning that best described “modest” was “self-effacing”.
As a consequence, the lexical description of “modest” is formulated as “not prominent,
self-effacing”. This procedure was repeated for each item. The complete set of lexical
descriptions that resulted is shown in table 7.1.
Table 7.1
Descriptions of lexical meaning of the items (in English and Dutch)
Item
• Cheerful
Vrolijk
• Open
Open
• Relaxed
Relaxed
• Pretty
Leuk
• Easy-going
Vlot
• Cute
Schattig
• Idiosyncratic
Eigenzinnig
• Provocative
Uitdagend
• Interesting
Interessant
• Lively
Pittig
• Obtrusive
Opdringerig
• Dominant
Dominant
• Untidy
Slordig
• Childish
Kinderachtig
• Silly
Dom
• Boring
Saai
• Aloof
Afstandelijk
• Serious
Serieus
• Honest
Eerlijk
• Modest
Bescheiden
Description
Happy, joyful
Blij, opgewekt
Not secluded from his/her environment, accessible
Sluit zich niet af voor omgeving, toegankelijk
Calm, laid-back
Niet overhaastig, ontspannen
Nice, attractive, charming
Aardig, aantrekkelijk, charmant
Loose, informal
Los, ongedwongen
Very sweet, adorable
Erg lief, snoezig
Out of the ordinary, unusual
Van het gewone afwijkend, eigenaardig
Provoking
Provocerend
Arousing your curiosity
Prikkelt de nieuwsgierigheid
Strong, full of energy
Krachtig, energiek
Noticeable in a unpleasant way
Vraagt op negatieve manier aandacht
Overpowering
Overheersend
Careless, messy
Rommelig, onverzorgd
Puerile, immature
Kinderlijk, onvolwassen
Daft
Onnozel, stom
Monotonous
Eentonig
Distant, cold
Terughoudend, koel
Sensible, mature
Verstandig , volwassen
Trustworthy, sincere
Te vertrouwen, oprecht
Not prominent, self-effacing
Niet op de voorgrond, ingetogen
123
124
7.4
Visualizing the meaning of items
This section describes the development and validation of the visuals. The visuals
are to be used together with the verbal items, in order to raise the unequivocalness of the
items. The use of visual items is not only based on depicting referents as defined by the
semiotic triangle, it also fits in current developments in consumer research. Visual items
are being applied more and more often because of their power of expression and
unequivocalness (Van de Ven, 2003). In addition, visual information is often understood
non-verbally (Doumont, 2002). Since our stimuli are mostly pictures of product variants,
a scale with partly visual items should make the application of the items to the products
easier. Comparing the pictures of products with the verbal items requires a mental
transition of a non-verbal (intuitive) impression to a (rational) verbal item. Comparing the
stimulus picture with a visual item is in that sense more direct and, therefore, possibly less
open to variation between respondents. First, the development of the visuals is described
in section 7.4.1. Section 7.4.2 subsequently describes the validation of the visuals
7.4.1
Development of the visuals
Pictures or visuals that are meant to represent something can be interpreted at
three different levels: (1) iconic, (2) indexical and (3) symbolic (Doumont, 2002; Mick,
1986). At the iconic level, a picture is interpreted literally; it imitates or resembles what it
represents. The picture of a rose then means a rose. At the indexical level, a picture is
interpreted as a metaphor; it has some factual correspondence to what it represents. The
picture of a rose at this level can be used to represent something that smells good. At the
symbolic level, the picture and what it represents only have a conventional relation (like in
the semantic triangle, figure 7.1). The picture of a rose at this level can be used to mean
love or passion. These levels are not mutually exclusive. A picture can function in all
three capacities.
Combining several pictures into one visual makes it possible to represent not just
one object or idea, but something more multifaceted (like a mood or context). Designers,
for example, use visual collages to depict their target group or a user context. They then
deduce product appearance characteristics from it (Muller, 1997). The visuals that are to
depict the meaning of the items also consist of a combination of pictures. In terms of the
semantic triangle (figure 7.1), the visuals consist of a collection of referents, which in
combination represent the common “thought”. Using only one picture to visualize each
item in the product personality scale is insufficient because a single picture illustrates but
only one instance of a concept (Doumont, 2002).
125
Pictures were collected from the categories flora, fauna, humans and artifacts.
The categories flora and fauna were chosen because they could provide instances of
symbolical pictures and we wanted to provide respondents with an example of figurative
use of the items. Humans were included because they represent the original use of the
items. Artifacts were included because they are referents that resemble durable products.
Since we did not want to distort the answers of respondents by giving them specific
examples of relevant product variants, durable consumer products were excluded.
The development of the visuals included two stages: (1) gathering pictures that
represented the items and (2) selecting a combination that correctly represented the
meaning of the items as described in section 7.3.
Participants
Eleven participants gathered pictures and discussed them. Seven of them were
male and four were female. Their ages ranged from 27 to 47, with an average of 34.
Participants are members of the design staff at Industrial Design Engineering, Delft
University of Technology and/or working in the field of design as a practitioner. We
preferred designers as participants because designers are very familiar with visualizing
meaning and have experience in gathering a lot of visual material.
Design
The set of the 20 verbal items was split up randomly into four sets of five items.
The participants were also divided into four groups. The participants of one group all
received one set of items. They collected pictures from the four different categories for
every item in the set and discussed the selection with each other during group sessions.
Procedure
Participants were sent an email with the set of five items for which they were to
gather pictures. They had two weeks to gather pictures for all five items. The pictures
needed to originate from all four selected categories (flora, fauna, humans and artifacts),
and had to communicate the meaning of the items as described by the lexical meaning
(see table 7.1). Both the content and the way the picture was taken had to be taken into
account.
After they had collected the pictures, the participants of each group participated
in a group session. During the group session, the participants compared and discussed
the pictures that each of them had collected. The goal of the group session was to select
the pictures that, according to the participants, communicated the meaning of an item
best. If participants could not come to an agreement about which picture to select, then
they could leave the category/item combination open. They were then asked to indicate
126
what an alternative picture would look like. The author acted as session leader in all
sessions. The group sessions lasted one-and-a-half to two hours. Participants received a
book token afterwards.
Results
The result of the group sessions should have been a filled matrix of twenty
(items) by four (categories) of pictures. However, some of the cells in the matrix
remained empty. In some cases, participants decided that none of the pictures that they
had collected communicated the item clearly enough. For example, in the case of
“honest”, the best picture from the category “humans” depicted a male adult and a child
playing. Though all of the participants agreed that the combination of a male adult with a
child communicated honesty, all of the pictures at hand were found to be too cheerful. In
some other cases, none of the participants found a picture that sufficed. In the case of
“untidy” and “fauna”, the participants could not think of any untidy animal. According to
the participants “untidy” meant “outside the norm”. Since animals always set their own
norm, they concluded that there is no untidy fauna. In both of these cases, the
respondents indicated what the missing picture should look like. This information could
then be used as guide in further collection.
Not only were some cells left vacant, selection also included pictures that could
not be used in the product personality scale. For example, a wooden cross placed on
ground zero in New York was selected as an artifact to depict “serious”. This is indeed a
serious symbol, but it does not communicate serious as “mature and sensible”. Moreover,
this symbol might needlessly hurt peoples’ feelings.
As a result of the vacancies, and because some cells could better be represented
with other pictures, a second collection and selection procedure was conducted. This
second procedure was conducted by four participants, two of them were male and two
female. Their ages ranged from 42 to 54, with an average of 48. None of them were
designers. During the previous group sessions it became clear that some obvious pictures
were excluded because of the fact that participants were designers. The participants
considered these pictures too cliché. Therefore, the second set of participants did not
include designers, but rather well-educated9 professionals with a larger than average
vocabulary and understanding of Dutch language. One had a background in Arts, one in
psychology, one in pedagogy, and one in management. The four participants were divided
into two groups consisting of one male and one female. Both groups were provided with
the partly filled out picture matrix, and the collected pictures that could fill in the blanks.
The comments of the participants in the group sessions served as guides. In a time period
9
All participants had at least a bachelor degree.
127
of three weeks, these participants collected pictures and made a final selection together
with the author. Most of the shortcomings of the previous selection were countered in
this second collection and selection stage (e.g. the cross on ground zero has been replaced
by a rope), but it remained difficult to fill all cells with appropriate pictures (e.g. “fauna”
and “untidy”). It was decided to limit the amount of pictures per item to three. The
visuals therefore consist of combinations of three pictures out of the four categories.
These visuals are shown in figure 7.2. The face-validity of these visuals was high enough
to proceed with an empirical validation.
7.4.2
Validation of the visuals
The validity of the visuals is crucial if the visuals are to increase the
unequivocalness of the items. Invalid visuals do not depict the meaning of the single
worded items, but rather complicate the items with even more possible interpretations.
Invalid visuals will thus decrease instead of increase the reliability of the product
personality scale. Van de Ven (2003) also notes the quality of visual items: “Visuals can
raise the quality of research. Of course, the quality of the visuals is essential. We are not talking about a
set of pictures, but about a carefully collected and validated set of visuals” (p. 49). The study reported
in this section investigates the validity of the visuals that resulted after the selection
procedure described in section 7.4.1.
The method we used to validate the visuals is based on recognition studies of
facial expressions. Most of these studies follow the same procedure. Respondents are
shown a set of facial expressions and asked, for example, to indicate “which of these faces best
expresses anger”. When the proportion of respondents who selected the correct facial
expression is significantly above chance level, then this facial expression is considered to
be valid (Desmet, 2002). Desmet successfully translated this method to validate visual
stimuli. The goal of his studies was to find out whether a set of animations validly
represented a set of emotions. Since the goal of the current research is similar to that of
Desmet’s study, i.e. we wanted to find out whether a set of visuals validly represents a set
of personality characteristics, we used a similar validation method. In order to find out
whether the visuals are valid representations of the items, we presented respondents with
three items (e.g. “cute”, “serious” and “pretty”) and asked them to indicate which item
best described a given visual (e.g. visual representing “serious”). When the proportion of
respondents that selected the correct item was significantly above chance level, the visual
was considered to be valid.
128
Cheerful
Obtrusive
Open
Dominant
Relaxed
Untidy
Pretty
Childish
Easy-going
Silly
Cute
Boring
Idiosyncratic
Aloof
Provocative
Serious
Interesting
Modest
Lively
Honest
Figure 7.2: The visuals per item
129
Respondents
Respondents were approached to participate in a study about personality
characteristics at the library of Delft University of Technology. Ninety-three respondents
agreed to participate. Fifty-seven of them were male (61%) and 36 were female (39%).
Their ages ranged from 18 to 65, with an average of 24. Participation was voluntary.
The respondents were mostly students from Delft University of Technology
(79%). Twelve percent of the respondents were students from other universities. In total,
they came from 22 different schools. The remaining 9% were otherwise engaged at the
library (e.g. security personnel and library personnel).
Stimuli
The visuals that were developed for clarifying the meaning of the items (see
figure 7.2) were presented to the respondents on a computer screen, together with three
items. Figure 7.3 shows an example. The items were presented together with their lexical
descriptions as formulated in section 7.3 (see also table 7.1).
Figure 7.3: Example of stimuli presentation
Design
The presentation of the visuals was randomized. Every respondent was
presented the visuals in a different order. Each visual was presented with the correct item
and two other randomly selected items. To make sure that the correct item was not
always presented at the same position, we also randomized the presentation order of the
items.
130
Procedure
Respondents who agreed to participate were seated behind a computer in a
secluded area at the library. The presentation of the visuals was preceded by an example.
The 20 visuals were presented to the respondents, one at a time. Respondents were asked
to indicate “what personality characteristic best described the meaning of the pictures”. It was stressed
that the three pictures in the visual should be interpreted as a whole. The selected
personality characteristic had to describe the meaning of the three pictures together. The
test concluded with three general questions (age, gender and study or occupation) and
lasted 5-10 minutes.
Results
In order to test the validity of the visuals, we calculated the amount of times
respondents chose the right item and subsequently computed the proportion of correct
answers per visual. If respondents recognized the meaning of the visuals, then the
proportion of correct answers should exceed the proportion of correct answers that can
be expected on the basis of chance. The criterion of validity was therefore set at the
proportion of correct answers being significantly higher than chance level (.33). The 95%
confidence interval of .33 was computed10 and for each visual the proportion of correct
answers was compared to these confidence limits (.25 and .43). Figure 7.4 shows the
proportion of correct choices for each visual. The vertical line and the shaded area
represent the 95% confidence interval of .33. The horizontal axis represents the visuals,
named after the item they were intended to visualize.
The results show that the proportion of correct choices exceeds chance level for
every visual. This means that all visuals can be considered valid. For most of the items,
the proportion of correct choices is far beyond chance level. Scores ranging from .72 to
.98 indicate that the majority of respondents selected the correct item. Though the
proportion of correct choices was lowest for the visuals depicting “easy-going” and
“obtrusive” (.61 and .59 respectively), two thirds of the respondents still selected the
correct item in these cases.
10
The confidence interval for proportions is calculated with the following formula:
N/N+z2 [P+(z2/2N)+-√(PQ/N)+( z2/2N)] where P is the proportion, Q=1-P and z is the
standard score in a normal distribution cutting of, in our case, the upper 2.5% of cases (Hays,
1988, p. 240).
131
Figure 7.4: Proportion of correct answers per visual over the total sample.
In order to ensure the validity of the visuals, it is important that they discriminate
between different items with related meaning. As we have seen in section 6.9.2 the items
constitute six clusters of meaning. The results of the analysis over the total sample may
mask confusion between related items because the visuals were presented with
alternatives randomly chosen from all 20 items, including items with completely different
meanings (e.g. “cheerful” with “serious”). The proportion of correct answers may
decrease when a visual is presented with alternatives with a closely related meaning group
(e.g. “cheerful” with “pretty”). Therefore, we selected the cases where at least one of the
two alternative items stemmed from the same cluster as the target item. We then
computed the proportion of correct choices to see whether it exceeded chance level.
Again, the 95% confidence interval of .33 was computed. Figure 7.5 shows the
proportion of correct choices for every visual that was presented with alternatives from
the same group. The vertical line and the shaded areas represent the 95% confidence
interval of .33. The amount of times a visual was presented with at least one alternative
from the same group differed per visual. Since confidence limits broaden as an effect of a
low n, the confidence limits differ per visual.
132
Figure 7.5: Proportion of correct answers per visual if visuals are presented with
alternative items from the same group.
The results show that the proportion of correct answers for the visuals “easygoing” and “obtrusive” do not exceed chance level. The proportion of correct answers
for “easy-going” is .43 (n = 42) and the 95% confidence limits range from .21 to .48. The
proportion of correct answers for “obtrusive” is .23 (n = 13) and the 95% confidence
limits ranged from .14 to .61. Since developing new visuals for these items takes a lot of
time, both “easy-going” and “obtrusive” have been excluded from the current analyses.
The scale now contains 18 items that can be expanded with a lexical description
and/or a visual. Both the lexical description and the visuals are considered valid
representations of the items. The validity of the visuals has been confirmed in this study.
The lexical descriptions are based on dictionaries and thus have not been validated
empirically. As such, they are assumed to raise the unequivocalness of the items. The
complete product personality scale, including the lexical descriptions and the visuals, shall
now be tested with regard to its reliability.
7.5
Testing the product personality scale
As a final step in scale development, data needs to be collected with the concept
scale (see section 6.1). This section reports two of such studies. The goal of both studies
was to find out whether the current product personality scale is reliable. The first study
was conducted using stimuli from a durable consumer product class (clock radios). As a
133
result of the small response set (n = 25), a second study was conducted to confirm these
findings. An additional goal of the second study was to broaden the scope of use of the
product personality scale. In this study we used logos as stimuli, instead of a durable
consumer product class.
Outside of the scope of durable consumer products, logos are an example where
the product personality scale could also be applied. Logos are designs or symbols that
is the
companies use as their special sign (Oxford Dictionary, 2000). For example,
logo of Nike. Although logos are not durable products in the pure sense of the term,
logos do meet the two criteria of durability and tangibility that determine the classification
as a durable product (see the definition of durable consumer products in section 1.4). As
long as the personality of the logo is assessed, and not the personality of the company
represented by the logo, then the product personality scale is still applicable.
The main goal of both studies remained to investigate whether the product
personality scale provides reliable product personality profiles for all respondents. That is
to say, different respondents should perceive the product personality of a particular
product variant consistently. This way, we can be sure that, once we have established the
product personality profile of a product variant, this profile is generalizable.
7.5.1
Method study 1
Respondents
Respondents were approached at the library of Delft University of Technology
and asked to participate in a short study about rating products on a set of characteristics.
Twenty-five respondents agreed to participate. They were all students, coming from Delft
University of Technology, Utrecht University or Leiden University. Twelve of them were
male (48%), and 13 were female (52%). Their ages ranged from 19 to 25, with an average
of 22. Since product designers have been shown to perceive products differently (Hsu,
Chuang & Chuang, 2000), students of Industrial Design Engineering were excluded from
participation.
Stimuli
The stimuli for this study were selected from the product class clock radios. In
order to ensure variation in appearance between the stimuli, color pictures from the
current (2003) range of clock radios were collected and arranged into groups of similar
appearance. From each of these groups, one product variant was chosen. In order to
minimize possible brand effects, the brand names and logos were removed. The color
pictures of the clock radios selected to serve as stimuli are presented in appendix M.
134
The items in the scale consisted of the word, the lexical description and the
visual. All 18 items were presented to the respondent together on an A3 format
questionnaire.
Design
Every respondent rated all six stimuli, thus filling out the product personality
scale six times. The 18 items were randomized into six random sequences of presentation,
a different sequence per stimulus. The presentation of the stimuli was also randomized.
Each stimulus was printed on a card of 10 by 15 centimeters and these six cards were
arranged in six different presentation orders.
Procedure
Respondents were seated in a secluded area at the library. They were given a set
of six product personality questionnaires (one questionnaire for each stimulus) and the six
cards depicting the stimuli. Each stimulus was rated on the 18 items using a 7-point scale
(1 = “not at all descriptive” and 7 = “totally descriptive”). It was stressed that the three
pictures and the written descriptions accompanying the items should be interpreted as a
whole.
When finished rating the first stimulus, respondents returned the stimulus picture
together with the first questionnaire. Then they continued on with the second stimulus.
This procedure continued until all six stimuli had been rated. This took approximately 10
minutes. Participation was voluntary.
7.5.2
Results study 1
Reliability of the personality profiles
This study was set up to test the reliability of the product personality scale. We
were especially interested in the reliability of the personality profiles. The personality of
each stimulus should be perceived consistently by different respondents. Therefore, we
calculated Cronbach’s alpha and the intraclass correlation coefficient for every product
variant. The results are depicted in table 7.2. The numbers of the clock radios in table 7.2
correspond to the numbers in appendix M. The results show high reliability values for all
six stimuli. The alphas all exceed .80, and the ri’s are all significant (p < .001). These
results indicate that the product personality profiles of all stimuli are perceived
consistently by all respondents.
135
Table 7.2
Reliability measures of the product personality scale per product variant (n = 25)
Clock radio 1
Clock radio 2
Clock radio 3
Clock radio 4
Clock radio 5
Clock radio 6
α
.92
.80
.95
.86
.98
.96
ri
.26**
.12**
.37**
.17**
.56**
.43**
** The correlation coefficients are significantly higher than zero, p < .001
Reliability of the individual items
To be reliable, the scale also needs to provide consistent results for every item.
Respondents should agree about product variant X being cheerful, and product variant Y
being less cheerful. We calculated Cronbach’s alpha and the intraclass correlation
coefficients for each item. The results are depicted in table 7.3.
The results show high reliability values for most items. All alphas but one exceed
.60, which is the lower limit for Cronbach’s alpha (Hair et al., 1998), and most alphas
exceed .80. The same goes for the ri’s. All ri’s but one are significant. Most of them at p <
.001. “Honest” shows relatively low values (α = .67 and ri = .07), but the alpha is still
higher than .60 and the ri is significant (p < .05). The item that stands out is “silly”. “Silly”
provides remarkably low values (α = -.06 and ri = -.002, n.s.). Values this low generally
indicate that the item is not relevant. The fact that “silly” is not relevant here, may be
caused by the nature of the stimuli. It seems reasonable that “silly” is not relevant to
clock radios.
Personality profiles
The ultimate goal of the product personality scale is to be able to calculate
product personality profiles per product variant. In order to ensure that the product
personality scale provides different personality profiles for different stimuli, the mean
scores of two stimuli on the product personality items are calculated and represented in
table 7.4. The resulting personality profiles are depicted in figure 7.7.
The results show that the stimuli indeed have significantly different personalities.
All but three items show significant differences. For example, the clock radio represented
by the dotted line in figure 7.7 (M1 and SD1 in table 7.4) is more cheerful, open, relaxed
and pretty than clock radio represented with the solid line (M2 and SD2 in table 7.4). The
first clock radio, on the other hand, is more idiosyncratic, interesting, lively and dominant
Yet, it is also perceived as more boring and aloof.
136
Table 7.3
Reliability of the items (n = 25)
α
.96
.91
.85
.93
.91
.90
-.06
.85
.77
.94
.94
.86
.91
.91
.67
.75
.87
.86
Selected items
Cheerful
Open
Relaxed
Pretty
Cute
Dominant
Silly
Childish
Untidy
Idiosyncratic
Interesting
Lively
Provocative
Modest
Honest
Serious
Aloof
Boring
ri
.48**
.27**
.19**
.36**
.28**
.26**
-.002
.19**
.12**
.37**
.37**
.19**
.28**
.28**
.06*
.10**
.21**
.20**
* Correlation coefficient is significantly higher than zero, p < .05
** Correlation coefficient is significantly higher than zero, p < .001.
Table 7.4
Mean scores on the product personality items of two clock radios
Items
Cheerful
Open
Relaxed
Pretty
Cute
Dominant
Silly
Childish
Untidy
Idiosyncratic
Interesting
Lively
Provocative
Modest
Honest
Serious
Aloof
Boring
* p < .05
** p < .001.
M1
SD1
M2
SD2
t(48)
3.80
4.60
4.36
3.96
3.12
2.76
4.60
3.24
2.28
2.48
2.80
3.08
2.48
5.00
4.88
4.28
3.60
3.92
1.12
1.19
1.15
1.34
1.76
.88
1.35
1.39
.84
1.16
1.16
1.32
1.09
.87
1.09
1.37
1.56
1.38
2.56
3.20
2.96
2.88
1.88
5.00
2.72
1.96
2.12
4.52
4.28
4.36
4.04
3.08
4.12
5.32
4.96
4.24
.92
1.19
1.17
1.05
1.67
1.32
1.72
.98
1.27
1.36
1.46
1.32
1.72
1.16
1.16
1.28
1.46
1.39
4.05*
4.16**
4.26**
3.17*
2.93*
-7.05**
.18
3.76**
.53
-5.71**
-3.98**
-3.43
-3.84**
5.40**
2.38*
-2.77*
-3.19*
-.82
low
Cheerful
Open
Relaxed
Pretty
Cute
Idiosyncratic
Provocative
Interesting
Lively
Dominant
Untidy
Childish
Silly
Boring
Aloof
Serious
Honest
Modest
137
high
●
W
●
W
●
W
●
W
●
W
●
●
W
W
●
W
●
W
●
W
W
●
●
W
W
●
●W
●
W
●
W
W
W
●
●
Figure 7.7: Product personality profiles of two clock radios
Dimensions of product personality
A final analysis concerns a confirmatory factor analysis. We ran a factor analysis,
forced into a three-factor solution, in order to test the existence of three dimensions.
Earlier results about product personality suggested the existence of three dimensions of
product personality that resemble the agreeableness, extroversion and conscientiousness
dimensions of human personality (see section 6.9.3). The results show a three-factor
solution that explains 60% of the variance (R2 = .59). All of the items have high loadings
on one of the factors (in absolute values, ranging from .54 to .85), and relatively low
loadings on the other two factors. Communalities of all items are high (> .48). The
factors can indeed be interpreted as agreeableness, extroversion, and conscientiousness.
Table 7.5 shows which items load on the three dimensions.
Table 7.5
Agreeableness
Explained Eigenvariance
value
23%
4.35
Extroversion
20%
3.51
Conscientiousness
16%
2.72
Dimension
Items with high factor loadings
pretty, cheerful, open, cute, relaxed,
aloof (-), boring (-)
provocative, idiosyncratic, interesting,
lively, dominant, modest (-)
serious, honest, silly (-), untidy (-),
childish (-)
138
7.5.3
Method study 2
Respondents
Sixty-six respondents participated in this study. A considerable proportion of
them (67%) were males (44 respondents). The remaining 33% were females (22
respondents). The ages of the respondents ranged from 19 to 31, with an average of 25.
Participation was voluntary.
Stimuli
The stimuli for this study were four logos designed to express the identity of a
boutique named “Het Mantelpak”. This boutique sells women’s clothes and targets
working women of all ages. Color pictures of the logos are depicted in figure 7.8.
Figure 7.8: Presentation of the stimuli
Design
Each respondent rated all four stimuli. The four stimuli were presented together.
The items appeared on the screen one at a time (see figure 7.8 for an example).
Respondents rated all four stimuli. When they were finished with the first item, then they
could proceed to the second item. This process continued until all oft the stimuli had
been rated on all items. The sequence of the 18 items was randomized. Each respondent
was presented this same sequence.
Procedure
The study was conducted via the internet. Respondents were invited to
participate via email, and could enter the study via a direct link in the message. This link
opened an introduction page. This page explained the design of the study and stressed
139
that the three pictures and the written descriptions accompanying the items should be
interpreted as a whole. After rating the logos on the first item, respondents could proceed
to the next item. The stimuli remained on the screen. Every item was rated using a 7point scale (1= “not at all descriptive” and 7= “totally descriptive”). Rating the logos on
all items took respondents approximately 10 minutes.
7.5.4
Results study 2
Reliability of personality profiles
This second study was conducted to confirm the reliability of the product
personality profiles that resulted from the product personality scale. We calculated
Cronbach’s alpha and the intraclass correlation coefficient per logo. The results are
depicted in table 7.6. The numbers of the logos in table 7.6 correspond to the numbers in
figure 7.8. The results again show high reliability values for all stimuli. The alphas all
exceed .90, and the ri’s are all significant (p < .001). This indicates that the personality
profiles of the logos are perceived consistently for all respondents. The reliability scores
of the item “silly” do show high values this time (α = .90 and ri = .12, p < .001).
Table 7.6
Reliability measures of the product personality scale per logo (n = 66)
Logo 1
Logo 2
Logo 3
Logo 4
α
.90
.95
.98
.96
ri
.12**
.24**
.49**
.27**
** The correlation coefficients are significantly higher than zero, p .001
Personality profiles
To find out whether different logos, like different product variants, also have
different product personalities, the mean scores of two logos on the product personality
items are calculated and represented in table 7.7. The resulting personality profiles are
depicted in figure 7.9. The results show that the two logos do have significantly different
personalities. All but three items show significant differences. For example, the “blue
logo” (logo 4) is more serious, honest and aloof than the “purple logo” (logo 2). Logo 2,
on the other hand, is more cheerful, open, and cute. In terms of the dimensions named in
table 7.5, logo 2 can be characterized as agreeable but not conscientious. Logo 4 shows
140
the opposite pattern. It is perceived as conscientious but not as agreeable. With respect to
extroversion, the two logos are not very different. Both are moderately extroverted.
Table 7.7
Mean scores on the product personality items of two logos
Items
Cheerful
Open
Relaxed
Pretty
Cute
Dominant
Silly
Childish
Untidy
Idiosyncratic
Interesting
Lively
Provocative
Modest
Honest
Serious
Aloof
Boring
Mlogo 2
SDlogo 2
Mlogo 4
SDlogo 4
t(130)
5.39
5.05
4.12
4.67
4.29
3.58
3.58
4.80
3.52
5.26
4.74
4.18
4.44
3.34
4.26
3.18
2.47
2.33
1.36
1.47
1.53
1.66
1.72
1.55
1.92
1.55
1.71
1.42
1.50
1.55
1.86
1.50
1.33
1.52
1.04
1.39
3.61
3.76
4.26
3.95
2.18
4.36
2.58
2.56
2.74
4.35
4.23
4.71
3.52
4.64
5.08
5.05
5.08
4.24
1.57
1.53
1.51
1.49
1.19
1.54
1.48
1.35
1.76
1.64
1.63
1.66
1.60
1.44
1.23
1.43
1.52
1.90
7.01**
4.93**
-.51
2.59*
8.19**
-2.93*
3.35**
8.86**
2.56*
3.37**
1.89
-1.90
3.06*
-5.04**
-3.67**
-7.26**
-11.48**
-6.60**
* p < .05
** p < .001.
low
Cheerful
Open
Relaxed
Pretty
Cute
Idiosyncratic
Provocative
Interesting
Lively
Dominant
Untidy
Childish
Silly
Boring
Aloof
Serious
Honest
Modest
high
W
W
W
W
WW
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
Figure 7.9: Personality profiles of two logos
7.5.5
141
Discussion
The goal of the studies reported in this section was to test the reliability of the
product personality scale based on data collected with the concept scale. The product
personality scale needed to provide reliable product personality profiles for all
respondents. Consistency in the way people perceive the personality of one product
ensures that, once the product personality profile of a product variant is established, this
profile is generalizable for all people. The results show that the personality profiles of all
stimuli, in both studies, are reliable. The alphas in the second study even exceed .90,
meaning that 80% of the variation is explained.
When looking at the reliability of the individual items, it appears that the
reliability is not always high. Most items show high reliability scores. Yet, sometimes a
single item appears to be unreliable (e.g. “silly” in study 1). Since the reliability of the
items is less crucial than the reliability of the profiles, the reliability of the product
personality scale is still considered to be sufficient. Nevertheless, in future studies one
may want to replace the items that show low reliability scores, i.e. “silly” and “honest”,
with other personality characteristics from the same cluster. New visuals and lexical
descriptions will then have to be developed and validated. This is also an option for the
two items that were removed from the scale because their visuals were not valid, i.e.
“obtrusive” and “easy-going”. Until future research is conducted, the current product
personality scale will suffice and can be applied to assess the product personality of
product variants from different product classes.
We used logos as stimuli in the final study in order to extend the applicability of
the product personality scale beyond durable consumer products. Logos resemble durable
products in important ways. This led us to assume that the product personality scale
could also be applied to logos. We found out that logos also have personalities that are
consistently perceived by respondents. There is another important link between product
variants and logos. Like the appearance of product variants, logos are also communicative
tools of a company. A company uses the appearance of their products, as well as their
logo, to express the image of their brand. The fact that product variants, and maybe even
logos, have personality implies that both can be used to sustain the brand’s personality.
This expansion of the use of the product personality scale does not imply, however, that
the scale can be used for any other kind of object. The product personality scale is not
suitable for example, for establishing the personality of fast-moving products. Fastmoving products may also have a personality. Yet, characteristics that may be relevant to
describe aspects unique to fast-moving products, such as taste and smell (e.g. bitterness,
sharpness), are not included in the product personality scale. In the case of logos, the
possibility that relevant characteristics have not been included is considered minimal.
142
Logos have no extra aspects in addition to their appearance and, in our case, the logos
represented a company that was unknown to the respondents. Measuring the product
personality of the logo of Nike would be more difficult because it is very well known. As
a result you would be assessing brand personality and not the personality of the logo.
Limitations
A limitation of the first study was the size of the sample set (n = 25). However,
each respondent rated every stimulus. The scale is thus filled out not 25, but 150 times.
As a consequence, we have 150 data points per variable. Moreover, the response set of
the second study was larger (n = 66) than in the first study (n = 25), allowing us to
conclude with more certainty that the product personality scale is indeed a reliable
measure.
A second limitation of this first study is the fact that we used only one product
class in the first study. The product personality scale is to be used with respect to all kinds
of durable consumer products. However, given the fact that the items are valid for all
product classes, finding the scale reliable for one type of durable products should be
sufficient. Yet, it is not a strong basis. We did confirm the reliability of the scale in the
second study. In this study, instead of using another class of durable consumer products,
logos were used as stimuli.
A limitation of the second study was the design. We randomized the order of the
stimuli, but we offered the stimuli in the same sequence to all respondents. The same is
true for the items. We only randomized the order of the items in the product personality
scale once. All of the respondents filled out the same questionnaire. The fact that all of
the respondents filled out an identical questionnaire may have caused the alphas in this
study to be remarkably high. However, even if some proportion of the alphas in the
second study is caused by the similarity of the questionnaires, the alphas and correlation
coefficients are still high.
7.6
Conclusions
The research described in this chapter was conducted as the final step in
developing a product personality scale. We conducted some extra developmental steps to
improve the unequivocalness of the items that resulted from chapter six and tested the
reliability of the final scale. The final scale contains 18 items, existing of a combination of
the original adjective, a lexical description, and a visual. The results of two studies show
that this final scale is a reliable measure of product personality. However, we have not
been able to prove that the scale with the visuals and lexical descriptions is more reliable
143
than simpy using the original selection of 20 personality characteristics as items. A future
study would have to employ a two-by-two factorial design (visual absent or present and
lexical description absent or present) to test this assumption.
Another important conclusion is that the results of the studies in this chapter
strengthen the suggestion of three factors of product personality that resemble the
dimensions agreeableness, extroversion, and conscientiousness in human personality.
The results also confirm the role of appearance in the perception of product
personality. The clock radio and logo perceived as most serious (clock radio M = 5.32,
SD = 1.28, n = 25 and logo M = 5.80, SD = 1.27, n = 66) are comparable in terms of
appearance to the car and vacuum cleaner that resulted from earlier studies reported in
chapter 6 (see figure 7.10). Like the car and the vacuum cleaner, the serious clock radio
and logo are also grey (with black) and have robust forms.
Figure 7.10: Vacuum cleaner, car, clock radio and logo perceived as serious
8
(Siemens advertisement, 2003)
8.1
Product personality
The previous chapters have reported research on the different aspects of product
personality. This chapter will discuss the findings and implications of this research with
respect to the three research issues. This first section recapitulates the goal of this thesis.
Section 8.2 describes the position of product personality in consumer research literature.
The perception of product personality is discussed in section 8.3. The influence of
product personality on consumer preference is discussed in section 8.4, and the
implications of systematic assessment of product personality are discussed in section 8.5.
How these results increase the actionability of product personality is discussed in section
8.6. The chapter concludes with ideas for future research (section 8.7).
Each chapter heading in this thesis is accompanied by a picture. Several of these
pictures are advertisements, which praise their products by using human personality
characteristics, such as friendly toilet paper (chapter one), sporty cars (chapter three), and
sexy mobile phones (this chapter). This way of advertising is used, because now that
adequate functionality is the norm, symbolical aspects of products are recognized as more
important (Weightman & McDonagh, 2003). The market offers a variety of product
variants in different shapes and colors, which all fulfill the functional needs of the
consumer (see figure 8.1 for an example). How do consumers choose among all these
variants? If asked, they will probably say that they consider price, quality and
functionality. Nevertheless, beyond price, quality and functionality there are symbolic
reasons that guide consumer behavior. These symbolic motivations are becoming
increasingly important.
145
Figure 8.1: Advertisement for mobile phones (BelCompany, January 2004)
Previous consumer research has suggested that consumers prefer products that
match their self-concept or personality (e.g. Aaker, 1999; Malhotra, 1988; Levy, 1959;
Sirgy, 1982). Consumer research has studied this symbolic use of products. Yet, the
knowledge gained from these fields of study is too abstract to be applied directly to
product design. In this thesis, we have presented the concept of product personality.
Product personality refers to the human personality characteristics that people use to
describe product variants. The contribution of product personality lies in the fact that
product personality describes the impression of a single product variant. Product design
works at the level of product variants.
In order to make product personality usable in product design, we set out to
increase the actionability of product personality. Actionability of product personality
means that it is clear what specific actions designers must take to create a product with a
certain personality (Shocker & Srinivasan, 1974). Information about the perception of
product personality, the influence of product personality on consumer preference, and
the systematic assessment of product personality provide designers with a basis for better
adapting the products that they design to the wishes of the consumer.
8.2
The symbolic meaning of product variants
People use the symbolic meaning of products to convey information about
themselves to themselves and to others (Holman, 1981; Solomon, 1983). This symbolic
meaning of products has been studied from a social perspective (focusing on products as
symbols of status and group membership), and from a more individualistic perspective.
Product personality best fits the tradition of the individualistic perspective.
The individualistic oriented study of the symbolic meaning of products has
largely focused on products as symbols of the self-concept. People have a self-concept
and are motivated to confirm this self-concept by the motive of self-consistency
146
(Rosenberg, 1979). Self-consistency influences consumer preference, because consumers
prefer products that are consistent with their self-concept (e.g. Belk, 1988; Malhotra,
1988; Sirgy, 1982). “All commercial objects have a symbolic character, and making a purchase involves
an assessment – implicit or explicit – of this symbolism….” (Levy, 1959, p.119). With respect to
product personality, this means that during the evaluation process, a consumer compares
his/her self-concept to the personality of the available product variants and will prefer a
product variant with a personality that matches with his/her self-concept.
What does product personality add to the existing concepts in symbolic
consumption? The fact that people prefer products with an image that is consistent with
their self-concept is not new. Yet, product personality is different from existing concepts.
First, it refers to the product itself. Most of the existing concepts assume that the
impression of a product, as described by personality characteristics, refers to the users of
that product. The product-user image reflects the image of the stereotype user (Sirgy,
1997), and brand personality is said to describe the “user component of brand image” (Biel,
1993, p. 71). More importantly, product personality differs from brand personality
because it describes a single product variant and not a brand. This distinction forms the
base of the contribution of product personality to product design and to consumer
research literature.
By means of product personality, consumers are able to communicate their
individuality and to meet their need for uniqueness. User-image and brand personality tap
more into the need to belong to a group. For example, by purchasing a mobile phone
people comply with the current norm of being reachable ubiquitously. By buying a Nokia,
consumers meet the norm of their reference group. By choosing the specific Nokia
variant with a faceplate that expresses their personality, the consumer finally fulfills
his/her need for uniqueness.
Figure 8.2: Different Nokia faceplates
8.3
147
Perception of product personality
The first issue we set out to investigate with was the perception of product
personality. We focused our research on the influence of product characteristics because
this aspect can be manipulated by designers. In chapter two, we argued that differences
between the product variants would have a stronger effect on the perception of product
personality than the differences between perceivers. In other words, the perception of
product personality was assumed to be a function of the product’s characteristics, and not
determined in the eye of the perceiver. Whether product variant X is cheerful is
determined by aspects of products X. This does not change when person B perceives it
instead of person A.
The results of our studies provide post hoc evidence that substantiate this
argument. If perception is a function of the product’s characteristics, consensus among
perceivers should be high (Kenny, Albright, Malloy & Kashy, 1994). The results of the
reliability tests in chapter 7 show high consensus between consumers. People agree about
the personality of a product variant. Both studies in section 7.5 (n = 25, and n = 66) yield
high alphas (generally > .80), as well as significant intraclass correlation coefficients
(mostly p <.001). These results imply that product personality is indeed largely
determined by characteristics of the products.
The role of product appearance
Since the perception of personality in humans is strongly influenced by
appearance characteristics, we investigated the influence of appearance characteristics on
the perception of product personality. The results of our first studies (section 3.2 and 3.3)
showed that watches that were described with similar personality characteristics had
comparable appearance characteristics. The masculine watches, for example, were all
robust and big. It also appeared that small functional differences that were apparent in
the form of the product influenced product perception. The masculine watches had more
additional features than the other watches. Moreover, irons designed to express a
particular personality characteristic (“happy”, “cute” or “tough”) also had similar shapes.
Cute irons were round and stocky, happy irons were round and open, and tough irons
were big and robust. It even appeared that the irons and watches described with the same
personality characteristics shared similar appearance characteristics. The tough irons, like
the tough watches, were big and robust.
The assumption that product variants from different product classes with the
same personality have similar features was strengthened after comparing product variants
with a similar personality from later studies. The serious product variants shown in figure
7.10 stem from studies reported in sections 6.9 and 7.5. They are all grayish and have
148
basic, rather robust forms. The personality of a product variant thus seems to be
determined to a significant degree by the appearance of that product. Future research
could investigate this further by systematically manipulating product appearance
characteristics.
Another result regarding the perception of product personality is that product
personality seems to describe three dimensions. These three dimensions are comparable
to three of the Big Five dimensions of human personality: agreeableness, extroversion
and conscientiousness. The same three dimensions of human personality were also
apparent in the dimensions of brand personality (sincerity, excitement and competence)
(Aaker, 1997).
A first indication of these three dimensions was found in chapter 6; the six
clusters represented in the selection of items were found to describe agreeableness,
disagreeableness, extroversion, introversion, conscientiousness and carelessness. The idea
of three dimensions of product personality was supported by the confirmative factor
analysis on the data of study 1 in section 7.5. The 18 items of the product personality
scale were forced into a three-factor solution. The resulting dimensions were easily
interpretable as agreeableness, extroversion and conscientiousness (see table 7.5).
However, the research in this thesis has only provided indirect evidence of the existence
of three dimensions of product personality. Future research should study this issue
directly.
Stability of product personality
The stability of product personality is assumed to be comparable to the stability
of human personality. As discussed in chapter 2, the debate about the cross-situational
(in)consistency of human personality was resolved by accepting the notion of relative
consistency (Schmitt & Borkenau, 1992). The idea of relative consistency suggests that
behavioral differences between two persons remain the same regardless of the situation,
even though their individual behavior may differ. We think that this notion also applies to
product personality. The absolute judgment of the personality of a product variant may
differ from one situation to another, though the relation between the personality of
product X and product Y will not change. For example, the watch in the middle of figure
8.3 is perceived as tougher when presented together with the refined and classic watch on
the left side of figure 8.3. Yet it appears to be less tough when it is presented with an
extremely rugged watch like on the right side of figure 8.3. Nonetheless, the watch in the
middle can be described as tough in both situations and it will always remain tougher than
the watch on the left.
149
Figure 8.3: A tough watch: the relative stability of product personality
An issue in line with the stability of product personality in different situations is
the question of whether product personality is stable throughout time. Product
personality is probably relatively stable throughout time. Yet, since it is a way of
describing the distinction between different product variants, product personality will
always be a comparative judgment. As such, it will be influenced by new product variants
that enter the market. If all new watches were to be as rugged as the watch on the righthand side of figure 8.3 and if all of the classic watches (such as the watch on the left-hand
side) were to disappear, then, in time, the watch in the middle would become refined.
Implications
The research with respect to the perception of product personality showed that
product appearance is an important factor in the perception of product personality and
that product personality is perceived similarly by different people. It is also suggested that
product personality is relatively stable in different situations. As a consequence, it is
possible for designers to create products with a personality that is perceived consistently
by different people and remains constant in varying situations.
Since product personality is determined by product appearance and is perceived
consistently by different people and in various situations, it can be used as a tool to
communicate symbolic meaning to consumers. By giving different product variants of the
same brand similar appearances, a company can tune its products to one another.
Companies can use product appearance to communicate the personality of a brand or the
corporate image. The Siemens Porsche line shown in figure 8.4, is an example where
products from different product classes are attuned to one another by means of
appearance, in order to express a similar image.
On the other hand, companies can also use product personality to differentiate
between segments, while keeping the functionality of the products equal (giving the same
product different appearances). An example of differentiating while keeping functionality
similar is provided by the faceplates of Nokia shown in figure 8.2.
150
Figure 8.4: The Siemens Porsche product line
Limitations
An important limitation of the research addressing the issue of perceiving
product personality is that we only studied the influence of product appearance on the
perception of product personality. The method we used resembled the method used in
“zero-acquaintance” studies in person perception. In zero acquaintance studies,
personality impressions are based on minimal exposure to the target, either in the form of
pictures, short descriptions (vignettes) or short pieces of film. The product personality
impressions in our studies were based on pictures of the stimuli. As a consequence, we
can only assume that the slam of a car door, the percolating of coffee in a coffeemaker,
the growling of a motorcycle, or the sliding of a cd-drive influence the perception of
product personality. This limitation is put in perspective, however, when it is realized that
people in real life often purchase products on the same minimal information. In most
cases, a consumer can only see (not use) the product before buying it.
The focus on appearance characteristics might also have influenced the scope of
product personality as a concept. Both human personality and brand personality exist of
five factors. Yet, we only found an indication of three. These three “factors” of product
personality resemble the first three dimensions of human personality, extroversion,
agreeableness and conscientiousness. The results of our studies give no indication of the
existence of two dimensions that resemble the remaining two dimensions of human
personality, neuroticism and openness to experience. The absence of neuroticism and
openness to experience in product personality could be caused by the fact that
neuroticism and openness are parts of personality that only become apparent in
interaction (Kenny et al., 1994). Since we studied still images of products, we excluded
interaction dimensions and, by doing so, we probably ruled out the possibility of finding
aspects of neuroticism and openness to experience in product personality.
151
The fact that we did not find an equivalent of neuroticism and openness to
experience in product personality could be caused by the fact that both dimensions are
difficult to transfer to products. It seems reasonable to assume that because of its
emotional valence, neuroticism does not only apply to non-humans. Though products
can evoke emotions (Desmet, 2002), attributing emotional states to products is a form of
anthropomorphism associated with primitive people who attribute emotions to
sculptures, trees, and animals (Caporael, 1986). Openness to experience relates to
intelligence and is quite a cognitive concept. Common sense argues that cognitive abilities
are restricted to living creatures and are hardly applicable to vacuum cleaners or
coffeemakers. However, the attribution of human characteristics to non-human entities is
widespread in modern life (Carporael, 1986). Maybe by extending the research with
interaction dimensions, and so-called “smart” products (Rijsdijk & Hultink, 2002), we will
find neuroticism and openness dimensions in product personality.
8.4
We wanted to increase the actionability of product personality because product
personality can influence consumer preference. Only if product personality affects
consumer preference is it effective to build products with a particular personality.
Therefore, we investigated the influence of product personality on consumer preference,
hypothesizing that “similars attract”.
The similarity-attraction theory of human interaction states that people are
attracted to other people who are similar to themselves. Similarity between individuals is
thought to be gratifying because each individual validates and reinforces the self-concept
of the other. Since products also allow people to validate and reinforce their self-concept,
we investigated whether the similarity-attraction relationship would apply to humanproduct interaction. In analogy to user-image congruence, we converted the similarityattraction relationship of human interaction to a product-personality congruence effect.
Consumers were assumed to compare themselves with the personality of a product
variant and expected to prefer product variants with a personality similar to their selfconcept.
The results of the first study (n = 48) investigating the influence of product
personality on consumer preference (section 4.2) confirmed this expectation. Productpersonality congruence turned out to be a significant predictor of product evaluation (ß =
.48; t(442) = 13.08, p < .001). The results also indicated that product personality
congruence had a positive influence on product evaluation, in addition to and
independent of the user-image congruence effect. The results of a second study (section
152
4.3) showed that the positive effect of product-personality congruence is also present in a
post-purchase situation. In two subsequent experiments (both n = 90), we investigated
whether people are thought to be more satisfied with, and more attached to product
variants with a product personality that is similar to their own personality. One study
investigated the satisfaction and attachment of a conscientious person with/to a
conscientious product variant as compared to a non-conscientious product variant. The
other study replicated this experiment using extroversion as stimulus dimension. The
results suggested that similarity indeed increased satisfaction as well as attachment. Three
out of four hypotheses were confirmed (see sections 4.3.2 and 4.3.4). However, the effect
of product-personality congruence was stronger for satisfaction than for attachment.
It was concluded that consumers do not only evaluate product variants more
positively when they perceive them as similar to themselves, they are also more satisfied
with, and more attached to product variants with a self-congruent product personality.
These results indicate that product-personality congruence, user-image congruence and
brand-personality congruence have a comparable effect on consumer preference. Userimage congruence and brand-personality congruence increase consumer preference
(Sirgy, 1982; Aaker, 1999), as does product-personality congruence. It is suggested that
brand personality can increase brand loyalty and, therefore, can result in a lasting
relationship between the consumer and his/her brand (Fournier, 1998). Similarly, our
results suggest that product-personality congruence can result in the development of a
lasting relationship between the person and the product variant because the person is
more satisfied with the product variant and may develop a sense of attachment.
The positive effect of product-personality congruence is especially important in
light of further individualization. People pride themselves on their independence and
unique styles. Consumers feel good about their choices because they fit their own selfconcept rather than the group norm. (Solomon, 1999). Product-personality congruence
links up with this development because it allows consumers to express their individuality
through the selection of a distinctive product variant.
However, the relative importance of the product-personality congruence effect
on consumer preference has to be noted. Consumer preference is known to be
determined by a great variety of factors, such as price, product quality, and product
attributes. The effect of product personality on consumer preference, though significant,
determines only a limited proportion of consumer choice. Together with user-image
congruence, product personality congruence explained more than one third of the
variance in product evaluation (table 4.3). However, the relation between product
evaluation and consumer choice is a relation between an attitude and behavior.
Correlations between attitudes and behavior have been criticized to be low (Wicker,
1969). Nevertheless, attitudes can and do predict behavior (Azjen & Fishbein, 1977).
153
Product personality is part of the symbolic value of a product, in addition to the userimage and brand image. As such, product personality determines part of a consumer’s
choice.
Implications
The positive influence of product-personality congruence on consumer
preference stresses the importance of product personality as a tool for directing product
development. We already stated that product personality could be designed. Now it turns
out that product personality has a positive influence on consumer preference. People
prefer products that are similar to themselves and that allow them to express their selfconcept. Therefore, it is apparent that if companies design product variants according to
a pre-determined personality that matches the personality characteristics shared by the
members of their target group, then they can create preference, stimulate satisfaction and
encourage product attachment.
The product-personality congruence effect can be used to expand the width of a
brand, or to enlarge its market share. In order to communicate a consistent image, the
product variants that fall under the umbrella of a global brand should have product
personalities that match the brand personality (see the Siemens Porsche line in figure 8.4).
However, the personalities of the product variants do not have to be identical to that of
the global brand. In order for product personality to sustain the brand personality, they
should share a family resemblance, but there is enough room for differentiation. When
offering a range of product variants that fulfill the same functional need, but have
different product personalities, a company enables more consumers to select a product
variant that expresses their own individuality (see the Nokia faceplates in figure 8.2).
Another way to take advantage of the fact that consumers prefer products with a
self-congruent personality is through mass customization. Mass customization enables
consumers to specify their individual preferences from a comprehensive list of options,
with the product being built to order before delivery (Weightman & McDonagh, 2003).
This allows consumers to create their own, unique product that matches their selfconcept. Consumers are more satisfied with products they have created themselves (Vink,
2003). Part of this satisfaction could be due to the fact that the self-made products are
made to fit the consumer’s self-concept. Several manufacturers use the opportunities of
the World Wide Web to offer this extra benefit to consumers. For example, Nike offers
consumers the possibility to create their own bag or shoes (nikeid.nike.com). Giving
consumers the opportunity to create a product variant that matches their self-concept
provides competitive advantage through the product-personality congruence effect.
154
Limitations
A limitation of the research in this thesis with respect to consumer preference
concerns the fact that the relation between brand personality and product personality has
not been studied directly. There are some indirect arguments in this thesis to suggest that
these two concepts are independent. We used different product variants from the same
global brand in one study (e.g. three Philips coffee makers in section 3.2, two Siemens
vacuum cleaners, and two Nilfisk vacuum cleaners in section 6.9). These product variants
that have the same brand proved to have different product personalities (see for example
figure 1.2, page 13). Furthermore, user-image congruence and product personality
congruence were found to be independent determinants of consumer preference. All
things taken into account, we feel confident assuming that brand personality congruence
and product personality congruence are also independent influences. Yet, future research
will need to confirm this assumption and investigate the exact relation between the two
concepts.
8.5
The results of the first studies in this thesis have shown that the concept of
product personality exists. People do describe products using human personality
characteristics. Further results showed that the appearance of a product variant is an
important determinant of product personality. This means that products can be designed
to have a particular product personality. The question that remained was: why would one
design a product with a particular product personality? Designing a product with a predetermined product personality is attractive, because it appears that people prefer
products with a personality that is similar to their own. However, in order to design
products with a pre-determined personality, it is important to know which appearance
characteristics consumers associate with a particular personality characteristic. As a final
step, we therefore set out to develop a scale for assessing product personality. This scale
can be used to gain information about product personality and the appearance
characteristics that are associated with specific personality characteristics.
After a careful consideration of the literature, we concluded that there was no
existing scale for assessing product personality. There are scales for assessing user-image
or brand personality, and there are several scales developed for human personality
assessment. Yet, none of these scales meet the defining components of product
personality. Scales of human personality were thought to be inappropriate because
product personality is not the same as human personality. Product personality is more
superficial than human personality, since products lack the complexity and flexibility of
155
humans. Human personality is used as metaphor, trying to “understand one thing in terms of
another” (Lakoff & Johnson, 1980, p. 5). People are experts in relating to other people and
they exploit this ability in dealing with objects (Laurel, 1991). Metaphors draw incomplete
parallels between dissimilar things, accentuating some aspects and concealing others
(Lakoff & Johnson, 1980). Regarding the description of products using human
personality characteristics as a metaphor means that product personality does not have to
be identical to human personality. It is an incomplete parallel and, as such, human
personality scales cannot be used to assess product personality. The product personality
scale had to include human personality characteristics that are used to describe the
impression of product variants (not its users or their brand image) from different classes
of durable consumer products.
The development of the product personality scale followed the general steps of
scale development. First, we formulated a definition of product personality and
established its defining components (section 5.2). Secondly, the format of measurement
was determined (section 6.3) and an initial item pool of 1142 personality items was
generated (sections 6.4 and 6.5). Next, this item pool was reviewed and reduced with
respect to defining components of product personality. It turned out that 78 personality
characteristics were relevant for describing durable consumer products (sections 6.6 to
6.8). These 78 characteristics were reduced to a more manageable set of 20 items (section
6.9).
In order to reduce the chance of multiple interpretations of one item, the
selected items were extended with lexical descriptions and visuals (section 7.3 and 7.4).
The final scale consisted of 18 of these extended items (see table 7.2). The two final
studies in this thesis (n = 25 and n = 66, section 7.5) showed that these 18 items provide
different personality profiles for different product variants. These personality profiles are
perceived consistently by all respondents (Cronbach alphas exceed .80, and intraclass
correlations are significant at p < .001) and discriminate between product variants of the
same product class. It may thus be concluded that the scale is a reliable measure of
Implications
The goal of developing a product personality scale was to allow systematic
assessment of product personality. This systematic assessment should provide insight into
the appearance characteristics associated with the personality characteristics of product
personality. Ultimately, this knowledge could be used to design new product variants that
have a particular product personality. The underlying assumption was, of course, that
product variants that are described with the same personality characteristics share certain
appearance characteristics. This assumption was based on the results of our first studies.
156
The results of the product personality scale in later studies (chapter 6 and 7) strengthen
this idea. See, for example, the products in figure 7.10. All of them are perceived as
serious and all of them have basic, rather robust forms. How the product personality
scale provides insight into this relation between product personality and product
appearance is also shown by the examples in figure 8.5 and is discussed in section 8.6.
Limitations
A first limitation of the scale is the fact that the product personality scale can
never be a fixed scale. The meaning of the items changes over time. For example, “gay”
used to mean “happy” and therefore was included in earlier scales (Wells et al., 1957).
Over the centuries, the primary meaning of “gay” has changed to “homosexual”. This
change of meaning implies that the items will need constant development and renewal.
This limitation is not unique to product personality, however. Every scale needs to be
updated once in a while. A measure of social desirability, for example, should contain
different items now than it used to contain in the past.
A second limitation concerns the language of the scale. The scale is in Dutch and
cannot simply be translated into another language, due to a discrepancy in the social
desirability of terms. “Aggressive”, for example, is mildly favorable in American English,
but its translation “agressief” is clearly unfavorable in Dutch. “Critical” (“kritisch” in
Dutch) shows exactly the opposite pattern (Hofstee, 1990). Translating the scale would
thus alter its content. If one were to translate the scale, then the same method we applied
could be used. The selection of 78 items could be translated using a two-way translation
method. Adding characteristics from qualitative research in the native language could
reduce further differences between languages.
Another problem with the applicability in other language groups or cultures is
the dimensionality of product personality. The question is whether the personality
characteristics that are found to be relevant in describing product personality (or the field
of meaning that they represent) are stable across cultures. Existing knowledge about
personality does not offer an unequivocal answer. Whereas the five factors of human
personality have been proven to be consistent in different cultures (Costa & McCrae,
1988), the dimensions of brand personality have not (Aaker et al., 2001). We suggest that
the applicability of the product personality scale depends on the distance between
cultures. The symbolic meaning of products is determined by the culturally constituted
world (McCracken, 1986). If the meaning structures of two cultures are alike, then the
perception of a product’s personality by people in either culture is probably comparable.
If the meaning structures of two cultures are very different, then other terms become
important. In this case, the perception of product personality will also differ.
157
The final limitation of the product personality scale lies in the use of pictures.
For one thing, culture also influences the meaning of symbols (e.g. a white dove
symbolizes peace in one culture and death in another). The pictures we used to expand
our items also serve as symbols. The expansion of the items with visuals and lexical
descriptions may have increased the cultural anchoring of the scale, since we added more
symbols each with their own cultural meaning. Though valid and reliable assessment of
product personality is possible through the use of the current product personality scale,
the applicability of the scale is limited to The Netherlands. A second limitation with
respect to the pictures is that we have not been able to establish that the product
personality scale with the visuals (and lexical descriptions) is a better measure than the
same scale with single worded items. Future research should look into the effect of
including visuals and lexical descriptions on the reliability of the scale. Based on the
current research, we only know that the scale with the visuals and lexical descriptions is
reliable. We do not know if this scale is better than a scale that consists of only verbal
items.
8.6
Actionability of product personality
This thesis presents the concept of product personality and has made a start in
making the concept more actionable. The appearance of a product is proven to be a
determinant of product personality. As a consequence, products can be designed to have
a particular product personality. This is an attractive strategy because it has been found
that people prefer products with a personality that is similar to their own. In order to
facilitate the design of products with a pre-determined personality, we have developed a
product personality scale. Frequent use of the scale will lead to the accumulation of
knowledge about the appearance characteristics that consumers associate with a particular
personality characteristic. That knowledge can subsequently be used as a base for the
design of new products, thereby allowing product personality to be designed.
An example of the kind of information that frequent measurement with the
product personality scale can provide is shown in figure 8.5. The product variants used as
stimuli in this thesis are combined per dimension of related product personality
characteristics. This example shows that in the combination of product variants with the
same personality characteristics can provide valuable information for product designers.
The fact that the knowledge that is generated by combining products with the same
associations is valuable for design is also mentioned by Zaltman (2003). With respect to
designing a friendly car, Zaltman (2003) states that especially the use of “nonautomobile
examples” provided a “more complete understanding” (p. 89) of what was meant by “friendly”.
158
In addition, the use of product examples proved that they are more useful than the verbal
cues that traditionally result from market research.
Agreeableness (+)
Cheerful
Open
Pretty
Cute
Relaxed
Agreeableness (-)
Aloof
Boring
Extroversion (+)
Dominant
Idiosyncratic
Interesting
Lively
Provocative
Extroversion (-)
Modest
Conscientiousness (+)
Honest
Serious
Conscientiousness (-)
Silly
Childish
Untidy
Figure 8.5: Product variants representing the dimensions of product personality
Based on the product variants shown in figure 8.5, agreeableness versus
disagreeableness could cautiously be interpreted as round shaped versus sharp edged.
Extroversion versus introversion could be characterized by closed versus space-extending
or open design. Conscientiousness versus carelessness could be associated with a
combination of basic, robust forms and the absence of colors versus atypical shapes and a
159
colorful design. In designing products with a specific product personality, one needs to
keep in mind, however, that all appearance aspects influence a product’s personality.
“Slight changes in design can drastically change the metaphor conjured up by customers. With watches, as
with human faces, the difference between a mean and a friendly look can be very subtle ..” (Zaltman,
2003, p. 88). Nonetheless, there seems to be a difference in the importance of appearance
features in the perception of product personality. A change of color or a change of form
probably leads to a larger change in a product’s personality, than changing any of the
other aspects of a product’s appearance. Yet, not only pure appearance characteristics,
such as product form and color, influence the perception of product personality.
Functional differences also influence the perception of product personality.
It may be advantageous to develop a database in order to track the combination
of product variants with the same personality characteristics. In a database, numerous
product variants can be stored along with their scores on the product personality items.
Continuous use of the scale adds to the source of information stored in the date base.
Changes in the relation between the personality characteristics and certain appearance
characteristics would then become visible.
8.7
Suggestions for future research
The previous sections have discussed the results and practical implications of the
research that we conducted in this thesis. However, the results of research also lead to
suggestions for further research. Some of these suggestions were mentioned in the text.
This final section discusses some more suggestions for future research. The suggestions
are mentioned in the order of the issues just discussed. Ideas about the perception of
product personality are mentioned first, followed by suggestions about the influence of
product personality on consumer preference. Finally, we will conclude with research
suggestions with respect to the assessment of product personality.
The influence of appearance on the perception of product personality
In this thesis, product appearance has only been investigated as a determinant of
product personality. Though the results indicate that appearance is an important factor in
the perception of product personality, further research is required for determining which
appearance characteristics are associated with specific personality characteristics. The
examples shown throughout the thesis, and the suggestions about specific relations
between product form/proportion/color and personality characteristics, are
circumstantial. Only the study of the irons (section 3.3) studied the relation between
product appearance and product personality directly. Yet, this study was focused on how
160
designers use product appearance to design a product variant expressing a certain
personality characteristic. The study was not focused on the relation between product
appearance and product personality according to the consumer. In order to further
increase the actionability of product personality, future research should address this issue.
For example, future research could systematically manipulate the appearance
characteristics of a product and assess the influence this may have on the perception of
the product’s personality.
The influence of situation on the perception of product personality
The research in this thesis has investigated the role of product characteristics on
the perception of product personality. The influence of the perceiver is considered to be
less important. However, perception is a combination of stimulus, perceiver and
situation. The influence of the situation on the perception of product personality is only
mentioned shortly in section 8.3 when the influence of other product variants was
discussed. The influence of other product variants that are presented together with the
target product should be further investigated. It is also interesting to look at the influence
of store image. One can imagine, for example, that a watch offered at the HEMA is
perceived differently than the same watch at a jeweler.
Relation between brand personality and product personality
It was already mentioned in section 8.4 that we have not studied the relation
between brand personality and product personality directly. Though, we did use different
stimuli from the same brand in one study and found different product personality, more
knowledge is needed about the exact relation between the two concepts. For example, are
they really independent factors in consumer preference? And, how are the dimensions of
brand personality and product personality related?
Dissimilarity-avoidance instead of similarity-attraction?
With respect to the influence of product personality on consumer preference, an
interesting thought has been posited in recent literature about the influence of the
undesired self. Next to the basic assumption of the similarity-attraction hypothesis (people
prefer products that are similar), people are said to also consciously avoid products that
are dissimilar to their self-concept (or similar to their undesired self). This line of thinking
is based on the idea that the so-called undesired self (Ogilvie, 1987), rather than the real
or ideal self, is a person’s point of reference. The argument is that “… without a tangible
undesired self, the real self would loose its navigational cues” (Ogilvie, 1987, p.380). Recently,
several authors have investigated this idea with respect to consumer research (Banister &
Hogg, 2001; Englis & Solomon, 1995, 1997; Hogg & Banister, 2001). They conclude that
161
the undesired self is an important factor in shaping consumer preference. “…The undesired
end state (= undesired self) will function as an incentive to avoid products with negative images”
(Hogg & Banister, 2001, p. 73), and “… ‘good taste’ exists only through a thorough knowledge of
what constitutes ‘bad taste’” (p. 76). This suggests that products that are perceived as
dissimilar to the self-concept will be avoided or even rejected. In addition, consumers
derive their perception of what is similar to their self-concept from what they perceive as
dissimilar. When taken to the extreme, this would mean that people would prefer
products with a personality that reflects the opposite of their undesired self.
The fact that the avoidance of dissimilar products influences consumer
preference does not exclude the influence of similarity on consumer preference. Ogilvie
(1987) contents that there is “….both a push and a pull…” (p. 383), implying that consumer
choice is determined by both what people avoid and what they prefer. Future research
should investigate how similarity and dissimilarity combine to determine product
personality preference.
Extending the applicability of the product personality scale
The items in the product personality scale are selected to measure product
personality defined as “the human personality characteristics used to describe and
differentiate between durable consumer product variants”. As a consequence, the scale
cannot be used out of this context without some extra research. However, we did find
that the scale can assess the personality of logos. We think that it can also be used to
assess the personality of other objects that are tangible, survive many uses, and have no
extra aspects that can influence the perceived personality beyond their appearance (such
as art or photographs). It is interesting for future research to investigate other fields of
application.
It would also be interesting to look into the cross-cultural differences in product
personality. Adapting the product personality scale to other cultures (or developing a new
scale) would provide insight into cross-cultural differences in product personality. This is
especially relevant, since more and more companies are entering the global market.
162
163
Summary
Product Personality (Pascalle C.M. Govers)
Personality characteristics are terms such as serious, happy and tough that people
use to describe the personality of other people. However, personality characteristics are
not only used to describe their impression of persons; they are also used to describe
products. The profile of personality characteristics used to describe and discriminate a
single product variant is what we call product personality.
Product personality is part of the symbolic meaning of a product. Symbolic
meaning refers to the fact that products have a significance that goes beyond their
utilitarian function. The symbolic meaning of products has been studied from a social
perspective (focusing on products as symbols of status and group membership), and from
a more individualistic perspective. Product personality best fits the tradition of the
individualistic perspective. The individualistic perspective has largely focused on products
as symbols of the self-concept. People have a self-concept and are motivated to confirm
this self-concept. People are therefore suggested to prefer products that match their selfconcept. Many studies have investigated this suggestion and they generally found
confirming evidence: people prefer those products that are consistent with their selfconcept.
These studies have, however, been conducted with the underlying goal of
understanding human behavior. As a consequence, most of the research is not directed at
products and it is difficult to apply these results directly in product design. A designer
may know that people use products as symbols to communicate their self-concept, but
this knowledge does not provide insight into how (s)he can design a product that matches
it. How should (s)he create a serious, happy or tough product? It is important to know
the answer to this question because the symbolic meaning of products has become
increasingly important. The market offers product variants in a variety of different shapes
and colors, which all meet the functional needs of the consumer. For the consumer, the
symbolic meaning of a product forms a means of choosing among all these variants. For
companies, the symbolic meaning is a way to differentiate themselves from the
competitor. The contribution of product personality lies in the fact that it describes the
symbolic meaning at a more specific level, the level of product variants. Product designers
work at the level of product variants, they create product variants and need information
at this level.
The question remains, however, which actions a designer would have to take in
order to create a serious, happy or tough product. Literature addresses this problem as
one of actionability and the aim of this thesis was to increase the actionability of product
personality. It did so by addressing three issues: (1) the perception of product personality,
164
Summary
(2) the influence of product personality on consumer preference, and (3) systematic
assessment of product personality. It was argued that information about the perception
of product personality, the influence of product personality on consumer preference, and
the systematic assessment of product personality would provide designers with a basis for
better adapting the products that they design to the wishes of the consumer.
In order to establish that product personality was indeed a concept worth
studying, we first needed to confirm the assumption that people describe products using
personality characteristics. The results of our first study showed that a reasonable
proportion of the initial verbal descriptions of watches (30%) was indeed personality
related. This proportion of personality related answers was remarkably high if we take
into account that personality characteristics make up only a little more than 40% of
person descriptions.
After we had established that product personality existed, we focused on how
people perceive product personality. Since the perception of personality in humans is
strongly influenced by appearance characteristics, we investigated the influence of
product appearance on the perception of product personality. We conducted two studies
(reported in chapter 3). The stimuli in the first study differed in appearance on many
accounts: color, size, form, etc. The differences between stimuli in the other study were
restricted to product form. Yet, we found differences in product personality between the
stimuli in both studies. Furthermore, a comparison of the product variants that have the
same product personality revealed that product variants from different product types that
are described using the same personality characteristics had similar visual appearance
characteristics. Based on these results, we concluded that product appearance is an
important determinant of product personality.
The second issue that we investigated was whether, and how, product
personality influenced consumer preference (chapter 4). Only if product personality
affects consumer preference, is it effective to build products with a particular personality.
We assumed that, since it is known that people evaluate other people more positively
when they perceive them as similar to themselves (in terms of demographics, attitudes
and personality), people might also evaluate products more positively when they perceive
them as similar to themselves in terms of personality. Studies in consumer research have
also reported findings that supported this assumption.
Our first study into the influence of product personality on consumer preference
studied the influence of product personality on product evaluation. The results confirmed
our expectation. Product-personality congruence appeared to be a significant predictor of
product evaluation. A subsequent study showed that the positive effect of productpersonality congruence is also present in a post-purchase situation. Respondents expected
a target person to be significantly more satisfied with and attached to product variants
Summary
165
that had a product personality that was similar to them. It was concluded that productpersonality congruence positively influenced product preference in a pre-purchase as well
as in a post-purchase situation.
The third and last issue that we addressed in this thesis was the systematic
assessment of product personality. After a careful consideration of the literature, we
concluded that there was no existing scale for systematically assessing product personality
(chapter 5). We therefore set out to develop a product personality scale ourselves. The
goal of this scale would be to enable the assessment of product personality of many
product variants from different product classes. The measurement results of this scale
could be used to find out which appearance characteristics consumers associate with a
particular personality characteristic. This knowledge could then be used for designing
products with a pre-determined personality.
The development of the product personality scale followed the general steps of
scale development. First, we formulated a definition of product personality and
established its defining components. Then, the format of measurement was determined
and an initial item pool of was generated. Next, this item pool was reviewed with respect
to defining components of product personality, and reduced to a manageable number
(chapter 6). A final stage in the development concerned the unequivocalness of the
remaining items. In order to reduce the chance of multiple interpretations of one item, we
added an extra developmental step. The selected items were extended with lexical
descriptions and visuals. To show that this (extended) scale was a reliable measure of
product personality we conducted two studies (chapter 7). The results of these studies
showed that the final product personality scale provided different personality profiles for
different product variants. These personality profiles were perceived consistently by all
respondents. We thus concluded that the scale we developed is a reliable measure of
In conclusion, this thesis presented the concept of product personality and has
made a start in making the concept more actionable. First, by showing that the
appearance of a product is a determinant of product personality and thereby showing that
products can be designed to have a particular product personality. Next, by indicating
that is an attractive strategy because people prefer products with a personality that is
similar to their own. Furthermore, in order to facilitate the design of product personality,
a product personality scale was developed. Frequent use of this scale will lead to
knowledge about the appearance characteristics that consumers associate with a particular
personality characteristic. This knowledge can subsequently be used as a base for the
design of new products, thereby improving the actionability of product personality.
166
Samenvatting
Productpersoonlijkheid (Pascalle C.M. Govers)
Persoonlijkheidseigenschappen zijn termen zoals vrolijk, eigenzinnig en serieus.
Het zijn termen die gebruikt worden om de persoonlijkheid van mensen te omschrijven.
Ze worden echter niet alleen gebruikt om mensen te beschrijven, ze worden ook gebruikt
om producten te beschrijven; een product kan ook vrolijk en eigenzinnig zijn. Het profiel
van dergelijke persoonlijkheidseigenschappen, gebruikt om de indruk van een individueel
product te beschrijven, wordt productpersoonlijkheid genoemd.
Productpersoonlijkheid is onderdeel van de symbolische betekenis van een
product. Met symbolische betekenis wordt verwezen naar het feit dat producten méér
dienen dan alleen hun functionele doel. Producten worden gebruikt als symbolen. De
symbolische betekenis van producten is in de literatuur op twee manieren benaderd: (1)
vanuit sociaal perspectief (producten als symbolen van status en groepslidmaatschap) en
(2) vanuit het perspectief van het individu. Productpersoonlijkheid sluit aan bij dit tweede
perspectief. Het onderzoek naar symbolische betekenis vanuit het perspectief van het
individu richt zich voornamelijk op het gebruik van producten als symbolen van het
zelfbeeld. Ieder mens heeft een zelfbeeld en gedraagt zich consistent met dat beeld. Mede
daardoor hebben mensen een voorkeur hebben voor producten met een symbolische
betekenis die overeenkomt met hun zelfbeeld.
Dit gegeven is echter niet als zodanig toepasbaar in de productontwikkeling. Een
ontwerper kan weinig met de informatie dat consumenten de voorkeur geven aan
producten die overeenkomen met hun zelfbeeld. Wanneer bekend is dat consumenten
een eigenzinnige en vrolijke mobiele telefoon willen, moet een ontwerper wel weten wat
de consument daarmee bedoelt. Hoe ziet een vrolijk en eigenzinnig product eruit? Het
antwoord op deze vraag is van belang, omdat de symbolische betekenis van producten
steeds belangrijker wordt. Producten worden aangeboden in allerlei vormen, maten en
kleuren, en elk individueel product vervult de functionele behoefte van de consument. De
symbolische betekenis van producten is voor consumenten een manier om te kiezen
tussen de voor hem/haar anderszins gelijke producten. Voor bedrijven is de symbolische
betekenis een manier om zich te onderscheiden van de concurrent. De bijdrage van
productpersoonlijkheid zit in het feit dat het de symbolische betekenis van individuele
producten omschrijft, en ontwerpers werken op het niveau van individuele producten.
De vraag blijft echter, wat moet een ontwerper doen om een product te
ontwerpen met een specifieke symbolische betekenis. Dit probleem wordt in de literatuur
een probleem van “actionability” genoemd. Dit proefschrift heeft tot doel de
“actionability” van productpersoonlijkheid te vergroten. Met dit doel voor ogen worden
drie onderwerpen behandeld: (1) de perceptie van productpersoonlijkheid, (2) de invloed
Samenvatting
167
van productpersoonlijkheid op de voorkeur van consumenten, en (3) het systematisch
meten van productpersoonlijkheid. De achterliggende gedachte was dat informatie over
de deze drie onderwerpen ontwerpers een basis zou geven om producten beter af te
stemmen op de wensen van de consument. In de behandeling van de drie onderwerpen is
steeds een vertaling gemaakt van kennis uit de psychologie.
De eerste studie had tot doel aan te tonen dat mensen inderdaad
persoonlijkheidseigenschappen gebruiken om producten te omschrijven. De resultaten
van deze eerste studie gaven aan dat ongeveer 30% van de termen die mensen gebruikten
om hun indruk van horloges te omschrijven gerelateerd waren aan menselijke
persoonlijkheid. Dit percentage is opmerkelijk omdat beschrijvingen van mensen voor
ongeveer 40% uit persoonlijkheidseigenschappen bestaan.
Nadat was aangetoond dat productpersoonlijkheid bestaat, heeft het onderzoek
zich gericht op de invloed van het productuiterlijk in de perceptie van
productpersoonlijkheid. Het is namelijk bekend dat bij de perceptie van mensen het
uiterlijk een grote invloed heeft. Er zijn twee studies uitgevoerd (hoofdstuk 3). De stimuli
in de eerste studie verschilden op verschillende aspecten in uiterlijk (kleur, vorm, grootte,
etc.). De stimuli in de tweede studie verschilden alleen in vorm (het betrof zwart-wit
schetsen van strijkijzers). In beide studies bleken de verschillen in productuiterlijk ook
verschillen in productpersoonlijkheid op te leveren. Bovendien bleek na een vergelijking
van de stimuli dat producten met dezelfde productpersoonlijkheid overeenkomstige
uiterlijke kenmerken hadden. Zowel stoere horloges als stoere strijkijzers bleken robuust
en grof van vorm te zijn. Later in het proefschrift werd deze observatie bevestigd toen
ook producten uit andere product categorieën vergeleken worden. Op basis van deze
informatie werd geconcludeerd dat het productuiterlijk een belangrijke determinant is van
productpersoonlijkheid, met als gevolg dat productpersoonlijkheid ontworpen kan
worden.
Vervolgens is gekeken naar de invloed van productpersoonlijkheid op product
voorkeur (hoofdstuk 4). Alleen wanneer productpersoonlijkheid de voorkeur van
consumenten beïnvloedt, is het voor een bedrijf zinvol een product met een bepaalde
productpersoonlijkheid te ontwerpen. Om met productpersoonlijkheid product voorkeur
te creëren moet echter ook bekend zijn hoe productpersoonlijkheid de voorkeur van
consumenten beïnvloedt. Vanuit de psychologie is bekend dat mensen andere mensen
met een achtergrond/persoonlijkheid/mening die overeenkomstig is positiever evalueren
dan mensen die van hen verschillen. Onderzoek naar consumentengedrag toont aan dat
mensen de voorkeur geven aan merken en producten met een imago dat overeenkomt
met hun zelfbeeld. Derhalve was de hypothese in dit proefschrift dat consumenten
producten met een persoonlijkheid overeenkomstig met hun eigen persoonlijkheid ook
positiever evalueren. De resultaten bevestigden deze hypothese. Overeenkomst tussen de
168
Samenvatting
persoonlijkheid van de consument en de productpersoonlijkheid resulteert in een
positievere evaluatie van het product. Tevens bleek dat ook de tevredenheid en
gehechtheid aan het product sterker zijn wanneer de persoonlijkheid van de persoon en
het product overeen komen. Op basis van deze resultaten werd geconcludeerd dat
overeenkomst tussen de consument en de persoonlijkheid van een product een positieve
invloed heeft, en dat deze invloed geldt zowel vóór als na de aanschaf van het product.
Tenslotte is er een productpersoonlijkheid schaal ontwikkeld. De schaal heeft tot
doel het meten van de productpersoonlijkheid van een grote hoeveelheid, verschillende
duurzame producten mogelijk te maken. De meetresultaten van een dergelijke schaal
kunnen gebruikt worden om inzicht te verkrijgen in de uiterlijke kenmerken die
consumenten associëren met bepaalde persoonlijkheidseigenschappen. Deze kennis kan
vervolgens weer gebruikt worden om producten met een bepaalde persoonlijkheid te
ontwerpen.
De productpersoonlijkheid schaal is ontwikkeld volgens de algemene stappen in
schaalontwikkeling. Eerst is een definitie van productpersoonlijkheid geformuleerd, op
basis daarvan zijn de definiërende kenmerken van productpersoonlijkheid vastgesteld. De
meetvorm is bepaald en vervolgens is er een grote hoeveelheid items verzameld. Daarna
zijn de verzamelde items beoordeeld waarbij de definiërende kenmerken van
productpersoonlijkheid als criteria dienden. Op deze manier is de verzameling items
gereduceerd tot een bruikbare hoeveelheid (hoofdstuk 6). Om de betekenis van de items
eenduidiger te maken zijn de items, als extra stap, uitgebreid met een afbeelding en een
omschrijving. Deze (uitgebreide) schaal is tot slot getest. De resultaten van twee studies
geven aan dat de productpersoonlijkheid schaal een betrouwbare methode is om
productpersoonlijkheid te meten. Verschillende producten hebben een verschillend
productpersoonlijkheidsprofiel, en dit profiel is consistent over consumenten.
Samengevat heeft dit proefschrift het concept “productpersoonlijkheid” ingeleid
en een begin gemaakt met “actionable” maken van het concept. Ten eerste door aan te
tonen dat productpersoonlijkheid in belangrijke mate bepaald wordt door het
productuiterlijk. Dit betekent dat productpersoonlijkheid ontworpen kan worden.
Vervolgens door te laten zien dat dit een aantrekkelijke strategie is omdat overeenkomst
tussen de consument en de productpersoonlijkheid leidt tot voorkeur bij de consument.
Tenslotte, is er een productpersoonlijkheid schaal ontwikkeld om het ontwerpen van
productpersoonlijkheid te faciliteren. Veelvuldig gebruik van de schaal zal leiden tot
kennis over de uiterlijke kenmerken die consumenten associëren met bepaalde
persoonlijkheidskenmerken. Deze kennis kan dan weer gebruikt worden door bedrijven,
waardoor de “actionabilty” van product persoonlijkheid verbeterd wordt.
169
References
Aaker, J. L. (1997). Dimensions of Brand Personality. Journal of Marketing Research, 34, 347356.
Aaker, J. L. (1999). The Malleable Self: The Role of Self-Expression in Persuasion. Journal
of Marketing Research, 36, 45-57.
Aaker, J. L., Benet-Martínez, V. & Garolera, J. (2001). Consumption Symbols as Carriers
of Culture: A Study of Japanese and Spanish Brand Personality Constructs.
Journal of Personality and Social Psychology, 81, 492-508.
Albright, L., Kenny, D. A. & Malloy, T. E. (1988). Consensus in Personality Judgments at
Zero Acquaintance. Journal of Personality and Social Psychology, 55, 387-395.
Anderson, N. H. (1968). Likableness ratings of 555 personality-trait words. Journal of
Personality and Social Psychology, 9, 272-279.
Anderson, S. M. & Klatzky, R. L. (1987). Traits and Social Stereotypes: Levels of
Categorization in Person Perception. Journal of Personality and Social Psychology, 53,
235-246.
Antill, J. K. (1983). Sex-Role Complementarity Versus Similarity in Married Couples.
Arrindell, W. A. & Luteijn, F. (2000). Similarity between Intimate Partners for Personality
Traits as Related to Individual Levels of Satisfaction With Life. Personality and
Individual Differences, 28, 629-637.
Asch, S. E. (1946). Forming Impressions of Personality. Journal of abnormal and social
psychology, 41, 258-290.
d'Astous, A. & Lévesque, M. (2003). A Scale for Measuring Store Personality. Psychology
and Marketing, 20, 455-469.
Augoustinos, M. & Walker, I. (1995). Social Cognition: An Integrated Introduction. London:
Sage.
Ball, A. D. & Tasaki, L. H. (1992). The Role and Measurement of Attachment in
Consumer Behavior. Journal of Consumer Psychology, 1, 155-172.
Banister, E. N. & Hogg, M. K. (2001). Mapping the Negative Self: From “So Not Me” to
“Just Not Me”. Advances in Consumer Research, 28, 242-248.
Barry, W. A. (1970). Marriage Research and Conflict: An Integrative Review. Psychological
Bulletin, 73, 41-54.
Baumeester, R. F. (1998). The Self. In Gilbert, D. T., Fiske, S. F. & Lindzey, G. (Eds.),
The Handbook of Social Psychology (pp. 680-740). New York: Oxford University
Press.
170
References
Belch, G. (1978). Belief Systems and the Differential Role of the Self-Concept. Advances in
Consumer Research, 5, 320-325.
Belch, G. E. & Landon, E. L. (1977). Discriminant Validity of a Product Anchored SelfConcept Measure. Journal of Marketing Research, 14, 252-256.
Belk, R. W. (1978). Assessing the Effects of Visible Consumption on Impression
Formation. Advances in consumer behavior, 5, 39-47.
Belk, R. W. (1988). Possessions and The Extended Self. Journal of Consumer Research, 15,
139-168.
Belk, R. W. (1992). Attachment to Possessions. In Altman, I. & Low S. M. (Eds.), Place
Attachment (pp. 37-62). New York: Plenum Press.
Bellenger, D. N., Steinberg, E. & Stanton, W. W. (1976). The Congruence of Store Image
and Self Image. Journal of Retailing, 52, 17-32.
Bentler, P. M. & Cho, C-P. (1988). Practical Issues in Structural Modeling. In Long, S. J.
(Ed.), Common Problems/Proper Solutions: Avoiding Error in Quantitative Research (pp.
161-192). Newbury Park, CA: Sage.
Bentler, P. M. & Newcomb, M. D. (1979). Longitudinal Study of Marital Success and
Failure. In Cook, M. & Wilson, G. (Eds.), Love and Attraction: An international
conference (pp. 189-194). Oxford: Pergamon Press.
Berkowitz, M. (1987). Product Shape as a Design Innovation Strategy. Journal of Product
Innovation Management, 4, 274-283.
Biel, A. L. (1993). Converting Image into Equity. In Aaker, D. A. & Biel, A. L. (Eds.),
Brand Equity and Advertising (pp. 67-82). Hillsdale, New Jersey: Lawrence Erlbaum
Associates.
Birdwell, A. E. (1968). A Study of Influence of Image Congruence on Consumer Choice.
Journal of Business, 41, 76-88.
Bloch, P. H. (1995). Seeking the Ideal Form: Product Design and Consumer Response.
Journal of Marketing, 59, 16-29.
Borkenau, P. & Liebler, A. (1992a). The Cross-Modal Consistency of Personality:
Inferring Strangers’ Traits from Visual or Acoustic Information. Journal of Research
in Personality, 26, 128-204.
Borkenau, P. & Liebler, A. (1992b). Trait Inferences: Sources of Validity at Zero
Acquaintance. Journal of Personality and Social Psychology, 62, 645-657.
Borkenau, P. & Liebler, A. (1995). Observable Attributes as Manifestations and Cues of
Personality and Intelligence. Journal of Personality, 63, 1-25.
Bosveld, K. & Van de Kolk, S. (2003). Vijf designers over de rol van design in Nederland.
Tijdschrift voor Marketing, februari, 16-20.
Brehm, S. S. & Kassin, S. M. (1990). Social Psychology. Boston: Houghton Mifflin
Company.
References
171
Bruce, M. & Whitehead, M. (1988). Putting Design into the Picture: The Role of Product
Design in Consumer Purchase Behaviour. Journal of Marketing Research Society, 30,
147-162.
Bruner, J. S. (1957). On Perceptual Readiness. Psychological Review, 64, 123-152.
Byrne, D. (1961). Interpersonal Attraction and Attitude Similarity. Journal of Abnormal and
Social Psychology, 62, 713-715.
Byrne, D. (1971). The Attraction Paradigm. New York: Academic Press.
Byrne, D. (1997). An Overview (and Underview) of Research and Theory within the
Attraction Paradigm. Journal of Social and Personal Relationships, 14, 417-431.
Byrne, D., Clore, D. L. & Smeaton, G. (1986). The Attraction Hypothesis: Do similar
Attitudes Affect Anything? Journal of Personality and Social Psychology, 51, 1167-1170.
Byrne, D., Clore, G. L. & Worchel, P. (1966). The Effect of Economic SimilarityDissimilarity on Interpersonal Attraction. Journal of Personality and Social Psychology,
4, 220-224.
Byrne, D. & Griffitt, W. (1969). Similarity and Awareness of Similarity of Personality
Characteristics as Determinants of Attraction. Journal of Experimental Research in
Personality, 3, 179-186.
Byrne, D. & Nelson, D. (1965). Attraction as a Linear Function of Proportion of Positive
Reinforcement. Journal of Personality and Social Psychology, 1, 659-663.
Caporael, L. R. (1986). Anthropomorphism and Mechanomorphism: Two Faces of the
Human Machine. Computers in Human Behavior, 2, 215-234.
Carli, L. L., Ganley, R. & Pierce-Otay, A. (1991). Similarity and Satisfaction in Roommate
Relationships. Personality and Social Psychology Bulletin, 17, 419-426.
Carver, C. S. & Scheier, M. F. (1996). Perspectives on Personality. Needham Heights: Allyn
and Bacon.
Cattell R. B., Eber, H. W. & Tatsuoka, M. M. (1977). Handbook for the Personality Factor
Questionnaire. Champaign: IPAT.
Churchill, G. A. (1979). A Paradigm for Developing Better Measures of Marketing
Constructs. Journal of Marketing Research, 16, 64-73.
Clore, G. L. & Byrne, D. (1974). A Reinforcement-Affect Model of Attraction. In
Huston, T. L. (Ed.), Foundations of Interpersonal Attraction (pp. 143-170). New York:
Academic Press.
Cook, M. (1979). Perceiving Others. New York: Methuen & Co.
Costa, P. T. & McCrae R. R. (1988). Personality in Adulthood: A Six-year Longitudinal
Study of Self-Reports and Spouse Ratings on the NEO Personality Inventory.
Costa, P. T. & McCrae, R. R. (1992a). Normal Personality Assessment in Clinical
Practice: The NEO Personality Inventory. Personality Assessment, 4, 5-13.
172
References
Costa, P. T. & McCrae, R. R. (1992b). Four Ways Five Factors are Basic. Personality and
Individual Differences, 16, 653-665.
Creusen, M. (1998). Product Appearance and Consumer Choice. Delft: Delft University of
Technology.
Critelli, J. W. & Waid, L. R. (1980). Physical Attractiveness, Romantic Love, and Equity
Restoration in Dating Relationships. Journal of Personality Assessment, 44, 624-629.
Cronbach, L. J. & Furby, L. (1970). How Should We Measure ‘Change’ - Or Should We?
Psychological Bulletin, 74, 68-80.
Csikszentmihalyi, M. (2000). The Costs and Benefits of Consuming. Journal of Consumer
Research, 27, 267-272.
De Raad, B., Mulder, E., Kloosterman, K. & Hofstee, W. K. B. (1988). PersonalityDescriptive Verbs. European Journal of Personality, 2, 81-96.
Deaux, K. & Taynor, J. (1973). Evaluation of Male and Female Ability: Bias Works Both
Ways. Psychological Reports, 32, 261-262.
Desmet, P. (2002). Designing Emotions. Delft: Delft University of Technology.
Devellis, R. F. (1991). Scale Development: Theory and Applications. Newbury Park: Sage.
Diamantopoulos, A. & Winklhofer, H. M. (2001). Index Construction with Formative
Indicators: An Alternative to Scale Development. Journal of Marketing Research, 38,
269-277.
Digman, J. M. (1990). Personality Structure, Emergence of The Five Factor Model.
Annual Review of Psychology, 41, 417-440.
Dittmar, H. & Pepper. L. (1994). To Have is To Be: Materialism and Person Perception
in Working-Class and Middle-Class British Adolescents. Journal of Economic
Psychology, 15, 233-251.
Doddema, M. & De Raad, B. (1997). Idioticon van de persoonlijkheid. Amsterdam: Uitgeverij
Nieuwezijds.
Dolich, I. J. (1969). Congruence Relationship between Self Image and Product Brands.
Journal of Marketing Research, 6, 80-84.
Doumont, J-L. (2002). A Word is Worth a Thousand Pictures, Too. Technical
Communication, 49, 219-224.
Dryer, D. C. & Horowitz, L. M. (1997). When Do Opposites Attract? Interpersonal
Complementarity Versus Similarity. Journal of Personality and Social Psychology, 72,
592-603.
Dumaine, B. (1991). Design That Sells and Sells and ….. Fortune, 11, 56-61.
Duncan, S. L. (1976). Differential Social Perception and Attribution of Intergroup
Violence: Testing the Lower Limits of Stereotyping of Blacks. Journal of Personality
and Social Psychology, 34, 590-598.
References
173
Englis, B. G. & Solomon, M. R. (1995). To Be and Not to Be: Lifestyle Imagery,
Reference Groups and The Clustering of America. Journal of Advertising, 24, 13-26.
Englis, B. G. & Solomon, M. R. (1997). I Am Not Therefore, I Am: The Role of
Avoidance Products in Shaping Consumer Behavior. Advances in Consumer
Research, 24, 61-63.
Ericksen, M. K. & Sirgy, M. J. (1989). Achievement Motivation and Clothing Behavior: A
Self-Image Congruence Analysis. Journal of Social Behavior and Personality, 4, 307326.
Ericksen, M. K. & Sirgy, M. J. (1992). Employed Females’ Clothing Preference, SelfImage Congruence, and Career Anchorage. Journal of Applied Social Psychology, 22,
408-422.
Eysenck, H. J. (1947). Dimensions of Personality. New York: Praeger.
Feingold, A. (1988). Matching for Attractiveness in Romantic Partners and Same-Sex
friends: A Meta-Analysis and Theoretical Critique. Psychological Bulletin, 104, 226235.
Fiske, S. T. & Taylor, S. E. (1984). Social Cognition. New York: Random House.
Forty, A. (1995). Objects of Desire. London: Thames & Hudson.
Fournier, S. (1998). Consumers and their Brands: Developing Relationship Theory in
Consumer Research. Journal of Consumer Research, 24, 343-373.
Fournier, S. & Mick D. G. (1999). Rediscovering Satisfaction. Journal of Marketing, 63, 523.
Fromkin, H. L. & Snyder, C. R. (1980). Uniqueness, The Human Pursuit of Difference. New
York: Plenum Press.
Gilbert, D. T. (1998). Ordinary Personology. In D. T. Gilbert, S. T. Fiske and G. Lindzey
(Ed.), The Handbook of Social Psychology. Oxford: Oxford University Press.
Goffman, E. (1951). Symbols of Class Status. British Journal of Sociology, 2, 294-304.
Goffman, E. (1959). The Presentation of Self in Everyday Life. New York: Doubleday.
Goldberg, L. R. (1981). Language and Individual Differences; The Search for Universals
in Personality Lexicons. In Wheeler, L. (Ed.), Review of Personality and Social
Psychology (pp. 141-165). Beverly Hills, CA.: Sage.
Gordon, W. & Langmaid, R. (1995). Qualitative Market Research. Aldershot: Gower.
Gosling, S. D., Ko, S. J., Morris, M. E. & Mannarelli, T. (2002). A Room with a Cue:
Personality Judgments Based on Offices and Bedrooms. Journal of Personality and
174
References
Govers, P. C. M. & Schoormans, J. P. L. (2000). Intangible Product Attributes of
Watches: an exploratory study into the consensus of consumers’ reactions. In B.
Wieringa, A. Smidts & G. Antonides (Eds.), Marketing in the New Millennium,
Proceedings of the 29th EMAC Conference. Rotterdam: Erasmus Universiteit
Rotterdam.
Govers, P., Hekkert, P. & Schoormans, J. P. L. (2004). Happy, Cute and Tough: Can
Designers Create a Product Personality that Consumers Understand? In
McDonagh, D., Hekkert, P., Van Erp, J. & Gyi, D. (Eds.), Design and Emotion, The
Design of Everyday Things (pp. 345-349). London: Taylor & Francis.
Grubb, E. L. & Hupp, G. (1968). Perception of Self, Generalized Stereotypes and Brand
Selection. Journal of Marketing Research, 5, 58-63.
Grubb, E. L. & Stern, B. L. (1971). Self-Concept and Significant Others. Journal of
Marketing Research, 8, 382-385.
Hair, J. F., Anderson, R. E., Tatham, R. L. & Black, W. C. (1998). Multivariate Data
Analysis with Readings. New Jersey: Prentice-Hall, Inc.
Hays, W. (1988). Statistics. Orlando, Florida: Holt, Rinehart and Winston.
Heath, A. P. & Scott, D. (1998). The self-concept and image congruence hypothesis: An
empirical evaluation in the motor vehicle market. European Journal of Marketing, 32,
1110-1123.
Hendriks, A. A. J., Hofstee W. K. B. & De Raad, B. (1995). Handleiding bij de Five-Factor
Persoonlijkheids Vragenlijsten FFPI. Lisse: Swets & Zeitlinger.
Hendriks, A. A. J., Hofstee, W. K. B. & De Raad, B. (1999). The Five-Factor Personality
Inventory (FFPI). Personality and Individual Differences, 27, 307-325.
Higgins, E. T. (1996). Knowledge Activation: Accessibiltiy, Applicability and Salience. In
Higgens, E. T. & Kruglanski, A. W. (Eds.), Social Psychology: Handbook of Basic
Principles (pp. 133-168). New York: Guilford Press.
Hirschman, E. C. (1981). Comprehending Symbolic Consumption: Three Theoretical
Issues. In Hirschman, E. C. & Holbrook, M. B. (Eds.), Symbolic Consumer
Behaviour, Proceedings of the Conference on Consumer Esthetics and Symbolic Consumption
(pp. 4-6). New York: New York University.
Hjelle, L. A. & Ziegler, D. J. (1981). Personality Theories: Basic Assumptions, Research, and
Applications. Singapore: McGraw-Hill, Inc.
Hoekstra, H. A., Ormel, J. & De Fruyt, J. (1996). Handleiding bij de NEO Persoonlijkheids
Vragenlijsten NEO-PI-R en NEO-FFI. Lisse: Swets & Zeitlinger.
Hofmeester, G. H., Kemp, J. A. M. & Blankendaal, A. C. M. (1996). Sensuality in Product
Design: A Structured Approach. In Bilger, R., Guest, S. & Tauber, H. J. (Eds.),
Design Briefings (pp. 428-435). Vancouver: ACM Press.
References
175
Hofstee, W. K. B. (1990). The use of everyday language for scientific purposes. European
Journal of Personality, 4, 77-88.
Hogg, M. K. & Banister, E. N. (2001). Dislikes, Distastes and the Undesired Self:
Conceptualising and Exploring the Role of The Undesired End State in
Consumer Experience. Journal of Marketing Management, 17, 73-104.
Holman, R. (1980). Clothing as Communication: An Empirical Investigation. Advances in
Holman, R. (1981). Product Use as Communication: A Fresh Appraisal of a Venerable
Topic. In Enis B. M. and K. J. Roering (Ed.), Review of Marketing (pp. 106-119).
Chicago: American Marketing Association.
Hong, J. W. & Zinkhan, G. M. (1995). Self-Concept and Advertising Effectiveness: The
Influence of Congruency, Conspicuousness, and Response Mode. Psychology and
Marketing, 12, 53-77.
Hsu, S. H., Chuang, M. C. & Chang, C. C. (2000). A Semantic Differential Study of
Designers' and Users' Product Form Perception. Industrial Ergonomics, 25, 275391.
Hughes, G. D. & Guerrero , J. L. (1971). Automobile Self-Congruity Models
Reexamined. Journal of Marketing Research, 9, 125-127.
Huijgen, M. & Verburg, M. (1996). Van Dale Basis Woordenboek van de Nederlandse taal.
Utrecht: Van Dale Lexicografie.
Inman, J. J. & Zeelenberg, M. (2002). Regret in Repeat Purchase versus Switching
Decisions: The Attenuating Role of Decision Justifiability. Journal of Consumer
Research, 29, 116-128.
Janlert, L. E. & Stolterman, E. (1997). The Character of Things. Design Studies, 18, 297314.
Jones. E. E. (1990). Interpersonal Perception. New York: W.H. Freeman and Company.
Jordan, P. W. (1997). Products as Personalities. In Robertson, S. A. (Ed.), Contemporary
Ergonomics (pp. 73-78). London: Taylor & Francis.
Jordan, P. W. (2002). The Personalities of Products. In Green, W. S. & Jordan, P. W.
(Eds.), Pleasure With Products: Beyond Usability (pp. 49-60). London: Taylor &
Francis.
Jöreskog, K. & Sörbom, D. (1993). LISREL 8 User's Reference Guide. Chicago:
Scientific Software International.
Kassarjian, H. H. (1971). Personality and Consumer Behavior. Journal of Marketing Research,
8, 409-418.
Kenny, D. A., Albright, L., Malloy, T. E. & Kashy, D. A. (1994). Consensus in
Interpersonal Perception: Acquaintance and the Big Five. Psychological Bulletin,
116, 245-258.
176
References
Kleine, R. E., Kleine, S. & Kernan, J. B. (1993). Mundane Consumption and The Self: A
Social-Identity Perspective. Journal of Consumer Psychology, 2, 209-235.
Kleine, S., Kleine R. E. & Allen C. T. (1995). How Is a Possession ‘Me’ or ‘Not Me’?
Characterizing Types and an Antecedent of Material Possession Attachment.
Journal of Consumer Research, 22, 327-343.
Kotler, P. (1997). Marketing Management: Analysis, Planning, Implementation and Control. Upper
Saddle River, New Jersey: Prentice-Hall.
Kotler, P. G. & Rath, A. (1983). Design; A Powerful but Neglected Marketing Tool. Die
Unternehmung, 37, 203-221.
LaBarbera, P. A. & Mazursky, D. (1983). A Longitudinal Assessment of Consumer
Satisfaction/Dissatisfaction: The Dynamic Aspect of the Cognitive Process.
Journal of Marketing Research, 20, 393-404.
Lakoff, G. & Johnson, M. (1980). Metaphors We Live By. Chicago: The University of
Chicago Press.
Landon, E. L. (1974). Self-Concept, Ideal Self-Concept, and Consumer Purchase
Intentions. Journal of Consumer Research, 1, 44-51.
Langer, E. & Abelson, R. P. (1974). A Patient by Any Other Name: Clinical Group
Differences in Labeling Bias. Journal of Consulting and Clinical Psychology, 42, 4-9.
Laurel, B. (1991). Interface Agents: Metaphors with Character. In Laurel, B. (Ed.), The
Art of Human-Computer Interface Design (pp. 355-365). Reading, Massachusetts:
Addison-Wesley.
Leibenstein, H. (1950). Bandwagon, Snob, and Veblen Effects in the Theory of
Consumer’s Demand. Quarterly Journal of Economics, 64, 183-207.
Levy, S. J. (1959). Symbols for Sale. Harvard Business Review, 37, 117-124.
Macrae, C. N., Stangor, C. & Hewstone, M. (1996). Stereotypes and Stereotyping. New York:
The Guilford Press.
Malhotra, N. K. (1981). A Scale to Measure Self-Concepts, Person Concepts, and Product
Concepts. Journal of Marketing Research, 18, 456-464.
Malhotra, N. K. (1988). Self-Concept and Product Choice: An Integrated Perspective.
Journal of Economic Psychology, 9, 1-28.
Maslov, A. H. (1943). A Theory of Human Motivation. Psychological Review, 50, 370-396.
Maxwell, S. E. & Delaney, H. D. (1990). Designing Experiments and Analyzing Data.
Belmont, CA: Wadsworth.
McCracken, G. (1986). Culture and Consumption: A Theoretical Account of the
Structure and Movement of Cultural Meaning of Consumer Goods. Journal of
References
177
McCrae, R. R. & Costa, P. T. (1996). Toward a new generation of personality theories:
Theoretical contexts for the five-factor model. In Wiggins, J. S. (Ed.), The FiveFactor Model of Personality: Theoretical Perspectives. New York: Guilford Press.
McCrae, R. R., Costa, P. T. & Piedmont, R. L. (1993). Folk Concepts, Natural Language,
and Psychological Constructs: The California Psychological Inventory and The
Five-Factor Model. Journal of Personality, 61, 1-26.
McCrae, R. R. & John O. P. (1992). An Introduction to the Five-Factor Model and Its
Applications. Journal of Personality, 60, 175-215.
Mick, D. G. (1986). Consumer Research and Semiotics: Exploring the Morphology of
Signs, Symbols and Significance. Journal of Consumer Research, 13, 196-213.
Mischel, W. (1968). Personality and Assessment. New York: Wiley.
Mischel, W. (1979). On the interface of Cognition and Personality: Beyond the PersonSituation Debate. American Psychologist, 34, 740-754.
Montepare, J. M. & Zebrowitz-McArthur, L. (1988). Impressions of people created by
age-related qualities of their gaits. Journal of Personality and Social Psychology, 55, 547556.
Mullen, M. R. (1995). Diagnosing Measurement Equivalence in Cross-National Research.
Journal of International Business Studies, 26, 573-596.
Muller, W. (1997). Vormgeven, Ordening en Betekenisgeving. Utrecht: Lemma.
Murphy, K. R. & Davidshofer, C. O. (1994). Psychological Testing. New Jersey: Prentice
Hall.
Nass, C., Moon, Y., Fogg, B. J., Reeves, B. & Dryer, D. C. (1995). Can Computer
Personalities Be Human Personalities. International Journal of Human-Computer
Studies, 43, 223-239.
Newcomb, T. M. (1961). The Acquaintance Process. New York: Holt, Rinehart and Winston.
Nias, D. K. B. (1979). Marital choice: matching or completion? In Cook, M. & Wilson, G.
(Eds.), Love and Attraction (pp. 151-155). Oxford: Pergamon Press.
Norman, W. T. (1963). Toward an Adequate Taxonomy of Personality Attributes:
Replicated Factor Structure in Peer Nomination Personality Ratings. Journal of
Abnormal and Social Psychology, 66, 574-583.
Ogden, C. K. & Richards, I. A. (1923). The Meaning of Meaning. London: Routledge &
Kegan Paul.
Ogilvie, D. M. (1987). The Undesired Self: A Neglected Variable in Personality Research.
Oxford Advanced Learner's Dictonary. (2000). Oxford: Oxford University Press.
Park, B. & Judd, C. M. (1989). Agreement on Initial Impressions: Differences Due to
Perceivers, Trait Dimensions and Target Behaviors. Journal of Personality and Social
178
References
Pervin, L. A. & John, O. P. (1999). Handbook of personality: theory and research. New York:
The Guilford Press.
Pervin, L. A. (1975). Personality: Theory, Assessment and Research. New York: John Wiley &
Sons.
Pervin, L. A. (1985). Personality: Current Controversies, Issues, and Directions. Annual
Review of Psychology, 36, 83-114.
Peter, J. P., Churchill, G. A. & Brown, T. J. (1993). Caution in the Use of Difference
Scores in Consumer Research. Journal of Consumer Research, 19, 655-662.
Prentice, D. A. (1987). Psychological Correspondence of Possessions, Attitudes and
Values. Journal of Personality and Social Psychology, 53, 993-1003.
Punj, G. & Stewart, W. (1983). Cluster Analysis in Marketing Research: Review and
Suggestions for Application. Journal of Marketing Research, 20, 134-148.
Richins, M. L. & Bloch P. H. (1991). Post-Purchase Product Satisfaction: Incorporating
the Effects of Involvement and Time. Journal of Business Research, 23, 145-158.
Rijsdijk, S. A. & Hultink, E. J. (2002). Product Smartness: Scale Development and
Validation. In Farhangmehr, M. (Ed.), Marketing in a Changing World, Proceedings of
the 31st EMAC Conference. Braga, Portugal: University of Minho - School of
Economics and Management.
Roozenburg, N. F. M. & Eekels, J. (1998). Product Ontwerpen, Structuur en Methoden. Utrecht:
Uitgeverij Lemma.
Rosenberg, M. (1979). Conceiving the Self. New York: Basic Books, Inc.
Rosenberg, S. & Sedlak, A. (1972). Structural Representatives of Perceived Personality
Trait Relationships. In Romney, A. K., Shepard, R. N. & Nerlove, S. B. (Eds.),
Multidimensional scaling: Theory and Applications in the Behavioral Sciences (pp. 134-162).
New York: Seminar Press.
Schifferstein, H. N. J. & Pilgrim, E. (2003). Consumer-Product Attachment: The
Construct and its Measurement. Manuscript Submitted for Publication.
Schmitt, M. & Borkenau, P. (1992). The Consistency of Personality. In Caprara, G-V. &
Van Heck, G. (Eds.), Modern Psychology, Critical Reviews and New Directions.
Hempstead: Harverster Wheatsheaf.
Schultz, S. E., Kleine, R. E. & Kernan, J. B. (1989). These Are A Few of My Favorite
Things’ Toward an Explication of Attachment as a Consumer Behavior
Construct. Advances in Consumer Research, 16, 359-366.
Shocker, A. D. & Srinivasan, V. (1974). A Consumer-Based Methodology for the
Identification of New Product Ideas. Management Science, 20, 921-937.
Shoda, Y. & Mischel, W. (2000). Reconciling Contextualism with the Core Assumptions
of Personality Psychology. European Journal of Personality, 14, 407-428.
References
179
Singh, R. & Ho, S. Y. (2000). Attitudes and Attraction: A New Test of The Attraction
Repulsion and Similarity-Dissimilarity Asymmetry Hypotheses. British Journal of
Sirgy, M. J. (1982). Self-Concept in Consumer Behavior: A Critical Review. Journal of
Sirgy, M. J. (1985). Using Self-Congruity and Ideal Congruity to Predict Purchase
Motivation. Journal of Business Research, 13, 195-206.
Sirgy, M. J., Grewal, D. & Mangleburg, T. (2000). Retail Environment, Self-Congruity,
and Retail Patronage: An Inegrative Model and a Research Agenda. Journal of
Business Research, 49, 127-138.
Sirgy, M. J., Grewal, D., Mangleburg, T. F., Park, J., Chon, K. S., Claiborne, C. B., Johar,
J. S. & Berkman, H. (1997). Assessing the Predictive Validity of Two Methods of
Measuring Self-Image Congruence. Journal of the Academy of Marketing Science, 25,
229-241.
Sivadas, E. & Venkatesh R. (1995). An Examination of Individual and Object-Specific
Influences on the Extended Self and its Relation to Attachment and Satisfaction.
Advances in Consumer Research, 22, 406-412.
Smets, G., Overbeeke, K. & Gaver, W. (1994). Form-Giving: Expressing the
Nonobvious. Human Factors in Computing Systems, 79-84.
Smit, E. & Van den Berg, M. (2002). Een Merk in Mensentermen. Tijdschrift voor
Marketing, juni, 58-60.
Solomon, M. R. (1983). The Role of Products as Social Stimuli: A Symbolic
Interactionism Perspective. Journal of Consumer Research, 10, 319-329.
Solomon, M. R. (1999). Consumer Behavior; Buying, Having and Being. New Jersey: Prentice
Hall.
Spector, P. E. (1992). Summated Rating Scale Construction. Newbury Park: Sage.
Stapel, D. A. & Koomen, W. (1998). When Stereotype Activation Results in
(counter)Stereotypical Judgments: Priming Stereotype-Relevant Traits and
Exemplars. Journal of Experimental Social Psychology, 34, 136-163.
Stewart, R. A., Powell, G. E. & Chetwynd, S. J. (1979). Person Perception and Stereotyping.
Westmead: Saxon House.
Swann, W. B., Stein-Seroussi, A. & Giesler, B. (1992). Why People Self-Verify. Journal of
Personality and Social Psychology, 62, 392-401.
Szalay, L. B. & Deese, J. (1978). Subjective Meaning andCulture: An Assessment through Word
Association. Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Tan, A. (2002). Gebruikershandleiding PEL-PANEL. Delft: Delft University of Technology.
180
References
Tepper Tian, K., Bearden, W. O. & Hunter, G. L. (2001). Consumers Need for
Uniqueness: Scale Development and Validation. Journal of Consumer Research, 28,
50-66.
Tharp, R. G. (1963). Psychological Patterning in Marriage. Psychological Bulletin, 60, 97-117.
Thompson, C. J. & Haytko, D. L. (1997). Speaking of Fashion: Consumer Uses of
Fashion Discourses and the Appropriation of Countervailing Cultural Meanings.
Journal of Consumer Research, 24, 15-42.
Trueman, M. & Jobber, D. (1998). Competing through Design. Long Range Planning, 31,
594-605.
Tsiros, M. & Mittal, V. (2000). Regret: A Model of Its Antecedents and Consequences in
Consumer Decision Making. Journal of Consumer Research, 26, 401-417.
Van Buiten, L. & Jans, J-P. (2002). Levi’s Innovatie strategie werpt vruchten af, “De
engineered Jeans zijn heel belangrijk geweest”. Marketing Tribune, 20, 50-52.
Van Dale, Groot Woordenboek der Nederlandse Taal. (1999). Utrecht: Van Dale Lexicografie.
Van de Ven, L. (2003). De Opkomst van het Beeld in Onderzoek. Clou, november, 49.
Van den Berg, G. M. (1987). Homals voor beginners. Leiden: Rijksuniversiteit Leiden.
Van Raaij, W. F., Antonides, G., Oppedijk van Veen, W. M. & Schoormans, J. P. L.
(1999). Product en Consument. Utrecht: Lemma.
Veblen, T. (1899). The Theory of the Leisure Class. New York: The New American Library.
Veryzer, R. W. (1995). The Place of Product Design and Aesthetics in Consumer
Research. Advances in Consumer Research, 22, 641-645.
Vink, N. Y. (2003). Mass Custimazation. Delft: Delft University of Thechnology.
Vonk, R. (1990). The Cognitive Representation of Persons. Leiden: Leiden University.
Vonk, R. (1999). Cognitieve Sociale Psychologie. Utrecht: Uitgeverij Lemma BV.
Vonk, R. & Heiser, W. J. (1991). Implicit Personality Theory and Social Judgment:
Effects of Familiarity with a Target Person. Multivariate Behavioral Research, 26, 6981.
Wall, T. D. & Payne, R. (1973). Are Deficiency Scores Deficient? Journal of Applied
Wallendorf, M. & Arnould, E. J. (1988). My Favorite Things: A Cross-Cultural Inquiry
into Object Attachment, Possessiveness, and Social Linkage. Journal of Consumer
Research, 14, 531-547.
Weightman, D. & McDonagh, D. (2003). People are Doing it for Themselves. In,
Proceedings of the 2003 International Conference on Designing Pleasurable Products and
Interfaces (pp. 34-39). Pittsburg, Pennsylvania: ACM Press.
Wells, W. D., Andruili, F. J., Goi, F. J. & Seader, F. (1957). An Adjective Check List for
the Study of “Product Personality”. Journal of Applied Psychology, 41, 317-319.
References
181
White, J. K., Salter, A. J. & Gann, D. M. (2003). Designing to Compete: Lessons from
Millenium Product Winners. Design Studies, 24, 395-409.
Wicker, A. W. (1969). Attitudes versus Actions. Journal of Social Issues, 25, 41-78.
Wong, N., Rindfleisch, A. & Burroughs, J. E. (2003). Do Reverse-Worded Items
Confound Measures in Cross-Cultural Consumer Research? The Case of the
Material Values Scale. Journal of Consumer Research, 30, 72-90.
Zaltman, G. (2003). How Customers Think; Essential Insights into the Mind of the Market.
Boston: Harvard Business School Press.
Zebrowitz, L. A. (1990). Social Perception. Pacific Grove, CA: Brooks Cole.
182
183
Appendix A
Color pictures of the stimuli used in the study reported in section 3.2
184
Appendix B
Overview of the verbal description categories used in the study reported in section 3.2
Category
A A-symmetrisch
B Bijzonder/apart *
Blits
C Chic
D Design
Digitaal
Donker
Duidelijk
Duur
E Eenvoudig *
Excentriek *
F Feestelijk
Futuristisch
G Gebruiksvoorwerp
Gematigd positief
Gewoon
Gezellig *
Goed afgewerkt
Goedkoop
Goud
Groot/fors
H Handig *
J Je moet durven
Jong
Jongensachtig *
K Kinderlijk *
Klassiek
Kleurig
L Lelijk
Leuk
M Mannelijk *
Meisjesachtig *
Metaal
Militaristisch
Modern
N Nep
Netjes *
Niet druk
Amount of verbal
descriptions
3
35
7
19
5
8
1
12
18
33
10
1
24
1
12
34
1
1
12
2
31
12
1
14
22
6
15
3
27
49
13
35
1
1
15
1
9
1
% of verbal descriptions
0.4
4.6
0.9
2.5
0.7
1.0
0.1
1.6
2.3
4.3
1.3
0.1
3.1
0.1
1.6
4.4
0.1
0.1
1.6
0.3
4.0
1.6
0.1
1.8
2.9
0.9
2.0
0.4
3.5
6.4
1.7
4.6
0.1
0.1
2.0
0.1
1.2
0.1
Appendix B
Niet echt grof
Niet modern
Niet netjes *
Niet ouderwets
Niet sportief *
Niet stijlvol
O Onduidelijk
Onhandig *
Opvallend *
Ouder
Ouderwets *
Over nagedacht
P Pronken/patserig
R Rechthoekig
Rond
S Saai *
Sieraad
Speels
Sportief *
Stevig
Stoer *
Strak
T ‘Te veel'
Tijdloos
Trendy
U Unisex
V Verfijnd
Vrouwelijk *
W Wild
Z Zwitsers
1
2
4
1
2
1
15
8
15
2
8
2
20
7
2
2
2
4
48
21
8
2
32
2
14
7
7
8
1
1
0.1
0.3
0.5
0.1
0.3
0.1
2.0
1.0
2.0
0.3
1.0
0.3
2.6
0.9
0.3
0.3
0.3
0.3
6.3
2.7
1.0
0.3
4.2
0.3
1.8
0.9
0.9
1.0
0.1
0.1
* Categories considered to be personality characteristics by more than half of the five
independent judges.
185
186
Appendix C
The 18 irons designed to be happy, cute or tough (study reported in section 3.3).
Happy
Cute
Tough
187
Appendix D
The scenarios used in the studies reported in section 4.3 in English and Dutch
Scenario conscientious person
Susan is 27 years old and married to Stephan. She works as a lawyer for a large law firm.
For her job she always has to look representative and she loves that; a suit is one of
favorite outfits. Her job is very demanding, but she is very ambitious, so she does it with
pleasure. Her goal is to be at the top of the lawyer’s world in five years, and she works
hard to achieve that goal. In her spare time, she likes to cook and read a good book. She
also likes to go on holiday; she then maps out the route in advance and makes lists of all
the things she should take along. Her house always looks very neat and tidy; everything
has its own place so she can quickly find things. Stephan always calls her a fusspot,
because she is always very punctual and scrupulous. She personally thinks that it is not so
bad, she just likes to dot the i’s and cross the t’s. Her friends would characterize her as a
real go-getter, someone who they can rely on and who always keeps her promises. They
often ask her for help with difficult decisions, because she is good in weighing the pros
and cons against each other.
Susan is 27 en getrouwd met Stephan. Ze werkt als advocate bij een groot advocatenkantoor. Voor haar
werk moet ze er altijd representatief uitzien en dat vindt ze heerlijk, een mantelpakje is haar favoriete
outfit. Het werk vraagt veel van haar, maar ze is erg ambitieus dus ze doet het met plezier. Haar doel is
om over 5 jaar tot de top van de advocatenwereld te behoren, en ze werkt er hard aan om dat doel te
bereiken. In haar vrije tijd houdt ze van lekker koken en een goed boek lezen. Ook gaat ze graag op
vakantie, ze stippelt dan van tevoren de route uit en maakt lijstjes van alle dingen die ze mee moet nemen.
Haar huis ziet er altijd keurig en opgeruimd uit, alles heeft zijn eigen plek zodat ze dingen snel kan
vinden. Stephan noemt haar altijd Suusje Precies, omdat ze erg stipt en nauwgezet is. Zelf vindt ze dat
wel meevallen, ze houdt er gewoon van om de puntjes op de i te zetten. Haar vrienden zouden haar
typeren als een echte doorzetter, iemand waar ze op kunnen bouwen en die altijd haar beloftes nakomt.
Ze vragen vaak haar hulp bij moeilijke beslissingen, zij kan namelijk goed alle voors en tegens tegen
elkaar afwegen.
188
Appendix D
Scenario extravert person
Susan is 27 years old and married to Stephan. She works as a pr-employee for a large
media-concern. For her job she always has to look representative, but actually she thinks
a suit is far too decent. She prefers wearing clothes in which she catches the eye,
especially when she goes out. She must not think of going through life as a dull person.
When she enters a room, she will not remain unnoticed; she herself will take care of that
with her enormous enthusiasm and liveliness. According to Stephan, the sun starts
shining when she walks in, but some people can irritate themselves of her exaggerated
presence. However, for her hobby her need for attention is very convenient: she is a
singer in a band. She also spends a lot of time on her social life: she has many friends and
likes to be in the company of people. “The more, the merrier” is one of her mottos. Her
friends would describe her as a special young woman who is always in for everything and
who does not mince her words. Her ideal holiday is an active one, but it should be in a
crowded area; it is nothing like her to stay in a remote cabin in the woods.
Susan is 27 jaar en getrouwd met Stephan. Ze werkt als pr-medewerkster bij een groot media-concern.
Voor haar werk moet ze er altijd representatief uit zien, maar eigenlijk vindt ze een mantelpakje veel te
braaf. Ze draagt liever kleding waarmee ze in het oog springt, vooral als ze uitgaat. Ze moet er niet aan
denken om als een grijze muis door het leven te gaan. Als ze een kamer binnenkomt, zal ze niet
onopgemerkt blijven, daar zorgt zijzelf met haar enorme enthousiasme en levendigheid wel voor. Volgens
Stephan gaat het zonnetje schijnen als zij binnenkomt, maar sommige mensen kunnen zich irriteren aan
haar overdreven aanwezigheid. In haar hobby komt haar behoefte om in de belangstelling te staan haar
juist wel van pas: ze is zangeres in een band. Ze besteedt ook veel tijd aan haar sociale leven: ze heeft veel
vrienden en vindt het fijn om in gezelschap te zijn. “Hoe meer zielen, hoe meer vreugd” is één van haar
motto’s. Haar vrienden zouden haar omschrijven als een bijzondere meid, die altijd overal voor in is en
die geen blad voor haar mond neemt. Haar ideale vakantie is een actieve doe-vakantie, maar wel in een
druk gebied, het is niets voor haar om in een afgelegen hut in het bos te zitten.
189
Appendix E
Wells, Andruili, Goi & Seader (1957)
# items = 108
Friendly
Strong
Angry
Fat
Popular
Slow
Good
Married
Strange
Patient
Modern
Secure
Fair
Vain
Tender
Comfortable
Particular
Merry
Small
Honest
Poor
Natural
Sharp
Serious
Masculine
Different
Gay
Big
Silent
Gentle
Young
Set
Cold
Superior
Firm
Average
Clean
Rough
Simple
Travelled
Dangerous
Wise
Hard
Understanding
Content
Old-fashioned
Successful
Cheap
Soft
Middle-class
Able
Warm
Democratic
Thin
Good-looking
Common
Quiet
Fancy
Old
Feminine
Practical
Moral
Cross
Kind
Difficult
Thinking
Powerful
Careful
Low-class
Tired
Important
Foreign
Interesting
Little
Brave
Rich
Plain
Bright
Weak
Loud
Busy
Nice
Original
Happy
Heavy
Smart
Active
Correct
Calm
Curious
Proud
Bitter
Cool
Single
Pleasant
Bad
Steady
Famous
Religious
Funny
Wonderful
Fine
Independent
Smooth
Ordinary
Tall
High-class
Sad
190
Appendix E
Malhotra (1981)
# items = 30
Jordan (2000)
# items = 34
Rugged – Delicate
Excitable – Calm
Uncomfortable – Comfortable
Dominating – Submissive
Thrifty – Indulgent
Pleasant – Unpleasant
Contemporary – Non contemporary
Organized – Unorganised
Rational – Emotional
Youthful – Mature
Formal – Informal
Orthodox – Liberal
Complex – Simple
Colourless – Colourful
Modest – Vain
Kind – Unkind
Honest – Dishonest
Serious minded – Light hearted
Bright – Dim
Stable – Unstable
Narcissistic – Humble
Flexible – Inflexible
Authoritarian – Liberal
Value driven – Non-value driven
Extrovert – Introvert
Naïve – Cynical
Excessive – Moderate
Conformist – Rebel
Energetic – Unenergetic
Violent – Gentle
Complex – Simple
Pessimistic – Optimistic
Aaker (1997)
# items = 42
Down-to-earth
Family-oriented
Small-town
Honest
Sincere
Real
Wholesome
Original
Cheerful
Sentimental
Friendly
Daring
Trendy
Exciting
Spirited
Cool
Young
Imaginative
Unique
Up-to-date
Independent
Contemporary
Reliable
Hard-working
Secure
Intelligent
Technical
Corporate
Successful
Leader
Confident
Upper-class
Glamorous
Good-looking
Charming
Feminine
Smooth
Outdoorsy
Masculine
Western
Tough
Rugged
Appendix E
191
Cattell, Eber & Tatsuoka (1977)
# items = 32
Reserved – Warm
Concrete-reasoning – Abstract-reasoning
Reactive – Emotionally stable
Deferential – Dominant
Serious – Lively
Expedient – Rule-conscientious
Shy – Socially bold
Utilitarian – Sensitive
Trusting – Vigilant
Practical – Imaginative
Forthright – Private
Self-assured – Apprehensive
Traditional – Open to change
Group-oriented – Self reliant
Tolerates disorder – Perfectionist
Relaxed – Tense
Eysenck (1975)
# items = 32
Passive
Careful
Thoughtful
Peaceful
Controlled
Reliable
Even-tempered
Calm
Quiet
Pessimistic
Unsociable
Sober
Rigid
Moody
Anxious
Reserved
Sociable
Outgoing
Talkative
Responsive
Easygoing
Lively
Carefree
Leaderly
Active
Optimistic
Impulsive
Changeable
Excitable
Aggressive
Restless
Touchy
Bipolar and unipolar items representing the Five Factor Model (adopted
from: Carver & Scheier, 1996)
# items = 58 (bipolair) + 29 (unipolair)
Bold – Timid
Forceful – Submissive
Self-confident – Unassured
Talkative - Silent
Spontaneous – Inhibited
Gregarious
Outspoken
Energetic
Happy
Seclusive
192
Appendix E
Friendly – Unfriendly
Warm – Cold
Kind – Unkind
Polite – Rude
Good-natured – Irritable
Jealous
Considerate
Spiteful
Touchy
Complaining
Cautious – Rash
Serious – Frivolous
Responsible – Irresponsible
Thorough – Careless
Hardworking – Lazy
Neat
Persevering
Planful
Careful
Eccentric
Nervous – Poised
Anxious – Calm
Excitable – Composed
Relaxed – High-strung
Concerned
Nervous
Fearful
Tense
Imaginative – Simple
Intellectual – Unreflective
Polished – Crude
Uncurious – Curious
Uncreative – Creative
Knowledgeable
Perceptive
Imaginative
Verbal
Original
Norman (1963)
# items = 40
Asch (1946)
# items = 36
Talkative – Silent
Frank – Secretive
Adventurous – Cautious
Sociable – Reclusive
Good-natured – Irritable
Not jealous – Jealous
Gentle – Headstrong
Cooperative – Negativistic
Fussy – Careless
Responsible – Undependable
Exacting – Unexacting
Persevering – Quitting
Poised – Nervous
Calm – Anxious
Composed – Excitable
Not hypochondriacal – Hypochondrical
Artistically sensitive – Insensitive
Intellectual – Narrow
Polished – Crude
Imaginative – simple
Generous – Ungenerous
Shrewd – Wise
Unhappy – Happy
Irritable – Good-natured
Humorous – Humourless
Sociable – Unsociable
Popular – Unpopular
Unreliable – Reliable
Important – Insignificant
Ruthless – Humane
Good-looking – Unattractive
Persistent – Unstable
Frivolous – Serious
Restrained – Talkative
Self-centred – Altruistic
Imaginative – Hard-headed
Strong – Weak
Dishonest – Honest
Appendix E
193
Anderson & Klatzky (1987)
# items = 32
Ambitious
Self-assured
Diplomatic
Boisterous
Charismatic
Extraverted
Entertaining
Exuberant
Lively
Spirited
Silly
Argumentative
Manipulative
Domineering
Bullying
Harassing
Harsh
Tough
Glib
Brainy
Studious
Intelligent
Contemplative
Meditative
Introspective
Inhibited
Shy
Withdrawn
Introverted
Lonesome
Melancholy
Self-conscious
Rosenberg en Sedlak, 1972 (adopted from: Zebrowitz, 1990)
# items = 108
Intelligent
Friendly
Hard worker
Helpful
Honest
Kind
Self-centred
Sense of humour
Sincere
Generous
Sensitive
Conceited
Ambitious
Stubborn
Understanding
Athletic
Outgoing
Considerate
Quiet
Shy
Easy-going
Selfish
Conservative
Thoughtful
Emotional
Interesting
Moody
Religious
Talkative
Determined
Humorous
Immature
Smart
Lazy
Egotistical
Liberal
Witty
Proud
Domineering
Easy to get along with
Idealistic
Nervous
Sarcastic
Trustworthy
Dedicated
Extrovert
Happy
Liar
Warm
Bigot
Brilliant
Confident
Conscientious
Fun-loving
Independent
Individualistic
Loving
Loyal
Insecure
Active
Aggressive
Concerned
Inconsiderate
Dependable
Snobbish
Narrow-minded
Open-minded
Outspoken
Pleasant
Polite
Affectionate
Cold
Considerate
Cool
Courageous
Cynical
Good-natured
Introvert
Responsible
Sweet
Arrogant
Calm
Cheerful
Demanding
Drinker
Hypocritical
Unintelligent
Naïve
Phoney
Unselfish
Sophisticated
Talented
Untrustworthy
Two-faced
Popular
Reliable
Irresponsible
Self-sacrificing
Sloppy
Sympathetic
Quick-tempered
Carefree
Concerned about
others
Forceful
Interested in others
Likeable
Nice
Optimistic
194
Appendix F
Pool A – The translated items from literature
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
Aanhalig
Aardig
Abstract redeneren
Actief
Afwijzend
Agressief
Alledaags
Alleenstaand
Altruistisch (onzelfzuchtig)
Ambitieus
Angstig
Aristocratisch
Arm
Arrogant
Atletisch
Attent
Autoritair
Avontuurlijk
Bazig
Begripvol
Beheerst
Behulpzaam
Bekrompen
Bekwaam
Belangrijk
Belangstellend in anderen
Beleefd
Bereisd
Beschaafd
Bescheiden
Beschouwelijk
Beschouwend
Betrokken
Betrouwbaar
Bezorgd
Bezorgdheid
Bijdehand
Bitter
Blij
Boos
Briljant
Beroemd
Buitenlands
Buitenmens
Buitensporig
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
Bullebak
Burgerlijk
Chaotisch
Charismatisch
Charmant
Comfortabel
Complex
Concreet redeneren
Conformist
Conservatief
Cool
Correct
Creatief
Cynisch
Democratisch
Dik
Diplomatiek
Direct
Dom
Dominant
Dorps
Drinker
Droevig
Druk
Dun
Dwaas
Echt
Eenvoudig
Eenzaam
Eerbiedig
Eerlijk
Eersteklas
Egocentrisch
Egoïstisch
Eigenaardig
Elegant
Emotioneel
Emotioneel stabiel
Energiek
Ethisch
Extravert
Familie georiënteerd
Fantasieloos
Fantasierijk
Fijn
Appendix F
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
Flexibel
Formeel
Frivool
Geduldig
Gedurfd
Geestig
Geheimzinnig
Gelijkmoedig
Gelukkig
Gemakzuchtig
Gematigd
Gemiddeld
Genereus
Georganiseerd
Geremd
Gereserveerd
Geschikt
Gesloten
Gespannen
Getalenteerd
Getrouwd
Gevaarlijk
Gevoel voor humor
Gevoelig
Gewaagd/Gedurfd
Gewelddadig
Geweldig
Gewetensvol
Gewoontjes
Gezellig
Glad
Glamour
Goed
Goedkoop
Grappig
Groepsgeoriënteerd
Grof
Grondig
Groot
Hard werkend
Harde werker
Hardvochtig
Hatelijk
Humeurig
Humoristisch
Humorloos
Hypochondrisch
Hypocriet
Idealistisch
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
188.
Iemand die snel opgeeft
IJdel
Impopulair
Impulsief
Individueel
Informeel
Intellectueel
Intelligent
Interessant
Intolerant
Introspectief
Introvert
Jaloers
Jeugdig
Jong
Kalm
Kieskeurig
Klagend
Klein
Kleurloos
Kleurrijk
Kluizenaar
Knap
Koel
Koppig
Koud
Krachtig
Kuddedier
Kunstzinnig
Lang
Langzaam
Leergierig
Leidend
Leider
Leugenaar
Levendig
Lichtgeraakt
Lichthartig
Liefhebbend
Loyaal
Lui
Luidruchtig
Lusteloos
Machtig
Makkelijk
Makkelijk om mee om te gaan
Manipulatief
Mannelijk
Meedogenloos
195
196
189.
190.
191.
192.
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
217.
218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
228.
229.
230.
231.
232.
233.
234.
235.
236.
237.
Appendix F
Meewerkend
Menselijk
Modern
Moe
Moedig
Moeilijk
Nadenkend
Naïef
Nalatig
Narcistisch
Natuurlijk
Nauwgezet
Nep
Nerveus
Netjes
Niet genereus
Niet hedendaags
Niet jaloers
Niet nieuwsgierig
Niet zelfverzekerd
Nieuwsgierig
Nijdig
Nuchter
Onaantrekkelijk
Onafhankelijk
Onattent
Onbaatzuchtig
Onbeduidend
Onbekommerd
Onbeleefd
Onbetrouwbaar
Onbezonnen
Onbuigzaam
Onderdanig
Onderhoudend
Oneerlijk
Onevenwichtig
Ongelukkig
Ongemakkelijk
Ongerust
Ongevoelig
Onplezierig
Onstuimig
Ontvankelijk
Onveranderlijk
Onverantwoordelijk
Onvolwassen
Onvriendelijk
Onzeker
238.
239.
240.
241.
242.
243.
244.
245.
246.
247.
248.
249.
250.
251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
267.
268.
269.
270.
271.
272.
273.
274.
275.
276.
277.
278.
279.
280.
281.
282.
283.
284.
285.
286.
Open
Open voor verandering
Opmerkzaam
Oprecht
Optimistisch
Opwindend
Origineel
Orthodox
Oud
Ouderwets
Overgevoelig
Passief
Perfectionist
Pessimistisch
Pietluttig
Plezierig
Populair
Praatgraag
Praktisch
Praktisch aangelegd
Pret maker
Prikkelbaar
Rationeel
Reagerend
Rebel
Relaxed
Religieus
Rijk
Ruig
Ruimdenkend
Rusteloos
Ruw
Sarcastisch
Schattig
Scherpzinnig
Schoon
Schuchter
Sentimenteel
Serieus
Simpel
Slecht
Slim
Slordig
Snobistisch
Soft
Solitair
Spontaan
Stabiel
Standvastig
Appendix F
287.
288.
289.
290.
291.
292.
293.
294.
295.
296.
297.
298.
299.
300.
301.
302.
303.
304.
305.
306.
307.
308.
309.
310.
311.
312.
313.
314.
315.
316.
317.
Statig
Sterk
Stil
Studiebol
Succesvol
Superieur
Sympathiek
Taai
Technisch
Teder
Teruggetrokken
Terughoudend
Tevreden
Tijdelijk
Toegeeflijk
Toegewijd
Toestaan van wanorde
Traditioneel
Treiterend
Trendy
Trots
Twistziek
Uitbundig
Uitgesproken
Uniek
Utilitarist
Vaag
Vastberaden
Vasthoudend
Veeleisend
Veranderbaar
318.
319.
320.
321.
322.
323.
324.
325.
326.
327.
328.
329.
330.
331.
332.
333.
334.
335.
336.
337.
338.
339.
340.
341.
342.
343.
344.
345.
346.
347.
348.
Verantwoordelijk
Verbaal
Verlegen
Verschillend
Vertrouwend
Verwaand
Vol zelfvertrouwen
Volhardend
Volwassen
Voorzichtig
Vreedzaam
Vriendelijk
Vrolijk
Vrouwelijk
Waakzaam
Warm
Wereldwijs
Wijs
Wreed
Yup
Zachtaardig
Zeker
Zelfbewust
Zelfopofferend
Zelfverzekerd
Zonderling
Zorgzaam
Zuinig
Zwaar
Zwaarmoedig
Zwak
197
198
Appendix G
Pool B – The items resulting from qualitative research
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
1-eiige tweeling
2-eïge tweeling
Aaibaar
Aandachttrekker
Aangenaam
Aanpassingsvermogen
Aansprekend iemand
Aantrekkelijk
Aanwezig
Aardig
Absurde humor
Achtergebleven
Afgepast
Afgeraffeld
Afgesloten
Afhankelijk
Afschuwelijk
Afstandelijk
Afstotend
Afwijkend
Agressief
Algemeen, nooit iets persoonlijks
Alledaags
Alles onder controle hebben
Alles op een rijtje hebben
Altijd opgewekt
Ambitieus
Amfoor
Anders
Apart
Arbeider
Arrogant
Artistiek
'Arty'
Authentiek
Banaal
Bang
Bang voor oordeel van vrienden
Bedeesd
Bedreigend
Behoudend
Behulpzaam
Bekend
Bekrompen
Belachelijk
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
Belangrijk
Beleefd
Beperkt
Berouwbaar
Bescheiden
Beschermend
Beste met iedereen voor
Betrouwbaar
Bewerkelijk
Bewust
Bijdehand
Bijzonder
Bizar
Blaaskaak
Blij
Blijft netjes
Bochel
Bochtig
Boeiend
Bollig
Bont
Bourgondisch
Breed
Breekbaar
Broos
Bruikbaar
Bruisend
buitenbeentje
Buitenlands
Buitenmens
Burgerlijk
Burgerlijk
Burgertrutje
Carrière
Charmant
Chic
'Classy'
Clown
Collegiaal
Commercieel
Compleet
Complex
Consequent
Conservatief
Conventioneel
Appendix G
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
Correct
Creatief
Cultureel type
Dame
Deftig
Degelijk
Denkt alles zelf te weten
Denkt overal aan
Denkt veel na voor hij iets zegt
Depressief
Design liefhebber
Deskundig
Dicht
Dienstbaar
Diep over dingen nadenkt
Diepgang
Diepgravend
Diepzinnig
Dik
Dik buikje
Dikdoenerij
Diplomatiek
Direct
Dodelijk
Doe-het-zelver
Doelgerichte indruk
Doelloos
Doelmatig
Doeltreffend
Doener
Doet meer voorkomen dan dat ie is
Doet wat ervan verwacht mag
worden
Doet wat hij moet doen
Doet zich anders voor dan hij is
Doet zich meer voor dan hij is
Doet zich niet anders voor
Doet zijn best prettig over te komen
Dom
Dominant
Donker
Doodeenvoudig
Doods
Doordacht
Doordrammer
Doorsnee
Doortastend
Doorzetter
Down-to-earth
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
Droevig
Druk
Dubbel
Duidelijk
Duidelijk aanwezig
Duidelijk manifesterend
Duidelijk wat je ermee moet
Duister
Durft risico's te nemen
Duur
Echt
Eenduidig
Eenling
Eenvoudig
Eenzaam
Eenzijdig
Eerlijk
Effectief
Eigen mening
Eigenaardig
Eigengereid
Eigenwijs
Eigenzinnig
Elegant
Elke dag willen ontmoeten
Eng
Enkele keer nuttig
Erg tevreden met zichzelf
Ernstig
Esthetisch
Evenwichtig
Excentriek
Explosief
Extravagant
Extravert
Fantasieloos
Fantasierijk
Fassade
Feestelijk
Fel
Fijn
Fijntjes
Fijnzinnig
Flets
Fleurig
Flexibel
Flink
Fors
Fraai
199
200
188.
189.
190.
191.
192.
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208.
209.
210.
211.
212.
213.
214.
215.
216.
217.
218.
219.
220.
221.
222.
223.
224.
225.
226.
227.
228.
229.
230.
231.
232.
233.
234.
235.
236.
Appendix G
Fragiel
Fris
Frivool
Fruitig
Functioneel
Gaaf
Gaat iets vanuit
Gaat met de tijd mee
Gaat niet diep
Gaat om uiterlijk i.p.v. Functioneel
Gaat voor eigen profijt
Gaat zich niet te buiten
Gastvrij
Gauw op uit gekeken
Geborgen
Gebruiksaanwijzing nodig
Gebruiksvriendelijk
Gebruikt alles
Gecamoufleerd
Gecompliceerd
Gedegen
Gedienstig
Gedistingeerd
Geduldig
Geeft geen warmte
Geeft houvast
Geeft veel tijd aan materiele dingen
Geeft zich niet bloot
Geen hoogte van te krijgen
Geen lange levensduur
Geen positief of negatief beeld
Geen prettige indruk
Geestelijk gehandicapt
Geestig
Geestrijk
Geheel
Geheimzinnig
Geinig
Geïnteresseerd
Gek
Gemaakt
Gemakkelijk
Gemakzuchtig
Gemaskerd
Gemeen
Gemiddeld qua IQ
Gemoedelijk
Gemotiveerd
Geniepig
237.
238.
239.
240.
241.
242.
243.
244.
245.
246.
247.
248.
249.
250.
251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
262.
263.
264.
265.
266.
267.
268.
269.
270.
271.
272.
273.
274.
275.
276.
277.
278.
279.
280.
281.
282.
283.
284.
285.
Genieter
Geordend
Geraffineerd
Gereserveerd
Gericht op gemak
Geschikt voor zijn werk
Geslepen
Gesloten
Gestroomlijnd
Gevaar
Gevaarlijk
Gevoel voor humor
Gevoel voor natuur
Gevoelig
Gevoelig voor groep
Gevoelvol
Gevuld aan alle kanten
Gewelddadig
Geweldig
Gewoon
Gezellig
Gezelschapsmens
Gezet
Gezicht spreekt boekdelen
Gierig
Glad
Gladjanus
Glans
Goed
Goed afgewerkt
Goed bruikbaar
Goed figuur
Goed functionerend
Goed gekleed
Goed in de hand
Goed te gebruiken
Goed te hanteren
Goed te presenteren
Goed uitzien
Goed verzorgd
Goede gastvrouw
Goede huisvader
Goede kleren
Goede luisteraar
Goede man
Goede relatie
Goede uitstraling
Goedkoop
Goedkoop
Appendix G
286.
287.
288.
289.
290.
291.
292.
293.
294.
295.
296.
297.
298.
299.
300.
301.
302.
303.
304.
305.
306.
307.
308.
309.
310.
311.
312.
313.
314.
315.
316.
317.
318.
319.
320.
321.
322.
323.
324.
325.
326.
327.
328.
329.
330.
331.
332.
333.
334.
Goedlachs
Graag in schijnwerpers
Graag op willen vallen
Gracieus
Grappig
Grappige schoonheid
Grieks
Griezelig
Grijpbaar
Grijze muis
Grof
Groot
Groot hoofd
Grootheid
Grote contrasten
Grote persoonlijkheid
Gul
Handelbaar
Handig
‘Handy’ man
Hard
Harde
Harde werker
Harmonieus
Hartelijk
Heb je niks aan
Hebbedingetje
Heeft iets van goede oude tijd
Heeft niet veel over zich
Heeft smaak
Heeft stijl/klasse
Heeft veel contacten
Heeft weinig vrienden
Heethoofd
Heft in handen neemt
Heftig
Helder
Herkenbaar
Het dringt op
Hooghartig
Hoogste woord
Houdt iets achter de hand
Houdt iets verborgen
Houdt van duurdere spulletjes
Houdt van gezelligheid
Houdt van goed gesprek
Houdt van het leven
Houdt van lekker eten
Houdt van lekker wijntje
335.
336.
337.
338.
339.
340.
341.
342.
343.
344.
345.
346.
347.
348.
349.
350.
351.
352.
353.
354.
355.
356.
357.
358.
359.
360.
361.
362.
363.
364.
365.
366.
367.
368.
369.
370.
371.
372.
373.
374.
375.
376.
377.
378.
379.
380.
381.
382.
383.
201
Houdt van reizen
Houdt van stappen
Houdt van uiterlijk vertoon
Houdt van vrolijkheid
Huiselijk
Huis-tuin-en-keuken
Humeurig
Humor
Humoristisch
Hygiënisch
Iel
Iemand die angst oproept
Iemand die doet wat hij moed doen
Iemand die geniet van het leven
Iemand die goed nadenkt
Iemand die het niet zo nauw neemt
Iemand die lekker ruikt
Iemand die maar in 1 ding goed is
Iemand die nieuws bijhoudt
Iemand die nodig is
Iemand die voor iets staat
Iemand die zijn plicht doet
Iemand met lef
Iemand van volume
Ijdel
Ijverig
Imponerend
Impulsief
In het oog springen
In verhouding
In voor grapjes
In zichzelf gekeerd
Indrukwekkend
Inefficiënt
Ingetogen
Ingewikkeld
Inhoud
Inhoudsloos
Innemend
Inspelen op trends
Integer
Intellectueel
Intelligent
Intensief hard werken
Intensief van leven genieten
Intensieve relatie
Interessant
Intiem
Introvert
202
384.
385.
386.
387.
388.
389.
390.
391.
392.
393.
394.
395.
396.
397.
398.
399.
400.
401.
402.
403.
404.
405.
406.
407.
408.
409.
410.
411.
412.
413.
414.
415.
416.
417.
418.
419.
420.
421.
422.
423.
424.
425.
426.
427.
428.
429.
430.
431.
432.
Appendix G
Inventief
In zichzelf gekeerd
Is er als je hem nodig hebt
Is het net niet
Is verbaasd als hij iets ontdekt
Italiaans
Je kunt er alle kanten mee op
Jong
Jong meisje
Jong overkomen
Jonge man
Jongensachtig
Joviaal
Kaal
Kan alle kanten op
Kan goed relativeren
Kan goed schipperen
Kan hip ding zijn
Kan je altijd een beroep op doen
Kan je mee overweg
Kan met iedereen omgaan
Kan nuttig zijn
Kan zich anders voordoen
Karakter
Karaktervol
Kernachtig
Keurig
Kil
Kinderachtig
Kinderlijk
Kitsch
Klasse
Klassiek
Klein
Kleurig
Kleurloos
Kleurrijk
Kleverig
Klungelig
Klusjesman
Klust graag
Kneuterig
Knullig
Knus
Knutselaar
Koel
Koestert
Komisch
Komt onverwacht uit de hoek
433.
434.
435.
436.
437.
438.
439.
440.
441.
442.
443.
444.
445.
446.
447.
448.
449.
450.
451.
452.
453.
454.
455.
456.
457.
458.
459.
460.
461.
462.
463.
464.
465.
466.
467.
468.
469.
470.
471.
472.
473.
474.
475.
476.
477.
478.
479.
480.
481.
Komt voor zijn mening uit
Komt vriendelijk over
Koninklijk
Koppig
Kordaat
Kort
Korte termijn persoon
Koud
Krachtig
Krakkemikkig
Kritisch
Kronkel
Kun je niet omheen
Kun je van op aan
Kunstig
Kunstmatig
Kwalitatief goed
Kwaliteit
Kwetsbaar
Laag opgeleid
Laat niet achterste van tong zien
Laat zich niet uitdagen trendy te zijn
Landelijk
Lang
Leeft alleen
Leeft intensief
Leeg
Leeghoofd
Leider
Lekker
Lelijk
Let erg op mening van anderen
Leuk
Leuke presentatie
Leve de lol
Leven is leuk
Leven versieren
Levensgenieter
Levenslustig
Licht
Licht uitstralen
Lichtgewicht
Lichtvoetig
Lief
Lief klein meisje
Ligt wel
Lijkt op schoonheidsfoutje
Log
Logisch
Appendix G
482.
483.
484.
485.
486.
487.
488.
489.
490.
491.
492.
493.
494.
495.
496.
497.
498.
499.
500.
501.
502.
503.
504.
505.
506.
507.
508.
509.
510.
511.
512.
513.
514.
515.
516.
517.
518.
519.
520.
521.
522.
523.
524.
525.
526.
527.
528.
529.
530.
Lol
Lol in het leven
Lollig
Lomp
Los
Luchtig
Maakt weinig indruk
Maatschappelijk ingesteld
Mager
Makkelijk
Makkelijk in de omgang
Makkelijk in de omgang
Makkelijk mee omgaan
Man
Man van weinig woorden
Manifesteert zich
Mannelijk
Massief
Mat
Meegaand
Meestal vervelend
Meisje
Mensen hebben hem graag
Merkwaardig
Met beide benen op grond
Met eigen stempel
Met iedereen meedoen
Met zichzelf ingenomen
Middelmatig
Middenniveau
Middle of the road
Mild
Minachtend
Misbruikt anderen voor eigen doel
Misleidend
Misplaatst
Modebewust
Modern
Modieus
Moe
Moeilijk in samenwerking
Moeilijk persoon
Moeilijk te doorgronden
Moet je voor uitkijken
Mollig
Monter
Mooi
Mooie dame
Mooie kop
531.
532.
533.
534.
535.
536.
537.
538.
539.
540.
541.
542.
543.
544.
545.
546.
547.
548.
549.
550.
551.
552.
553.
554.
555.
556.
557.
558.
559.
560.
561.
562.
563.
564.
565.
566.
567.
568.
569.
570.
571.
572.
573.
574.
575.
576.
577.
578.
579.
203
Nadrukkelijk aanwezig
Nagedacht
Naïef
Natuur
Natuurlijk
Natuurmens
Natuurvriend
Negatief
Neiging naar klassiek
Nep
Net
Net niet af
Net niet wat het in eerste instantie is
Net wat anders
Netjes
Neutraal
Niet achterste van tong laten zien
Niet altijd eerlijk
Niet altijd makkelijk in omgang
Niet bang om conflict aan gaan
Niet helemaal zoals het hoort
Niet makkelijk mee om te gaan
Niet mooi, niet lelijk
Niet om mee te spelen
Niet te grijpen
Niet te raken
Nietig
Niets bijzonders
Niets te bieden
Nietszeggend
Nieuwsgierig
No nonsense
Nodig
Nonchalant
Noodzakelijk
Nooit overdreven
Normaal postuur
Nors
Nostalgisch
Nuchter
Nuttig
Ochtendhumeur
Ochtendmens
Oersaai
Onaangenaam
Onaangepast
Onaantrekkelijk
Onaardig
Onafhankelijk
204
580.
581.
582.
583.
584.
585.
586.
587.
588.
589.
590.
591.
592.
593.
594.
595.
596.
597.
598.
599.
600.
601.
602.
603.
604.
605.
606.
607.
608.
609.
610.
611.
612.
613.
614.
615.
616.
617.
618.
619.
620.
621.
622.
623.
624.
Appendix G
Onbehaaglijk
Onbekend
Onbemind
Onbenul
Onbereikbaar
Onbetrouwbaar
Ondegelijk
Ondernemend
Ondersteunend
Ondoorgrondelijk
Ondoorzichtig
Onduidelijk
Onecht
Oneerlijk
Ongemakkelijk
Ongepolijst
Ongeschoold
Ongezellig
Ongrijpbaar
Onhandelbaar
Onhandig
Onherkenbaar
Oninteressant
Onnozel
Onopvallend
Onpersoonlijk
Onplezierig
Onpraktisch
Onprettig
Onstandvastig
Onsympathiek
Ontoegankelijk
Ontspannend
Ontzag inboezemend
Onverschillig
Onverwacht
Onverzettelijk
Onverzorgd
Onvriendelijk
Onwelwillend
Onzeker
Onzorgvuldig
Oog voor kwaliteit
Oogt goed
Op gegeven moment aan gaan
ergeren
625. Op uiterlijk gelet
626. Op voorgrond dringend
627. Op willen vallen
628.
629.
630.
631.
632.
633.
634.
635.
636.
637.
638.
639.
640.
641.
642.
643.
644.
645.
646.
647.
648.
649.
650.
651.
652.
653.
654.
655.
656.
657.
658.
659.
660.
661.
662.
663.
664.
665.
666.
667.
668.
669.
670.
671.
672.
673.
674.
675.
Opdringerig
Open
Openhartig
Openlijk
Openstaan
Opgeblazen
Opgeruimd
Opgewekt
Oppervlakkig
Opportunistisch
Oprecht
Opschepperig
Optimistisch
Opvallend
Ordinair
Origineel
Oubollig
Oud
Oud mannetje
Oudere leeftijd
Oudere vrouw in café, Jordaans type
Ouderwets
Oudheid
Over de top
Overal bij nodig
Overbodig
Overdone
Overdreven
Overdreven aandachttrekkend
Overheersend
Overleg
Overmatig
Overweldigend
Persoon van deze tijd
Persoon waar mensen van
afhankelijk zijn
Persoonlijkheid
Pessimistisch
Pijnlijk
Pinnig
Pittig
Plezierig
Plomp
Poes pas
Poezelig
Politie, burgerwacht
Pompeus
Populair
Positief
Appendix G
676.
677.
678.
679.
680.
681.
682.
683.
684.
685.
686.
687.
688.
689.
690.
691.
692.
693.
694.
695.
696.
697.
698.
699.
700.
701.
702.
703.
704.
705.
706.
707.
708.
709.
710.
711.
712.
713.
714.
715.
716.
717.
718.
719.
720.
721.
722.
723.
724.
Postmodern
Potsierlijk
Prachtig
Praktijkmens
Praktisch
Praktisch ingesteld
Praktisch inzicht
Pralerig
Prater
Precies
Pretendeert iets te zijn, is het niet
Pretendeert meer dan het is
Pretentie
Prettig
Pretvogel
Prikkelbaar
Prikkerd
Prikt mensen door opmerkingen
Probleemloos
Professioneel
Protserig
Pruttelaars
Punctueel
Puntig
Raar
Rank
Realist
Recht
Recht door zee
Recht op zijn doel af gaan
Recht toe recht aan
Recht voor z'n raap
Rechtlijnig
Redelijk van lengte
Retro
Rijk aan fantasie
Rijzig persoon
Robuust
Romantisch
Rommelig
Rond
Rondborstig
Rustgevend
Rustig
Ruw
Ruwe bolster blanke pit
Saai
Schattig
Scherp
725.
726.
727.
728.
729.
730.
731.
732.
733.
734.
735.
736.
737.
738.
739.
740.
741.
742.
743.
744.
745.
746.
747.
748.
749.
750.
751.
752.
753.
754.
755.
756.
757.
758.
759.
760.
761.
762.
763.
764.
765.
766.
767.
768.
769.
770.
771.
772.
773.
Scherpzinnig
Schichtig
Schoon
Schreeuwerig
Schreeuwt net iets te hard
Schriel type
Sensueel
Sentimenteel
Sexy
Sfeer creëren
Sfeervol
Shockerend
Sierlijk
Simpel
Slaat de plank niet mis
Slaat plank mis
Slank
Slappe kwaliteit
Slechte smaak
Slim
Slordig
Smaak
Smaakvol
Smakeloos
Smerig
Snel
Snel tevreden persoon
Sneu
Snoeperig
Snufjes
Snugger
Sober
Sociaal
Soepel
Soepel in omgang
Soft
Solidair
Solide
Somber
Spannend
Speels
Spichtig
Spits
Spitsvormig
Spontaan
Spreekt niet aan
Staat me aan
Standaard
Standvastig
205
206
774.
775.
776.
777.
778.
779.
780.
781.
782.
783.
784.
785.
786.
787.
788.
789.
790.
791.
792.
793.
794.
795.
796.
797.
798.
799.
800.
801.
802.
803.
804.
805.
806.
807.
808.
809.
810.
811.
812.
813.
814.
815.
816.
817.
818.
819.
820.
821.
822.
Appendix G
Star
Statig
Status
Stekelig
Steriel
Sterk
Stevig
Stevig in zijn schoenen staat
Stijfjes
Stijgt boven de massa uit
Stijl
Stijlvol
Stille muis
Stoer
Storend
Straalt kwaliteit uit
Straf
Straight
Strak
strak type
Stralend
Streberig
Streng
Strippig
Succesvol
Suggereert oud te zijn
Sympathiek
Technisch
Tenger
Terug getrokken
Terughoudend
Teveel doorslaan
Tevreden met omgeving
Tevreden met weinig
Theatraal
Tijdgebonden
Tijdloos
To the point
Toegankelijk
Toegewijd
Toeters en bellen
Tolereren
Traditioneel
Trekt niet aan
Trendgevoelig
Trendvolger
Trendy
Troostend
Truttig
823.
824.
825.
826.
827.
828.
829.
830.
831.
832.
833.
834.
835.
836.
837.
838.
839.
840.
841.
842.
843.
844.
845.
846.
847.
848.
849.
850.
851.
852.
853.
854.
855.
856.
857.
858.
859.
860.
861.
862.
863.
864.
865.
866.
867.
868.
869.
870.
871.
Tuttig
Tweeslachtig
Uit balans
Uit de hoogte
Uitdagend
Uiterlijk bol/rond
Uitgaander
Uitgekiend
Uitgesproken
Uitnodigend
Uitslovertje
Uitspattingen
Uitstraling
Uitstraling van ergens boven staan
Vaag
Vakman
Valt op
Van andere wereld
Van de hak op de tak
Van deze tijd
Van vroegere tijden
Vasthoudend
Veel met zijn handen bezig
Veel te vertellen
Veelzijdig
Venijnig
Verborgen
Verdoezelend
Verdraagzaam
Verfijnd
Verhelderend
Verkeerd
Verlegen
Verleidelijk
Verrassend
Verschrikkelijk
Verschuilend
Vertederend
Vertrouwd
Vertrouwen
Vervelend
Vervormd
Verwarrend
Verzorgd
Vies
Vinnig
Vlot
Voelt zich thuis
Vol
Appendix G
872.
873.
874.
875.
876.
877.
878.
879.
880.
881.
882.
883.
884.
885.
886.
887.
888.
889.
890.
891.
892.
893.
894.
895.
896.
897.
898.
899.
900.
901.
902.
903.
904.
905.
906.
907.
908.
909.
910.
911.
912.
913.
914.
915.
916.
917.
Volhouder
Volwassen
Voor verbetering vatbaar
Voor zijn mening uitkomen
Vooraanstaand
Voordring type
Voorgrond
Voorspelbaar
Voortrekker
Vooruitstrevend
Voorzichtig
Voorlijk
Vreemd
Vriendelijk
Vrolijk
Vrouw
Vrouwelijk
Vurig van binnen
Waait met alle winden mee
Waar je tegenop kijkt
Waardevol
Wakker
Walgelijk
Warm
Warmhartig
Warmte
Warrig
Wat ie doet, doet ie goed
Waterig
Weet niet veel
Weet niet wat je er aan hebt
Weet nooit wat je krijgt
Weet veel
Weet wat hij wil
Weet wat je eraan hebt
Weinig zinnigs te zeggen
Weinigzeggend
Wekt irritatie op
Wel willen, maar niet kunnen
Wereldvreemd
Werkpaard
Werkt graag
Weten wat je ermee aanmoet
What you see is what you get
Wiebelig
Wil alles tot in detail geregeld
hebben
918. Wil veel bereiken
919. Wil zich goed presenteren
920.
921.
922.
923.
924.
925.
926.
927.
928.
929.
930.
931.
932.
933.
934.
935.
936.
937.
938.
939.
940.
941.
942.
943.
944.
945.
946.
947.
948.
949.
950.
951.
952.
953.
207
Wild
Willen vlammen
Wispelturig
‘Would be’
Wulps
Zacht
Zachtaardig
Zachtzinnig
Zakelijk
Zal je nooit in de steek laten
Zal mening niet opdringen
Zal niet gauw kwetsen
Zeer conventioneel
Zelfbewust
Zelfingenomen
Zelfstandig
Zeurkous
Zichzelf niet duidelijk manifesteert
Zielig
Ziet er beter uit dan dat hij is
Zit vol met trucs
Zoet
Zonder uitspattingen
Zonnig
Zorgzaam
Zwaar
Zwaar aan tillen
Zwaar op de hand
Zwaar postuur
Zwaar van gemoed
Zwaarlijvig
Zwaarmoedig
Zwierig
Zwoel
208
Appendix H
Color pictures of the stimuli used in the study reported in section 6.9
Set 1
5
6
7
8
2
3
4
Set 2
1
Set 3
13
14
16
15
Set 4
8
9
11
10
209
Appendix J
Dendograms resulting from cluster analysis of all items per subset
Dendogram based on male respondents
LIEF
òø
SCHATTIG
CHARMANT
òú
òôòø
ROMANTIS
GEVOELIG
ò÷ ó
òø ó
NONCHALA
BLIJ
òú ó
òú ó
VROLIJK
OPEN
òôòú
òú ó
VRIENDEL
AARDIG
òú ùòòòòòòòòòòòø
òú ó
ó
GEZELLIG
FLEXIBEL
òú ó
òú ó
ó
ó
RELAXED
VROUWELI
ò÷ ó
òø ó
ó
ó
KWETSBAA
INFORMEE
òôò÷
ò÷
ó
ùòø
UITSLOVE
OPDRINGE
òø
òú
ó ó
ó ó
OVERDREV
DOMINANT
òôòòòòòø
òú
ó
ó ó
ó ó
AGRESSIE
LEUK
ò÷
òø
ó
ó
ó ó
ó ó
AANTREKK
GEWELDIG
òú
òú
ùòòòòòòò÷ ó
ó
ó
INTERESS
PITTIG
òôòòòø ó
òú
ó ó
ó
ó
STOER
EIGENZIN
ò÷
òø
ó
ó
ó ó
ùò÷
OPVALLEN
APART
òôòø ó
òú ó ó
ó
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòø
EXCENTRI
UITDAGEN
ò÷ ó ó
òø ùò÷
ó
ó
ó
ó
WILD
UITBUNDI
òú ó
òú ó
ó
ó
ó
ó
POPULAIR
ENERGIEK
òôò÷
òú
ó
ó
ó
ó
VLOT
JONG
òú
òú
ó
ó
ó
ó
GRAPPIG
ONVOLWAS
ò÷
òø
ó
ó
ó
ó
KINDERAC
SLORDIG
òú
òôòø
ó
ó
ó
ó
CHAOTISC
ONZORGVU
òú ó
òú ó
ó
ó
ó
ó
ONBETROU
RAAR
òú ùòòòòòòòòòòòòò÷
ò÷ ó
ó
ó
DOM
ZIELIG
òø ó
òú ó
ó
ó
ORDINAIR
ENG
òôò÷
ò÷
ó
ó
SERIES
VOLWASSE
òø
òú
ó
ó
CONSEQUE
NUCHTER
òú
òôòø
ó
ó
PRECIES
VOORSPEL
òú ó
ò÷ ùòòòòòòòø
ó
ó
BETROUWB
EERLIJK
òø ó
òú ó
ó
ó
ó
ó
NETJES
VERZORGD
òôò÷
òú
ó
ó
ó
ó
INTELLIG
MANNELIJ
òú
ò÷
ó
ó
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
STRENG
ZAKELIJK
òûòø
ò÷ ó
ó
ó
AFSTANDE
GESLOTEN
òø ùòø
òú ó ó
ó
ó
ONGEZELL
BEKROMPE
òôò÷ ó
òú
ó
ó
ó
VERVELEN
ONINTERE
ò÷
òø
ùòòòòò÷
ó
ONAANTRE
OUDERWET
òôòø ó
òú ó ó
SAAI
RUSTIG
ò÷ ùò÷
òø ó
BESCHEID
BURGERLI
òôòú
ò÷ ó
ONOPVALL
òòò÷
210
Appendix J
Dendogram based on female respondents
BLIJ
òø
VROLIJK
òú
JONG
òú
RELAXED
òôòø
NONCHALA
òú ó
INFORMEE
òú ó
VRIENDEL
òú ó
AARDIG
òú ó
FLEXIBEL
òú ó
OPEN
òú ùòø
VROUWELI
ò÷ ó ó
LIEF
òø ó ó
SCHATTIG
òú ó ó
GEVOELIG
òôòú ó
GEZELLIG
òú ó ó
ROMANTIS
ò÷ ó ó
INTERESS
òø ó ó
GEWELDIG
òôò÷ ùòòòòòòòòòòòòòòòø
LEUK
òú
ó
ó
AANTREKK
òú
ó
ó
CHARMANT
ò÷
ó
ó
VLOT
òø
ó
ó
òú
ó
POPULAIR
ó
ENERGIEK
òôòø ó
ó
PITTIG
òú ó ó
ó
STOER
ò÷ ó ó
ó
UITDAGEN
òø ùò÷
ó
WILD
òú ó
ó
UITBUNDI
òú ó
ó
EXCENTRI
òôò÷
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòø
APART
òú
ó
ó
EIGENZIN
òú
ó
ó
OPVALLEN
òú
ó
ó
GRAPPIG
ò÷
ó
ó
ONBETROU
òø
ó
ó
ORDINAIR
òú
ó
ó
DOM
òú
ó
ó
ZIELIG
òôòòòø
ó
ó
ONVOLWAS
òú
ó
ó
KINDERAC
òú
ó
ó
ó
ó
SLORDIG
òú
ó
ó
ó
CHAOTISC
òú
ó
ó
ó
ONZORGVU
òú
ùòòòòòòòòòòòòòòò÷
ó
KWETSBAA
òú
ó
ó
RAAR
ò÷
ó
ó
OVERDREV
òø
ó
ó
OPDRINGE
òôòø ó
ó
UITSLOVE
ò÷ ùò÷
ó
DOMINANT
òø ó
ó
AGRESSIE
òôò÷
ó
ENG
ò÷
ó
STRENG
òø
ó
AFSTANDE
òú
ó
GESLOTEN
òú
ó
MANNELIJ
òú
ó
ONGEZELL
òôòø
ó
ZAKELIJK
ò÷ ó
ó
OUDERWET
òø ùòòòòòø
ó
BEKROMPE
òú ó
ó
ó
ONINTERE
òôò÷
ó
ó
ONAANTRE
òú
ó
ó
SAAI
òú
ó
ó
VERVELEN
ò÷
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
RUSTIG
òø
ó
BESCHEID
òôòø
ó
BURGERLI
òú ó
ó
VOORSPEL
òú ó
ó
ONOPVALL
ò÷ ó
ó
BETROUWB
òø ùòòòòò÷
EERLIJK
òôòú
NETJES
òú ó
VERZORGD
ò÷ ó
SERIES
òø ó
VOLWASSE
òú ó
CONSEQUE
òú ó
PRECIES
òôò÷
NUCHTER
òú
INTELLIG
ò÷
Appendix J
Dendogram based on vacuum cleaners
VRIENDEL
òø
AARDIG
òú
RELAXED
òú
LEUK
òú
AANTREKK
òôòø
GEWELDIG
òú ó
LIEF
òú ó
SCHATTIG
òú ó
CHARMANT
òú ó
GEZELLIG
òú ó
ROMANTIS
ò÷ ùòòòòòòòòòòòòòø
POPULAIR
òø ó
ó
FLEXIBEL
òú ó
ó
BLIJ
òôòú
ó
VROLIJK
òú ó
ó
ENERGIEK
òú ó
ó
VLOT
òú ó
ó
JONG
ò÷ ó
ó
GEVOELIG
òø ó
ó
OPEN
òú ó
ó
NONCHALA
òú ó
ó
INFORMEE
òôò÷
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòø
VROUWELI
ò÷
ó
ó
DOM
òø
ó
ó
ZIELIG
òú
ó
ó
ORDINAIR
òôòø
ó
ó
VERVELEN
ò÷ ó
ó
ó
KWETSBAA
òø ùòòòòòòòø
ó
ó
RAAR
òú ó
ó
ó
ó
ONZORGVU
òôò÷
ó
ó
ó
SLORDIG
òú
ó
ó
ó
CHAOTISC
òú
ó
ó
ó
ONBETROU
òú
ó
ó
ó
ONVOLWAS
òú
ùòòòòò÷
ó
KINDERAC
ò÷
ó
ó
STOER
òûòø
ó
ó
INTELLIG
ò÷ ùòø
ó
ó
DOMINANT
òø ó ó
ó
ó
AGRESSIE
òôò÷ ó
ó
ó
ENG
ò÷
ùòòòòò÷
ó
OVERDREV
òø
ó
ó
OPDRINGE
òôòø ó
ó
UITSLOVE
ò÷ ó ó
ó
APART
òø ùò÷
ó
EXCENTRI
òôòú
ó
EIGENZIN
òú ó
ó
OPVALLEN
ò÷ ó
ó
UITBUNDI
òø ó
ó
òú ó
ó
UITDAGEN
òôò÷
ó
GRAPPIG
òú
ó
INTERESS
òú
ó
PITTIG
ò÷
ó
ONINTERE
òø
ó
ONAANTRE
òôòø
ó
WILD
ONGEZELL
òú ó
ó
BEKROMPE
ò÷ ó
ó
RUSTIG
òø ùòòòø
ó
BESCHEID
òôòú
ó
ó
VOORSPEL
ò÷ ó
ó
ó
OUDERWET
òø ó
ó
ó
SAAI
òú ó
ó
ó
ó
BURGERLI
òôò÷
ó
ONOPVALL
ò÷
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
NETJES
òø
ó
VERZORGD
òôòòòø ó
CONSEQUE
òú
ó ó
PRECIES
òú
ó ó
BETROUWB
òú
ó ó
EERLIJK
ò÷
ùò÷
STRENG
òø
ó
AFSTANDE
òú
ó
GESLOTEN
òú
ó
SERIES
òôòòò÷
VOLWASSE
òú
NUCHTER
òú
ZAKELIJK
òú
MANNELIJ
ò÷
211
212
Appendix J
Dendogram based on cars
BLIJ
òø
VROLIJK
òú
GEZELLIG
òú
VRIENDEL
òú
AARDIG
òú
OPEN
òôòø
RELAXED
òú ó
NONCHALA
òú ó
FLEXIBEL
òú ó
INFORMEE
òú ùòø
JONG
ò÷ ó ó
APART
òø ó ó
EIGENZIN
òú ó ó
EXCENTRI
òôò÷ ùòòòø
OPVALLEN
ò÷
ó
ó
CHARMANT
òø
ó
ó
ó
ó
ROMANTIS
òú
LIEF
òôòòò÷
ó
SCHATTIG
òú
ó
GEVOELIG
òú
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòø
VROUWELI
òú
ó
ó
GRAPPIG
ò÷
ó
ó
LEUK
òø
ó
ó
AANTREKK
òôòø
ó
ó
INTERESS
òú ó
ó
ó
GEWELDIG
ò÷ ùòòòòò÷
ó
UITDAGEN
òø ó
ó
òú ó
ó
UITBUNDI
òôò÷
ó
PITTIG
òú
ó
WILD
STOER
òú
ó
VLOT
òú
ó
ENERGIEK
òú
ó
POPULAIR
ò÷
ó
RUSTIG
òø
ó
BESCHEID
òôòø
ó
òú ó
ó
ONOPVALL
ò÷ ó
ó
SERIES
òø ùòòòòòòòòòòòòòòòø
ó
BURGERLI
VOLWASSE
òôòú
ó
ó
ZAKELIJK
ò÷ ó
ó
ó
BETROUWB
òø ó
ó
ó
EERLIJK
òú ó
ó
ó
NETJES
òú ó
ó
ó
VERZORGD
òú ó
ó
ó
CONSEQUE
òôò÷
ó
ó
PRECIES
òú
ó
ó
NUCHTER
òú
ó
ó
VOORSPEL
òú
ó
ó
INTELLIG
ò÷
ó
ó
SLORDIG
òø
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
CHAOTISC
òú
ó
ONZORGVU
òú
ó
ONBETROU
òú
ó
KWETSBAA
òú
ó
ONVOLWAS
òôòø
ó
KINDERAC
òú ó
ó
DOM
òú ùòòòòòòòòòø
ó
RAAR
ò÷ ó
ó
ó
ZIELIG
òø ó
ó
ó
ó
ó
ENG
òú
ó
ó
ORDINAIR
ò÷
ùòòòòò÷
OUDERWET
òôò÷
OVERDREV
òø
ó
OPDRINGE
òú
ó
UITSLOVE
òôòòòòòø
DOMINANT
òú
ó
ó
MANNELIJ
òú
ó
ó
AGRESSIE
ò÷
ùòòòòò÷
STRENG
òø
ó
AFSTANDE
òôòø
ó
ONGEZELL
òú ó
ó
GESLOTEN
ò÷ ùòòò÷
ONINTERE
òø ó
ONAANTRE
òú ó
SAAI
òôò÷
VERVELEN
òú
BEKROMPE
ò÷
ó
213
Appendix K
Dendograms resulting from cluster analysis of the six groups of items individually
BLIJ
òûòòòø
VROLIJK
ò÷
GEZELLIG
òòòòò÷
JONG
òòòòòòòòòòò÷
FLEXIBEL
òòòòòòòûòòòòòø
ó
OPEN
òòòòòòò÷
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòø
ùòòòòòø
ùòòòòòø
ó
ó
NONCHALA
òòòòòòòòòòòòòôòòòú
ó
INFORMEE
òòòòòòòòòòòòò÷
ó
ó
AARDIG
òûòòòòòòòòòø
ó
VRIENDEL
ò÷
ó
ùòòòø ó
ùòø
RELAXED
VROUWELI
òòòòòòòòòòòòòòò÷
ùò÷
ó ó
ó ó
AANTREKK
òûòòòø
ó ó
LEUK
ò÷
GEWELDIG
òòòòò÷
ùòòòòòòòòòòòòòòòòòòòòòòò÷ ó
POPULAIR
òòòòòø
ó
ó
VLOT
òòòòòôòòòòòòòòòòòòòòòòò÷
ó
ENERGIEK
òòòòò÷
ó
CHARMANT
òòòòòûòòòø
ó
ROMANTIS
òòòòò÷
LIEF
òòòûòòòø ó
ùòòòòòòòòòòòòòòòòòø
ó ó
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
SCHATTIG
òòò÷
GEVOELIG
òòòòòòò÷
ùò÷
APART
òûòø
EIGENZIN
ò÷ ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòø
OPVALLEN
òòò÷
GRAPPIG
òòòòòûòòòòòòòòòòòòòòòòòòòòòòòòòø ó
ó
INTERESS
òòòòò÷
ùò÷
ó
UITDAGEN
òòòûòòòòòòòø
ó
ó
WILD
òòò÷
EXCENTRI
òòòòòûòòòòò÷
ó
UITBUNDI
òòòòò÷
ó
PITTIG
òòòòòòòûòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
STOER
òòòòòòò÷
ùòòòòòòòòòòòòòòòø
ùòòòòòòòòòòòòòòòòòòò÷
ó
OPDRINGE
òûòòòø
OVERDREV
ò÷
UITSLOVE
òòòòò÷
AGRESSIE
òòòòòòòòòòòûòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
DOMINANT
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòø
ó
214
Appendix K
CHAOTISC
òûòø
SLORDIG
ò÷ ùòòòòòòòø
ONZORGVU
òòò÷
ONBETROU
ùòòòòòòòø
RAAR
òòòòòòòòòòòòòòòòòòò÷
KINDERAC
òòòòòòòûòòòòòòòòòòòòòòòòò÷
ó
ONVOLWAS
òòòòòòò÷
ùòòòòòòòòòòòòòø
ùòòòòòø
ùòòòòòòòòòø
DOM
òòòòòòòòòòòòòòòûòòòòòòòø
ZIELIG
òòòòòòòòòòòòòòò÷
ó
ó
ùòòòòòòòòòòò÷
ó
ORDINAIR
òòòòòòòòòòòòòòòòòòòòòòò÷
ENG
òòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòòò÷
ó
AFSTANDE
òûòòòòòø
STRENG
ò÷
GESLOTEN
òòòòòòò÷
BEKROMPE
òòòòòòòòòòòûòòò÷
ùòòòòòòòòòòòòòòòòòòòòòòòòòòòø
ONGEZELL
ó
ó
MANNELIJ
òòòòòòòòòòòòòòòòòòòòò÷
ó
ONINTERE
òûòòòòòø
ó
ONAANTRE
ò÷
SAAI
òòòòòòò÷
OUDERWET
òòòòòòòòòòòòòòòòòòòòòòò÷
ùòòòòòòòø
ùòòòòòø
ùòòòòòòòòòòòòòòòø
ó
ùòòòòòòòòòòòòòòòòòòòòòòòòò÷
SERIES
òûòòòòòòòø
VOLWASSE
ò÷
CONSEQUE
òûòòòòòø ó
ó
PRECIES
ò÷
ùòòòø
ùòòòø
ùò÷
NUCHTER
òòòòòòò÷
VOORSPEL
òòòòòòòòòòòòò÷
ó
ùòòòòòòòòòòòòòòòòòòòø
ZAKELIJK
òòòòòòòòòòòòòòòòò÷
ùòòòòòòòòòòòø
BETROUWB
òûòòòòòòòòòø
ó
ó
EERLIJK
ò÷
ùòòòòòòòòòòòòòòòòòòòòòòòòò÷
ó
ó
ó
NETJES
òòòûòòòòòòò÷
ó
VERZORGD
òòò÷
ó
BESCHEID
òòòòòòòûòòòòòòòòòòòòòòòòòø
RUSTIG
òòòòòòò÷
BURGERLI
òòòòòòòòòòòòòòòòòòòûòòòòò÷
ONOPVALL
òòòòòòòòòòòòòòòòòòò÷
ó
ùòòòòòòòòòòòòòòòòòòòòòòò÷
215
Appendix L
List of the dictionary definitions of the items (in Dutch)
Vrolijk:
Van Dale, groot woordenboek:
•
In lichte stemming; opgeruimd, blij, jolig, uitgelaten
•
Waarin of waarover men zich vermaakt
•
Druk, levendig
•
Aangenaam stemmend, levendig van kleur
•
Levendig van voorstelling
Van Dale, basis woordenboek:
•
Blij en opgewekt, waar je vrolijk van wordt
Idioticon van de persoonlijkheid:
•
Opgewekt, gezellig, joviaal
•
Leutig, lustig, feestnummer, giebel, giechel
Open:
•
Niet gesloten, niet dichtgemaakt; geopend
•
Toegankelijk
•
Onbedekt, niet overdekt
•
Niet gevuld
•
Zonder bepaalde aanleiding; onbepaald
•
Onafgesloten, nog niet verrekend
•
Niet bezet, niet door een ander ingenomen, niet vervuld
•
Met openingen
•
Zonder beslissing
•
Niet afgesloten, waar je zo in of bij kunt, het tegenovergestelde van dicht en
gesloten
•
Leeg, niet ingevuld
•
Met een karakter of houding waaruit blijkt dat je je niet voor je omgeving afsluit
Idioticon van de persoonlijkheid
•
Opgewekt, joviaal, gezellig
•
Kent geen geheimen, staat open voor mensen, dingen
Relaxed:
•
Op zijn gemak; ontspannen
•
Lekker op je gemak, ontspannen
•
Rustig, bedaard
•
Bedaard, doet alles op z’n dooie akkertje, doet dingen op zijn gemak, niet
overhaastig
216
Appendix L
Leuk:
•
Lauw
•
Kalm, onverschillig
•
Aardig, amusant, vrolijk, grappig
•
Aardig, aantrekkelijk, charmant
•
Waar je om kunt lachen; aardig, grappig
•
Waarvan je in een goede stemming komt
•
Bedrijvig, energiek, vief
•
Aanminnelijk, bekoorlijk, bevallig, charmeur, heeft charme
Vlot:
•
Niet vastzitten, voldoende water onder zich hebbend om te kunnen drijven of
varen; drijvend
•
Zwevend, drijvend
•
Gemakkelijk van of in zijn beweging (vnl. van taal); vloeiend
•
Zonder oponthoud of moeilijkheden; vlug, onbelemmerd
•
Gemakkelijk in de omgang, zich gemakkelijk bewegend, niet stijf, verlegen of
kleinzielig; soepel, natuurlijk, los, ongedwongen, onbekrompen
•
Een indruk van losheid, ongedwongenheid gevend
•
Snel of gemakkelijk verlopend
•
Gemakkelijk in de omgang
•
(van kleding) Iets dat leuk staat en modern is
•
Opgewekt, gezellig, joviaal
•
Snelle jongen, kek type, heeft flair
Schattig:
•
Erg lief; allerliefst, alleraardigst, snoezig
•
Erg lief
•
Erg leuk om te zien
Idioticon van de persoonlijkheid – niet vermeld
Eigenzinnig:
•
Zijn eigen zin, wil volgend (m.n. tegen de waarschuwingen van anderen in)
•
Gezegd van iemand die zijn eigen zin doet zonder zich aan te trekken van wat
anderen zeggen
Appendix L
217
•
Opstandig, dwars, veeleisend
•
Laat zich niet overhalen/overreden, volgt zijn eigen zin
Uitdagend:
•
Een reactie uitlokkend; provocerend, seksueel prikkelend, gewaagd
•
Gezegd van iemand die dingen doet of zegt om een reactie bij iemand uit te
lokken; provoceren
•
Fel, lawaaierig, druk
•
Provocerend, doet iets om een reactie uit te lokken
Interessant:
•
De aandacht/nieuwsgierigheid prikkelend;
karakteristieke kenmerken aantrekkelijk
•
Indruk geven van belangrijkheid
•
Baatzuchtig, inhalig
•
Indringerig
•
Financieel aantrekkelijk
•
Gezegd van iets dat je belangstelling opwekt
Idioticon van de persoonlijkheid – niet vermeld
belangwekkend,
boeiend,
door
Pittig:
•
Vol pitten
•
Krachtig, energiek
•
Waar pit in zit (tegenovergestelde van flauw), pikant, heet
•
Kernachtig, geestig, snedig
•
Aantrekkelijk door expressiviteit
•
Moeilijk, maar te doen
•
Gezegd van iemand die pit (= kracht, energie, levendigheid) heeft
•
Scherp gekruid
•
Behoorlijk moeilijk
•
Fel, lawaaierig, druk
•
Kittig
Opdringerig:
•
Neiging hebbende om zich aan anderen op te dringen (= op onbescheiden of
lastige wijze iemands gezelschap zoeken)
218
Appendix L
•
Gezegd van iemand die voortdurend op een vervelende manier contact zoekt
•
Egocentrisch, tegendraads, ontactisch
•
Plaatst zich op de voorgrond, vraagt aandacht, drukt anderen weg, bemoeit zich
overal ongevraagd mee
Dominant:
•
Overheersend, de overhand hebbend
•
De werking van anderen onderdrukkend
•
Overheersend
•
Bazig, schreeuwerig, dominant
•
Doordrammerig, imponerend, imposant, is zeer aanwezig, heeft
overwicht/overtuigingskracht, domineert zijn omgeving,
bluft/overbluft/overblaft/overdonderd/verduvelt/overheerst/ overschreeuwt.
Slordig:
•
Niet verzorgd, niet netjes; onordelijk, onverzorgd, haveloos
•
Onachtzaam of onordelijk in zijn werk of in zijn gedragingen; negligent,
nonchalant
•
Sporen dragend van onachtzaamheid of zorgeloosheid
•
Zonder morele bekommering
•
Meer dan behoorlijke grootte; dik, flink
•
Niet netjes, onverzorgd
•
Chaotisch, inaccuraat, onbedachtzaam
•
Sloddervos, rommelkont, laat dingen verslonzen, ziet er onverzorgd/haveloos uit,
raffelt dingen af, doet alles rommelig, het is bij hem een zooitje
Kinderachtig:
•
Kinderlijk
•
Als van of voor een kind; pueriel
•
Erg klein, gering
•
Nog geheel een kind; onmondig
•
Als volwassene de eigenschappen van een kind vertonend waar die niet gewenst
zijn, zonder waardigheid of flinkheid, of wel onnozel, flauw of laf; infantiel, pueriel
•
Waar je te groot voor bent
•
Gezegd van iemand die gauw gekwetst of beledigd is; flauw, kleinzielig
Appendix L
219
•
Schrikachtig, bangelijk, vreesachtig
•
Pueriel, een flauwerd
Dom:
•
Niet gemakkelijk begrijpend, beperkt verstand; stupide
•
Onwetend
•
Redeloos
•
Van weinig verstand, van domheid getuigend; stupide
•
Waarbij het verstand niet gebruikt wordt
•
Vermoeid van geest
•
Stijf en ongevoelig: suf
•
Niet slim, niet verstandig; onnozel
•
Schrikachtig, bangelijk, vreesachtig
•
Traag/kort van begrip, niet erg slim, stom
Afstandelijk:
•
Zich op een afstand houdend, zich niet persoonlijk bij iets betrokken voelen;
terughoudend, koel
•
Alsof je er niet bij betrokken bent, onverschillig
•
Gesloten, zwijgzaam, introvert
•
Gedistantieerd, bewaard afstand, weert mensen af, wijst toenadering af
Saai:
•
Weinig levendig, waarvan weinig uitgaat, niet amusant; vervelend, droog
•
Niet vlot; traag
•
Waarbij of bij wie je je verveelt; eentonig, droog
•
Timide, saai, tam
•
Langdradig, grijze muis
Serieus:
•
Ernstig
•
Oprecht, gemeend
•
Ernstig, echt gemeend
220
Appendix L
•
Voorzichtig, serieus, perfectionistisch
•
Neemt alles in ernst op
Eerlijk:
•
Wars van leugen en bedrog; oprecht, deugdzaam, rechtschapen, tgov oneerlijk
•
Zich niet vergrijpend aan andermans eigendom, te vertrouwen
•
Zoals een rechtschapen of betrouwbaar persoon eigen is; oprecht, openhartig
•
In overeenstemming met te goede trouw
•
Zonder onzichtbare gebreken
•
Indruk van lichte sierlijkheid en harmonie gevend
•
Ouderwets degelijk, zuiver en goed
•
Zonder overbodige vergraaien, zuiver functioneel
•
In overeenstemming met iemands eer; fatsoenlijk, behoorlijk
•
Eervol
•
Zonder leugen of bedrog; oprecht, fair
•
Oppassend, correct, welgemanierd
•
Speelt open kaart, komt altijd ergens rond voor uit
Bescheiden:
•
Geen te hoge gedachten van zichzelf hebbend en zich dienovereenkomstig
gedragen, zonder enige aanmatiging; ingetogen, zedig, nederig
•
Beleefd, voorzichtig geuit, niet opdringerig; discreet
•
Zich niet opdringend, niet schril of schel
•
Niet groot
•
Verstandig, oordeelkundig
•
Gezegd van iemand die zich niet op de voorgrond plaatst
•
Niet groot of niet veel
•
Zachtmoedig, geduldig
•
Houdt zich op de achtergrond, blijft achter de schermen, effaceert zich (cijfert zich
weg), weet zijn plaats.
221
Appendix M
Color pictures of the stimuli used in the study reported in section 7.5.1
Stimulus 1
Stimulus 2
Stimulus 3
Stimulus 4
Stimulus 5
Stimulus 6
222
223
Curriculum Vitae
Pascalle Govers (1974) started studying Psychology at Tilburg University in 1993.
In September 1998 she received her master’s degree with distinction (annual best
graduation project in psychology). In December of that same year she started her
doctoral dissertation research at the faculty of Industrial Design Engineering at Delft
University of Technology. As of January first, 2004, she works as an Assistant Professor
at Nyenrode University.
Her research investigates the “irrational” aspects of consumer behavior, and is
mostly concerned with symbolic consumption. Her current research includes studies into
the understanding and measuring of product personality. She has presented her research
at several international conferences, and won the best presentation award at the first
Ph.D-day of Industrial Design Engineering at Delft University of Technology.

Product Personality

Transcription

Similar documents

Personality Disorders

Get to know yourself! know yourself!

Eysenck`s trait theory - The Grange School Blogs

Battle of the Codes flier

SLIM ad - The National institute for Marketing in Sri Lanka

PowerPoint

Activity: Financial Personality Quiz

Media Kit - Advertise on Cities 97

In the world of business, graphic design has a lovely and important

Miss Congeniality: