4 Data Collection Methods

4
1
Q
.
'
Data Collection Methods
ualitative researchers typically rely on four methods for gathering information: (a) participation in the setting, (b) direct obsenation, (c) indepth interviewing, and (d) analyzing documents and material culture. These
methods form tbe core of qualitative inquiry-the staples of the diet. Several
specialized methods supplement these. This chapter provides a bnef discusion of both theprimary and secondary methodsforconsiderationin designing
qualitative study. This discussion doesnot replace the many excellent,
etailed references on specific data collection methods (we refer lo. severa1 at
the end of this chapter); its purpose is to guide the proposal wnter in
stipulating the methods of choice for his study and descnbing, for the reader,
how the data generated will inform his research questions. How theresearcher
plans to use specific methods, howevec depends on several considerations.
.
Extending the discussion in Chapter 1, Brantlinger's.(l997) workprovides
a useful summary of seven categones of crucial assumptions. The iüst is how
the researcher views the nnture ofthe resenrch: 1s the inquiñy technical and
neutral, intending to conform to traditional research within her discipline, or
is it controversial and critica1 with an explicit political agenda? Second, how
Daia Collection Melhods
does she construe her location, her posiiioning relotive to the participanrs.
Does she view herself as distant and objective or intimately involved in the
lives of the participants? Third, what rs the "direetion of he? 'gaze' " (p. 4):
1s it riutward toward "others"-externalizing the research problem-or does
it inciude explicit inner contemplation? Fourth, what is the purpose clf the
research: Does she assume that the primary purpose of the study is professional and essentially private (e.g ,promoting her career), or 1s it intended to
be useful and informative to the participants or the site? Related to the fourth
category is thefifth Who is the inrendedaudienceof the srudy-the scholarly
community or the panicipants themselves? Sixth, what 1s her poIitical positioning: Does she vicw research as neutral. or does she claim a politically
explicit agenda? Finally. the seventh arena is how she views the exercise of
agcncy: Doef she view herself and the participants as essentially passive or
as "engaged in local praxis" (p. 4)? Assumptious made in these seven rategories shape how the specific research methods are eonceived and implemented
throughout a study: explicit discussion of assumptions strengthens the overall
logic and integnty of the proposal.
Primary Rlethads
Developed primanly from the disciplines of cultural anthropology and
qualitative swiology, participant observation (as this method is typically
called) is both an overall approach to inquiry aud a data-gathering method.
Participant observation is to some degree an essentialelement of al1 qualitative
studies. As its name suggests, participant observation demands firsthand
involvement.in the social world chosen for study Immersion in the setting
allows the researcher to hear, see, and hegin to expenence reality as the
participants do. Ideally, the researcher spends a considerable amount of time
in the setting, learning about daily life there. This immersion offers the
researcher the o p p w n i t y to learn directly from his own experience of the
setting. These personal reflections are integral to the emerging analysis of the
cultural group of interest.
This method for gathering data is basic to al1 quaiitative studies and forces
consideration of the role or stance of the researcher as a participant obsenrer.
We haveexplored issues nf rolemorefully in Chapter 3; we here reiterate that,
'
..
,>
at the proposal stage, it is helpful to elaborate on the olannéd
participation, what the nature of that involvement is likely to be, how
will be revealed about the study's purpose to the people in the settins hopt
intensively the researcher will be present, and how focused the participatioo
will be. The researcher Bhould stipulate bow his participation will infonn the
research questions.
F'
I, . Observation
-
-a:
Observation entails the systematic noting and rewrding of events, behavS, and artifacls (objects) in the social setting chosen for study. The observational record is frequently referred to asfieldnotes-detailed, nonjvdgmental, concrete descriptions of what has been o b s e d . For studies relying
exclusively on observation, the researcher makes no spffiial effort to have a
particular role; often, to be tolerated as an unobausive observer is enougb.
Classroom ohservational studies are one example ofien found in education.
Through observation, the researcher documents and describes complex actions and interactions. Without orber sources of i%miation, however, the
meaning of these actions can only be inferred. This method a3sumes that
behavior is purposeful and expressive of deeper values and beliefs. Observa",
tion can range from highly suuctured, detailed notation of jehavior guided
by checkiists to more holistic description of events and beh$vior.
In the early stagts of qualitative inquiry, theresearcher flpically enters the
. setting with broad areas of interest but without predetennined categones or
strict observational checklists. The value here is that the researcher is able lo
discover the recumng panems of behavior and relationships. After sueb
panems are identified and described through early analysis of field notes.
*'
checklists become more appropriate and context sensitive. Pocused observa' tion then is used at later stages of the study, usually Po check analytic themes
. to see, forexample, if they explain behavior and reiationships over a long time
bin a v d e t y of settings.
Observation is a fundamental and highly importan1 method in al1 qualitative inquiry: It is used to discover complex interactíons in natural social
settings. Even in in-depth interview studies; observation plays gn important
role as the researcher notes the interviewee's body language and affect in
: addition to her words. It is. bowever, amethod that requires a great deal of the
researcher. Discomfort, uncomfortable ethieal dilemmas and even danger, the
*difficulty of managing a relatively unobuusive role, and the challenge of
"
u
'7
108
D E S I G N I N G QUALITATIVE RESEARCH
identifying the "big picture" while finely observing huge amounts of fastmoving and complex behaviorare just afew of the challenges. At theproposal
stage, the researcher should describe the purpose of obserying, in what phase
of the study it is likely to be most fruitiul, and how field notes will be used 10
respond to t t e research questions.
Figure 4.1 provides an example of field notes conducted for a smdy of
kindergarten teachers. In these field notes, O'Hean-Curan (1997) has formatted the descriptive notes in a column on the left while reserving a second
column on the right for her obsewer's comments. These comments include
her emerging analytic insights and comments about the actions. Observer's
comments are often a quite fruitful source of analytic insights and clues to
focus data collection moretigtitly. They may also provide important questions
for subsequent interviews.
In-Depth hterviewing
Qualitative researche$ rely quite eñtensivaly on in-depth interviewing.
Kahn and Cannell (195J) describe iiiterviewing as "a conversation with a
purpose" (p. 149); it mty be the overall strategy or one of several methods
employed in a study. Interviewing varies in terms of a priori stnicture and in
the latitude the interviewee has in responding to questions. Patton (1990,
pp. 280-290) categorizes interviews into three general types: the informal
converfational interview, the general iiiterview guide approach, and the standardized open-ended interview.
Typically, qualitativein-depth interviews aremuch morelikeconversations
than formal events with predetermined response categories. The researcher
explores a few general topics to help uncover the participant's views but
othenvise respects how the participant frames and structures the responses.
This, in fact, is an assumption fundamental to qualitative research: The
participant's perspective on the phenomenon of interest should unfold as the
participant views it, not as theresearcher views it. A degreeof systematization
in questioning may be necessary in, for example, a multisite case study or
when many participants are interviewed. The most important aspect of the
interviev er's approach concerns conveying the attitude that the participant's
views are. valuable and useful. This may, however, evoke ethical dilemmas for
the interiiewer; we discuss these more fully below.
Interviews have particular strengths. An interview is a useful way to get
large amounts of data quickly. When more than one persou participares (e.g.,
Data Collection Methads
109
SAMPLE FIELD NOTES
niesday, November 13,1997 1240 p m.
Observation
Observer's Commenü
There are 17 children in the room. There are
3 adults: 1 teacher, 1 classroom assistant, and
1 student teacher (the student teacheris an
older womm).
The room is in the basement of the schwl.
The school is a brick builtling approximately 90
to 100 years old. The room is about 40 feet by
30 feet. The room is carpete and is sectioned off
by fumiture. There is an atea with big books and
a chari in the left-band back comer of the room.
Next 10 that is a shelf with amixture of srnall
books, tapes, and big books in baskets. Next to
that is a small area with toy kitchen furniture and
dolls. There is an area with several tables in front
of the kitchen area. There are many small chairs
pulled up lo the table. In the front of the room is
marea 4 t h a sand table. There is a semicirde
table in the left-hand front comer of the room
The walls arz coloríul with papers that have been
made by the children. One wall has papers with
apples on them. Another wall has pictures of
children with their names on the front of the,
papers. There are several small windows in the
room and the fluorescent lighting seems lo be the
major source of light.
The children have just come into the room.
They have put their coats and backpacks onto
their hooks in the hall outside.
Figure 4.1. Sample Field Notes
77ie teacher seerns to
have done o greal job
Tke space rtselfts no1
oprima1
Most of rhe chrldren
nppear lo know
the routine.
Data Coll~clionMcrhods
focus group inter\ iews, discussed later). the inter\~iewprocessgathers a wide
variety of informallon across a larger number of S U ~ J ~ Cthan
~ S if there were
fewer particlpantkthe familiar trade-off between breadth and depth. Imme&ate follow-up and clanfication are possihle. Combined with ohservation,
interviews aUow the researcher to understand the meanings that people hold
for their everyday activities.
Interviewing has limitations and weaknesses, howeuer. Interwewsinvolve
personal interaction; cooperation rs essential. Inte~vieweesmay be unwilling
or may be uncomfonable sharing al1 that the interviewer hopes to explore. or
they may be unaware Of recurrinp, patterns in their lives. The interviewer may
not ask questions that evoke long narratives from participants either because
of a lack of expenise or familiarity with local language or because of lack of
skill. By the same token, she may not broperly comprehend responses to the
questions orvarious elements of the conversahon. And atrimes, intemiewees
may have good reason not to be tmthfui [see Douglas, 1976,for a discussion).
Interviewers should have superb listening skills and be skillful al personal
interaccion, question framing, and gentle probing for elaboration. Volumes of
data can be obtained through interviewing, but the data are time-consnming
to analyze. Finally, tbere is the issue of the quality of the data. When the
researcher is using in-deplh interviews as the sole way of gathering data, she
i
sh<uld
have demonsuated through the conceptual framework that the purpose
of the study h lo uncover and describe the participants' perspectives on
events-that is, that the subjective view is what matters. Studies making more
objectivist assumptions would triangulate interview data with data gathered
through other methods.
Figure 4 2 ;)rouides elabcrated notes from an interview conducted for a
study of students of color in a community college. Koski (1997) was particularly interested in how these students identified and defined effective teachers.
She was intrigued with the notion of culturally relevant pedagogy and conducted several in-depth interviews with teachers identified by students as
especially effective. She has formatted the notes from the internew to provide
space for her comments, as did O'Hearn-Curran in the field notes presented
in Figure 4.1.
In addition to generic in-depth interviewing are several more specialized
forms of interviews, including ethnographic interviewing, phenomenological
interviewing, elite interviewmg, focus group interviewing, and interviewing
cbildren. We now describe each of these methods briefly.
.
111
SAMPLE INTERVIEW NOTES
Inlerview with DC
October 15,1997
DCir on &visor with on OC&LM'C
depmnent. Tiz infervieww u ser
up by the +m
1:30-3:40
Soting: DC's office in h e scadcmic dcpanmcnt It's
bnght and livcly-colorful Lspy an one wall. p r i okay." Bookr;&d papen are evwywherr. On the
comer of h e ddesk are some wwd gamss: tieiac-toe,
pynmid, and othm.
DCis a m d . dad<<olored w o m with hcrhaiiin
amall but longish braidr al1 ovw her hcad. Shc wcars
large glasses and a pinkish shack of lipstick ihat eomplemnits her caloring. Shc is linly. with a m d y
smilc and a quick laugh. Shc commcnu on ha
heighi: 'Tm smaller than al1 my advism. so I'm not
a t h m t lo anyonc."
1cxplain q b t I'm ini-ted
in and whatmy
projsst ir sbout. 1tell her that I would likc t h m
thingr from her: Ons ir an idea of what she 8s an adviser hi& are h e attnbules of a e w d lcachcr and
.
DC l¿rtenr very inrently k m
i#
1
Tius ir M O X I O Y L ~ ~ O ~ ~ ~ N
t a n to fot the projwf.
DC: ''o~ay.~ o o d .
ask me a qucrtion:
KK: ''Tell me a lirtk bit sbout w b t you do."
DC: "l'm an adviscr here. We get thrhc iin frcsh off
w&,
Jor m undJor hcr 1 hatn't
sum who, todo. Thir arnarol
Shehnnd<mnfomth@~k)iar
workd on wilh a student. l w i
educstbnal plan 1 Ike it whui &cy know what'r
' Ihcn somconc comes in nnd tclk
erpected of them:'
herrk ha?oson impononrphone
DC: 'Tht educational plan liss a01ody courses m col1 chal 1luy m ' 1 irnnrfer Shc
leoves for &u: ten mnutcr.
k tal<en but clubs and other smduit anivitieg. it Iim
1 m @fe ro bok omund
ihe advising evuits the student wiU suend."
D C ~ ~KK:
~ K . msny w k n t s do yo" have?'
DC: "About 1W."
XK: "1@! Are you sble to have a relstianship with
DC: "1 feel I'm an advocatc for stndentr. 1do whatevw needs lo be done to gct thcm through fhis. 1 te11
hcm not to overload. to d a . about this. . ..1think.
beidg honcrt wirh rtudaim ir impaiíant. If I dw't
Data Collection Mcthods
DESIGNING QUALITATIVE RESEAICCH
112
113
l
Ethnographic Interviewing
-*
i
Based on the discipline of cognitive anthropology, ethnographic interviewing elicits the cognitive stnictures guiding participants' worldviews.
Described a; "a particular kind-of speech event" (Spradley, 1979, p. ,18),
ethnographic questions are used by the ethnographer to gather cultural data.
Spradley (1979) identifies three main types of questions: descriptive, strüctural, and contrast. Descriptive questions allow the researcher to collect a
sample of the participant's language. Stnictural questions discover the basic
units in his cultural knowledge, and contrast questions provide the ethnographer with the meaning of vanous terms in the participant's language.
The value of the ethnographic interview lies in its focus on culture through
the participant's perspective and through a firsthand encounter. This approach
is especially useful for elieiting participants' meanings for events and behaviors and for generating a typology of cultural classification schemes. It also
highlights the nuances of the culture. The method is flexible in formulating
working hypotheses and avoids oversimplification in description and analysis
because of its nch narrative descriptions.
There are weaknesses ;n this method, however. The ethnographer may
impose her values through h e phrasing of questions or inierpretation of data
If the member of the cultural group chosen to participate does not represent
that culture, the subsequent analysis will be impovenshed. Thesuccess of this
method, as in al1 interviewing, is highly dependen1 on theresearcher's interpersonal skills.
Phenomenological Interviewing
Phenomenological interviewing is a specific type of in-depth interviewing
grounded in the tradition of phenomenology. Phenomenology is the study 0f
lived experiences and the ways we undeístand those experiences to develop a
worldview. It rests on an assumption that there is a structure and essence t0
shated expenences that can be narrated. The purpose of this type of interviewing is to describe the meaning of a concept or phenomenon that several
individuals share.
As developed by Seidman (1998), three in-depth interviews compose
phenomenological inquiry. The fint focuses on past experience with the
phenomenon of interest; the second focuses on present experience; and the
third joins these two narratives to describe the individual's essential experience
I
I
1
I
I
1
l
with thk phenomenon. Prior tointerviewing, however, the phenomenological
researcher has written a full description of her own experience, thereby
bracketing her experiendes from those of the interviewees. This phase of the
i n q u i j is referred to as. epoche. The purpose of this self-examination is for
the researcher to gain clarity from her own preconceptions, and it is part of
the "ongoing process rather than a single fixéd event" fPatton, 1990, p. 408).
The next phase is called phenomenological reduction; here, the researcher
identifies the essence of the phenomenon (Patton 1990, p. 408). The researcher then clusters the data around themes that describe the "textures of
the experience" (Creswell, 1998, p. 150). Thefinal stage. srructural synrhesis,
involves the imaginative exploration of "al1 possible meanings and divergent
perspectives" (Creswell, 1998, p. 150) and culminates in a description of the
essence of the phenomenon and its deep stnicture.
The primary advantage of phenomenological intewiewingis that itpermits
an explicit focus on the researcher'spersonal experience combined with those
of the interviewees. It focuses on the deep, lived meanings that events have
for individuals. assuming that these meanings guide actions and interactions.
It is, however, quite labor-intensive and requires a reflective t& of mind on
the part of the researcher.
1
t
"Elite" Interviewing
I
1
1
t
An elite interview is a specialized case of interviewing that focuses on a
particular type of interviewee. "Elite" individuals are those considered to be
' influential, prominent, andlor well-informed people in an organization or
community; they are selected for interviews on the basis of their expertise in
areas relevant to the research.
Elite interviewing has many advantages. Valuable information can be
'
gained from these participants because of the positions they hold in social,
political, financial, or administrative realms. Elites can usually provide an
overall view of an organization or itsrelationship to oiher organizations. They
are more likely than other participants to be familiar with the legal and
financial structures of the organization. Elites are also able to report on an
organizations' policies, past histories, and future plans from a particular
perspective.
Elite interviewing also presents disadvantages. It is often difficult to gain
access to elites because rhey are usually somewhat elusive and busy people
operating under demanding time constraints; they are also often difficult to
Dara Colleciian Methods
'.
contact initially. The inierviewer may Iiave torely on sponsorship, recommendations, and introductions for assistance 1in making appointmenls with elite
individuals.
Another disadvantage in intsrviewng elites n that the interviewer m y
have to adapt the planned-for sbucture of the injerview, based on the wishes
and pradílections of the person interviewed. Although this is true with al1
in-depth interviewing, elite individuals are typically quite savvy and may
resentthe restrictions of n a r r w or 111-phrased questions. They may want an
active interplay with the interviewer. Well practiced at meeting the public and
being in conbol, an eliteperson may tum rhe interview around, thereby taking
charge of it. Elites respond well to inqniriec about hroad areas of content and
t~ a high proportion of inreiligent, provocative, open-ended questions that
allow thcm the freedom to use their knowledge and imagination.
Working with elitesplaces great demands on the abtlity of the intervicwer,
who must establish compaence by displaying a thorough knowledge ofthe
topic or, lacking such knowledge, by projecting an accurateconceptualization
of the problem through shrewd questioning. The intervicwer's hard wwk
usnally paya off, however, in the quality o$ information obtained. Elites often
conmbute insight and meatnng to the interview process hecause they are
intelligent and quick-thinking people. at bome in the realm of ideas, policies,
and generalizakcps.
ished because the participant had not reflected on the topic and feels unpre
pared lo respond. Often, the questions in a focus group settlng are deceptively
simple; the tnck is to promote the participants' expression of their views
through the creation of a supportive environment.
The advantages of focus group interviews are that this method is socially
oriented, studying participants in an atmosphere more natural than arüficial
experimental circumstances and more relaxed than the exposure of a one-toone interview. When combined with participant observation, focus groups are
especially useful for gaining access, focusing site selection and sampling, and
even for checking tentative conclusions (Morgan, 1997). The format allows
the facilitator ihe flexibility lo explore unanticipated issues as they arise in
the discussion. The results have high face validity: Because the rnethod is
readily understood, thefjndings appear believable. Furthermore, focus groups
are relatively low cost, they provide quick results, and they can increase the
sample size of qualitative studies by interviewing more people at one time
(Krueger, 1988).
There are, however, certain disadvantages lo this method as well: The
interviewer has less control over a group interview than an individual one,
s discussed;
which can result in lost time wbile dead-end or i r r e l e v a ~ s s u eare
the data are difíicult lo analyze, because context is essential to understanding
the participants' comments; the method requires the use of special room
arrangements and highly trained observer moderators; the groups can vary a
great deal and can be hard to assemble; and finally, there are logistical
prohlems arising fiom the need to manage a conversation while getting good
quality data.
i
i
Focus ~rouiIdterviewing
The method of inteiviewmg participants in focus gioups comes largely
from marketing rese& but has been wideiy adapred to social science and
applied research. The groups are generally compmed of 7 to 10 people
(although they range from as small as 4 to as large gs 12) who are unfamiliar
to one another and have been selected because they s h m certain characieristics relevant to the study's questions. The interviewer creates a supportive
environment, asking focused questions, to encourage discussion and the
expression of differing opinions rnd points of view. n e s e interviews may be
conducted several times With different individuals SO that the researcher can
identify trends in the perceptions and opinions expressed, which are revealed
throogh careful, systematic anaIysis (Krueger, 1988, p. 18). .
Thismethod assumes that an individual's attitudes and beliefs do notfonn
in a vacuum: People often need to listen to otherst opinions and understandings in order to farm their own. One-to-ow interviewsmay b e impwer-
,
ii
I n t e ~ i e w i n gChildren
e,.
:
Children may be the primary focus of a study or one of many groups the
researcher wants to interview. Their perspectives may bcsought because they
offer fresh insights. Inneasingly, there are calls for including children's
F'perspectivesas relevant and insightful in learning more about aspects of their
!*.
;: worlds. This is especially true in education where. al1 too often those most
. .
.:.. ~affected by educational policy and programm.atic decisions-the students;,-,are absent fiom inquiry. There are special considerations, however, when the
s.,..,
': qualitative researcher proposes a study that involves children.
? First are'age considerations. Interviewing preschoolers,,for example, is
7x~~ quite different from interviewing early adolescents. Both age gioups have
P
%L
~-:..-
g.-
..
Data Collection Me~hods
Vigncfle 22
Imagine 12 uniuersity students, on a chilly Saturday moming, sprawled
out on a classroom floor formulating their thoughts for a proposal on scattered sheets of newsprint Laughter, silence, ano intense discussion highIight the writing process of these authors who are first-generation refugee
and immigrant (newcomer) students from China, Cambodia, Vietnam,
Laos, and Korea participating in an undergraduate seminar on crosscultu~alexperiences in community development. After brainstorming various ideas, the studenrs collectively decided on a proposal incorporating
photography and arr as the vehicle to ponray the lived, and often misunderstood, experiences that they share with a group of 15 Vietnamese and
Khmer youth they have been working with.This proposal has lead to a
student-initiated and directed research project. A CollectiveVisual
Portrayal, funded by the University of Massachusetts Chancellor's
Counsel on Community, Diversity and Social Justice.
The organic and evolutionary educational process of CIRCLE over the
past 3 years has allowed me to conduct different levels of preliminary research. This dissenation builds on this research, exploring how newcomer
students view their educational experiences after having participated in a series of community development courses and outreach projects that validate
their cross-cultural Iives. Drawing on a range of qua;litative data sources
&d infonned by my comprehensive examinstion research, 1 also focus on
how newcomer students and the facilitators working with them define themselves as laaders, activists, and educators through their academic course
work and community involvement.
f i i s research acknowledges the real tensions that exist in any qualilative
research endeavor. Certain qualitative research models can b e rigid. oneway streets if they seduce participants into a process of inqoiry in which
the researcher alone is the analyzer and interpreter of data. This study consciously tried to counter such situations by applying participatory research
as &re guide of the inquiry (Mnguire, 1987; Reardon, Welsh, Kreiswinh, Br
Forester, 1993). Study participants have been involved in this inquiry as
"researchers" and valued members of a learning team in order to produce
knowledge that may help stimulate social change.
119
Stemm~ngfrom a commitment to participatory processes, the research 1
am conducting is collahorative in nature, emerging from the students and
the communities 1work with. Collaboration and participation in develqping
cntical learning en\ ironments produce pooled resources and shared expertise leading to inteprated and collective activities, Collahoraúon, action,
and reflection enhance the legitimacy of each participant's knowledge
(Brice Heath & McLaughlin, 1993) and set the stage for the sources of
multilevel data collection employed in this study.niesesix sources of data
have evolved in complement with the development of CIRCLE courses and
community ouueach activities and support the concept of a pedagogy for
affirmation, advocacy and action. They include the following:
1. Journal enuies and self-reflectionpapers: These s m as &edatnfor
textual narrative analysis in this dissenation reseash.
2. A focus group interview with 8 UndergraduatestudenU at the end of a
course where personal practice in community development is discussed
and course conteut and process are reflected on.
3. In-depth inteiviews with 10 shidents: These contrib te to ihe dissertation's
oral narrative analysis data collection process.
lt
d.Video and photography documentatian have been collaboratively collected
by stody participants (students) and researchers around classroom and
communiry events, evaluations, and siudent presentations: These form the
dissenation's visual narrative analysis.
5 . Oral history intcrviews conducted by students and youth with each other as
pan of A Collective Visual Portrayal project.
6 . Research field notes, reflections, and academic papers for courses and
conferenccsover ihe 4 years of my involvement with and participation in
CIRCLE: These provide critica1 insights into my own theoretical
development in relalion to this research and my role as researcher in tbis ,,
study.
!
. .
--
:;.';F ::4 ri.
,
.
&S,>>
:
:
. . .
,,
,
,Ir
* * * * *
,.
~haddudr-~eniande>s
(í997) discussion of the various sources of qualitative data-some generated as part of the CIRCLEproject, others tp be generated
specifically for the dissertation-is eloquently cóngnient with her assumptions
ahout the nature of this work, its purpose and audience, and her politicalstance.
Note thitsheplans torely on severa1methods: documentsin theform ofjoumals,
124
DESlGNlNG Q U A L I T A T I V E RESEARCH
of eyewitnesses, documents, records, and relics. Secondary sources include
the reports of persons who relate the accounts of actual eyewitnesses and
summaries, as in history books and encyclopedias.
The researcher should consider various sources of historical data: for
example, (a) contemporary records, including instmctions, stenographic records, business and legal papers, and personal notes and memos; (b) confidentia1 reports, including military records, journals and diaries, and personal
letters; (c) public reports, including newspaper reports and memoirs or autobiographies; (d) questionnaires; (e) govemment documents, includimg archives
and regulations; (f) opinions, including editorials, speeches. pamphlels, letters to the editor, and public opinion polls; (g) fiction, songs, and poetry; and
(h) folklore.
Historical analysis is particularly useful in obtaining knowledgc of previously unexamined areas and in reexamining questions for which answers
are notas definiteas desired. It allows for systematic and direct classification
of data. Histoncal research traditions demand procedures to verify the accuracy of statements about the past, to establish relationships, and to determine
the direction of cause-and-effect relationships. In fact, many research studies
have a historical base or context, so systematic histoncal analysis enhances
the vustworthiness and credibility of a study.
There is a dialectic tension in historical analysis between contemporary
social thought and that of the past; this dialectic may throw current thought
into reiief Sensitivity should be given to the interpretation of the statements
of others. Historical analysis cannot use a direct observation approach, and
there is no way to test a histoncal hypothesis. There are also weaknesses in
the classification of historical data. Documents may be falsified deliberately
on the part of the
or may have been subject to incorrect inte~~retations
recorder. Words and phrases used in old records rnay now have different
rneanings. The rneanings of artifacts are perceived and interpreted by the
investigator. EITO~S
in recording as well as frauds, hoaxes, and forgenes pose
problerns in dealing with the past. Thus, the researcher should retain a modest
skepticism about the da&.
F i l m Videos, and Photographs
Films and photographs havé a long history in anthropol&y. Called visual
anihropology orfilrn elhnography, this uadition relies on films and photographs to capture the daily life of the group under study. Films provide visual
Data Collection Methods
records of passing n a t u o events and may be used as permanent resources,
The concept and methed of theresearch film have emerged and are compatible
with a variety of research methods and have been used to describe how people
navigate in public places (Ryave & Schenkein, 1974) and the use of &ice
(Whyte, 1980), to present findings (lackson, 1978), and to empower participants (Ziller & Lewis. 1981).
Research filming is a powerful tool for inquiry into past events. Film has
the unique ability to capture visible phenomens seemingly objectively-yet
always from the perspective of the filmmaker. Research film -thodology
requires the documentation of the time, place, and sub~ectof the filming, as
well as the photographer's intent and interests. Also, a great wealtk of visual
informalion emanates from al1 natural events: To attempt a "comp1ete"record
of even a small event would be a fmitless pursuit.
There are three kinds of sampling in films: opportunity, programmed, and
digressive (Sorenson, 1968). Opportunity sampling documents unanticipated
or poorly understood phenomena as they occur. Programmed sampling involves filming according to a predetemined plan-deciding
* i
what,
1
where, and when to film. Grounded in the research proposal's
conceptual
framework, this sampling sirategy stipulates which events are ltkely to be
significant. Such íilming is guidcd by the research design rather than by
intuition, as in opportunity sampling. Digressive sampling 1s deliberate
searching beyond the obvious to the novel, to the places and events that are
usually outside typical public recognition.
Researchers choose to use ethnographic film for its obvious suengths. The
visual samples increase the value of any record. Film documints life frises
and ceremorlies, transmits cultural events to successive generations, and
documents social conflicts (court, speakers, Senate sessions, and so on). The
film researcher is limited only by what the mind can imagine and the camera
can record. And, of course, events can be documented in the natural setting.
Film is particularly valuable for discovery and validation. It docurnents
nonverbal behavior and communication, such as facial expressions, gestures,
and emotions. Film preserves activity and change in its original f o m . It can
be used in the future to take advantage of new methods of seeing, analyzing,
and understanding thepjocess of changs. Film is an aid to the researcher when
the nature of what is iought is known but the elements of it cannot be
discovered because of the limitations of the human eye. Film allows for the
preservation and study of datafrom nonrecurring, disappearing. or rare events.
With films, interpretation of inforrnation can be validated by another
128
DESIGNING QUALITATIVE RESEARCH
have been useful in the study of the behavior of students in the classroom and
of marital partners undergoing counseling!
There areseveral advantages to theuse of proxemics. It is unobwsive, and
usually it is diff~cultforasubjecttomisleadtheobserverdeliberately.Because
proxemics i s concerned with nonverbal behavior, subjects would bave 18 be
skillful to "lie" about their feelings. Proxemics Is useful for studying the way
individuals react to others regarding space and the invasion of their temtory.
Likewise, proxemics can be used in cross-cultural studies because people's
use of personal space varies greatly from one culture to the next. Finally,
proxemic analysis is useful for studies in areas such as the effect of seating
arrangements on student behavior or the effect of crowding on workplace
productivity.
The greatest disadvantage of proxemics as a data collecfion method is that
to gain accurate information the researcher must be skilled in the interpretation of the observed behaviors. If the researcher is observing a conference or
a business meeting, the manner in which the subjects take their se?ts can be
of vital importance, but the data must be interpreted carefully.,'Exclusive
reliance on proxemics could be misleading because it might sugg$t relationships that do not exist Because of the relative youth of proxemics as a data
collection method, few space measurement instmments are available in the
fieid of research, further limiting its diverse use. The use of proxemics is
increasing throughout research arenas It provides a revealing and interesting
method of gathering information about individual social behavior.
Data Collection Methods
1ZY
Museum of Science and Indushy had io be replaced every 6 weeks. Tbe tile
in other parts of the museum did not require replacement for years. The
selective erosion of the tiles, indexed by the replacement rate. provided a
measure of the relative popularity of exhibits Webb et al.. 1966).
Unobtmsive measures are particularly useful for tnangulation. As a supplement'to interviews, nonreactive research provides another perspective on
the phenomenon, elaborating its complexity. These methods can be used
without arousing notice from subjects, aud data collection is relatively easy
because it often involves using data (e.&, bills, archiva1 records, sales records)
already collected by someone else.
When used in isolatian, however, unobtnisive measures may distort the
picture. Erosion and survival may be affected by activities unknown to the
researcher. For example, tiles near the hatchmg-chick exhibit may wear out
because the exhibit is clase to tbe candy machine, not because of the exhibit's
popularity. Some researchers consider the use of unobtrusive methods (e.&
monitoring exchanges on newsgroups or searching through garbage to be
unethical: They feel that those studied should be informed of the nature of the
research.
When the researcher needs information on measures of frequency or
attendance, wheu direct obsewation would be impossible or would bias the
data, this method is useful. Unobtnisive data collection is often aided by
hardware such as audiotapes, hidden cameras, one-way mirrors, gauges, aud
infrared photos.
4
Unobtrusive Measures
Questionnaires and Surveys
Unobtrusivemeasures are methods for collection of datathat do not reqriire
the cooperation of the subjects and, in fact, may be "invisible" to them. Webb,
Campbell, Schwartz, and Sechrest (1966) describe these measures as "nonreactive research:' because the researcher is expected to observe or gather data
without interfering in fhe ongoing flow of everyday events. Data collected in
this manuer are categorized as documents, archiva1 records, and physical
evidence. Of the three.documents and archivalreccrds are themostfrequently
used in qualitative studies and were discussed earlier.
Physical evidence not produced specifically"'fur the purpose of research
often constitutes data; the following example pmvides an illustration. During
the 1960s, the floor lile around the hatching-chick exhibit at the Chicago
Researchers administer questionnaires to same sample of a population to
learn abtiut the distribution of charactenstics, attitudes, or beliefs. In deciding
to survey the group of people chosen for study, researchers make one critica1
assumption-that the characteristic or belief can be described or measured
accurately through self-report. In using questionnaires, researchers rely totally on the honesty and accuracy of participanls' responses. Although this
limits the usefulness of questionnaires in delvinginto tacit beliefs and deeply
held values, there are still many occasions when sumeying the group under
study can be useful.
Questionnaires typically entail severa1 questious that have smctured response categories and may include some that are open-ended. The questions
Dala
are examined jsomeumes quite vigorously) for bias, sequence, clarity, and
face validiry. Questionnaires are usually tested through administration ro i
E
small grcups l o detemine theií usefulness and, perhaps, reliability.
Sample surveys consist of the c~llectionof data in a standardized fomat, 4
uwally from a probability sample af the population. The survey is ihe
preferred methqd if the reSea~cherwishes to obtain a small amonnt of inforrnation from a large number of rubjeets.
Sumey research is the appropriate mode of inquiry for making inferences
about a large group OS peaple from data drawn cm arelatively small number ;
of individuals from that group. The basic aim of survey research is to describe i
and explain statistically the variability of certain features of apopulafion. n i e
generd logic of suwey research gives a distinctive style lo the research
process; the type of susvey instnrment 1s determined by the information
needed. There are three types of surveys. mail. telepljone, and personal
interview. Any method af data collection. h e v e r , from observation lo
content annlysis, can be and has been used in survey research.
Mosi survey studies involvecross-sectional measurements made at a single
point in ume or longitudinal measuremeuts taken at several different times.
Other f o m s of survey research inelude trend studies examining apopulation
by studying separate samples at different poinls in time, mhoft studies
d
examining a bounded population, and panel studies examining a slhglemmple
of individuals af several points in time Analysis of surveydata tikes the form
of quantitative analysis that relies mainly on erther descripbve br inferential
statistics.
The relative advantages and disadvantages of sumey research are highligbted through the following critena: (a) appropliateness of the method to
the problem studied, @) accuracy of measurement, (c) generalizability of the
findings. (d) adminiswive convenience, and (e) avoidance O£ ethical or political diffjeulties in the research process.
There are s o m definite advantages af survcys when the goal8 of research
require obtaining quantitative data on a certain problem or populafion. Surveys facrlitatereseareh in politically or ethically sensitive areas. They are used
in proyams for public welfare oreconemic development. Laege surveys often
focus on sensitiveorcontrovasial mpcswithin thep&EcdbmWn-.
Seengihs of s u w y s include their accuracy, generalizability, and convenience. A m r a c y in measnremept is enhanced by quantification, replicahility,
and control over observer effects. Survey results can be generglized to a larger
Colleciion Mcihods
3 .
k
" .
j 131
populatron within known limits of error. Surveys are amenable ro rapid
statistical analysis and are comparatively easy to administw and manage.
Surveys have weaknesses, however. For example, they are of little yalue
for examining complex social relationships or intricate pattems of interaction.
The strengths of surveys can also be weaknesses. Although controlling accuracy, a survey cannot assure without further evidenee that the sample represents a broader universe. Thus, the method of drawing the sample and the
sample size is critica1 to the accuracy of the study and its potential for
generalizability. Also, even though surveys are ccmvenient, they aregenerally
a relatively expensive method of data collection. Finally, surveys may result
in an invasion of privacy or produce questionable effects in the respondent or
thecommunity. Someresearchprojects may enhance theposition orresource~
of a particular group, and conflicts frequently arise between sponsors and
research teams concerning how problems are defined.
-
Projecrive Techniqrres ond
PsychoIogical Tesring
:
:3
$
Some types of interpretive psychological strategies were developed many
years ago by clinical psychologists lo o6rain personality data. These strategies
have been used fairly extensively in comparative studies about culture and for
analysis of personality dynamics. Based on an internal, pereeptual frame of
reference, the techniques assume that one can get a valid picture of a person
by assessing the way the individual projects his personality ontoasome standard, amhiguons stimuli.
Standardization and ambignity are common elements in tests of this nature,
judgments are the primary interpretation bases of realthough sdciiniEa~
sponses to thesestimuli. Results are typically expressed in thefom o f a verbal
report assessing the subject's dominant needs and ambitions, tolerance of
fmstrations. attitudes toward anthority, major intemal conficts, and so on.
The reputation and qualifications of the tester sometimes play a role in how
the report is received and how much credihility is attacbui lo the interpretation.
Two of the most well-known and frequently used psychological strategies
of this notion are the Rorschach inkblot test and the niematic Apperception
Test (TAT). The original idea behind both includes the assnmption that the
stimuli are ambigoous so that the subject has to be imaginative and "projective"
132
Data Collection Merhods
DESlGNlNG QUALlTATlVE RESEARCH
questions and in the particular setting for a given study. Tables 4.1 and 4.2
display the strengths and Iimitations of each data collection method, based on
how it is generally used in qualitative studies. The tables should help researchers to select the best combination of methods: Limitations in one
method can be compensáted for by the strengtlis of a complementary one.
In drafting the proposal, the researchershould consider whether themethod
will provide adequate information, he wst-effective, and be feasible in t e r m
of the subdeties of the setting and the resources available for the study. The
relative emphasis on participation in many qualitative studies. for example,
suggests certain methods over others. t u t z and Iannaccone (1969) provide
guidelines for method selection based on role, as shown in %ble 4.3. These
choices should be logically linked lo the conceptual framework and research
questions, the overall strategy of thestudy, and early decisions about role.
Vignette 23 describes how a researcher selected specific data collection
methods lo elicit information about a long-term heaith care facility.
in response to those stimuli. %e Rorschach test uses pictures (symmetncal
inkblots); usually presented in a. predetemined order, with the subject's
reporting what each picture resembles or suggests. The number, quality, and
variety of thesubject'sresponses Bre compared with specificpersonality types
and with pribr expenences with the responses of other people to the same
stimuli. In the TAT, the subject iZ asked to tell stories about a set of picture
scenes. Test results areused to assess personality traits such as aggressiveness,
dependence, and sexual conflicts.
Although projective insmments have been ihe object of considerable
criticism for many years, they are still commonly employed in clinical :
contexts by psychologists. Yet questions remain as to their validity and
reliability; environmental and cultural factors may aiso affect results. Today,.
concern focuses on more. concrete aspects of personality traits, such as
self-esteem and styles of interpersonal behavior, rather than on the vague ,
generalizations that charackrized earlier jnterpretive schemes.
Recently a number of other psychological tests and measurements have
been developed for use in qualitative and anthropological research. Examples
include the study of (a) the perception of illusions, which uses optical and
auditory illusions to examine differencei in perceptian related to differences
in types of environments; (b)judgments of aesthetic qualities, yhich rely on
pictures of artobjects or musicaistimuli to elicit opinions concerning aesthetic
excellence; (c) psychomotor skills, which use physicai activity measures to
indicate personality qualities such as introversion and extroversion; (d) games
people engage in. to provide significan1 information about coqmunity and
social behavior; and (e) games as a laboratory device, which uses a specific
game involving family members to determine arelationship between communication patterns.and socioeconomic differences. Various other qualitative
methods have been devised for studying entire commuoities, group living
patterns, and social integration of individuals in differentresidential contexts;
these are referenced at the end of this chapter.
Choosing Data Collection Methods
How might one's view of life be shaped by residence in a long-tem
health care facility? A doctoral student in health care management
(Kalnins, 1986) wanted to examine-in depth and detail-the contexts,
processes, and mteractions that shaped patients' perspectives She reasoned
that a qualitative approach would bc most fruitful in picking up everyday
actions and interactions about complex social strucuires.
Kalnins's major purpose in the study was to understand the meanings
given to events by participahts. She referred to the work of Schatzman and
Strauss (1973) by noting the foliowing:
:i /
Combining Data Conection Methods
Many qualitative studies combine several data collecuon methods over the
course of the study, as seen in Shadduck-Hernandez's (1997) proposal discussed in Vignette 22. The researcher can assess the strengths and limit$ions
of each method, then decide if that method will work with the particular
..
The researcher musf get elose to the peoplc whom he [sic] studies; he understands that their actions are best comprehended when observed on the spotin the natural, ongoing environment where they livé and work. The researcber
himself [sic] mua be at the location, not only to watch but also to listen to the
symboiic sounds that characterize this world. A dialogue with persons in their
natural situations will reveal the nuances of meaning from which their perspectives and definitions are continually forged. (Kalnins. 1986. pp. 5-6)
fmnrinurdp. 1361
D E S l G N l N G Q U A L l T A T l V E RESEARCH
136
.5
,
.
~ o l r
,
Mcrhod
137
.
Table 4 3 ~ d t CollectionMethods
a
Reiated to Observation Role
-
Data Collcct~onMcthods
~biervaiionand
rccording of
descripdv~data
Rccarding dinet
quotations of
surtimint
Unsvuctarcd
interview
1-Pdnicipanr
a
t Observer
+
Il-Obslrvcr
m finicipnr
'.
+
+
+
.---
-
+
P a R i c u l ~ yuschil to
~
+
*
e
+
-
~n&action
freanenc~tallies
+
+
'
n
S
?
"PY
+
i
.. Snme as ab-
~
+
-
Radio and
tclevirion rcports
+
.,
gunrded intcraction
wd scntimmt
Dcrailed
interartion pides
Writtcn ncords
Ncwspapcr
Officiai minutes
Lctten
Speechcr
I I I ~ ~ s I N ~ ~ , ~ ~
NonponicipMr .
Commcnr
, Role 1 in areas of
swctwed
interview guides
paper-and-pencil
reSS
Quesrionnains
Scales
Achievementor
nbility
"
+
+
t
S
*
1
lf &e researchcx is
sliillhil. a stnictun
ernerges
Most uschil in survcy
work (e.g.. ccnau)
h4oit ureful in smallgmup work
Mesni?gful in
lcadership studis
Very helpful in cenain
cireumstances Sor
ccnain purposes
Very imponant to
Rolel in chwkmg
relisbility of obscrvcd
data
Snme as abovc
..
SOURCE: Lue and lannaceanr (1969,p. 113). ñcp*nred wlUi pnniriion.
NOTE:+= likely $0bc uacd; * -msy oscsaionally kue-= difficulrnimpaniibls lo u r
,
1;
, From the variety of data collection strategies, she proposed a combination
of direct observation, patticipant observation, and semistructured interviewing. Her beginning point would be direct observation of residents and
staff in various areas of the facility, "witnessing events which particularly
preoccupied the hosts, or indicated special symbolic importante to them"
(Schatzman & Strauss, 1973, p. 59). This would allow her to get a holistic
view and to gather data that would inform the interview process.
Kalnins's plan as participant ohserver would be to observe the residents
and staff in the natural setting of the long-term healtb care facility, requiring her "commitment tu adopt the perspective of those studied by sharing
hn their day-lo-day expenences" (Denzin, 1970, p. 185). In her proposal,
Kalnins antieipated that participant observation and interviewing would Nn
concurrently, allowing data from each to be used to substantiate events, explore emerging hypotheses, and make furtherdectsions about the conduct
of the research. She referred to Becker and Geer (1969), atguing that participant observation allows the researcha to (a) check definitions of terms
the pariicipants use in the interview in a more natural setting (¡.e., casual
conversations with others); (b) observe events t$e participants cannot report
hecause they do not want to, feeling that to speak of some particular subject would be impolitic, impolite, or insensitive; and (c) observe situaúons
desaibed in interviews and thus become aware of diffenng pe pectives
presented by the participants (p. 326).
Her role as participmt observer would mean that Kalnins would become
immersed in the lives and activities of those she was studying. She understood the interactive-adaptive nature of participant observation, reflecting
the complex relationship hetween field observation and emerging theory,
and the impact of this relaáonship on decisions ahout funher data collection. Her decisions ahout the data to be collected and methods for collecting those data would be guided by Wilson's (1977, p. 255) list of five relevant types of data employed to get at meaning structures: (a) the form and
content of verbal interaction between pamcipants, (b) the fonn and content
of verbal interaction with researcher, (c) nonverbal behavior, (d) pattems of
actions and nonaction, and (e.) traces, archiva1 records, artifacts. and documents.
To generate facts, opinions, and insights (Yin, 1984), Kalnins planned for
open-ended stmctured interviews (using questionnaires) that would enable
the exploration of many topics but that could focus on cultural nuances,
firsthand encounters, and the perceptions, meanings, and inteipretations of
others. Information would also be gathered from various documents and archives, lending a historical perspective to the sntdy.
#
Data Collcclion Methodr
Table 4.4 Informalion Types and iífelhods o í Oblaining Information
Vignette 23 illustrates how a researcher chost an may of data collection
methods, know;ng thal each meihod had particular suengths and how each
wciuld help elicit WaU1d&rd irif&nmion. Pwtkipant obswvatim would
immerse he: in the lives of !he pstjents, giving her a personal perspective on
life in a long-tcrm heaith care facility. Observation would provide focused
data about sperifíc setcings find events wlthin the facihty, Indepth, semi9tnictured inreniiews would expbre the meaning of &en& an6 actions held by
panicipanis and ajbw her to t ~heri emerging intexptaions. One advantage .
of using mul~ipleqditative methods is zhe potential @ e v a k unwrpected ,
data. A m h e r i s the fundamenfal flexibility of qualít&ve methods: Shecoukd
modify herdata coUEftion mate@ hased on what she Iparned in the field. Thts
vignette show ttiar data cokkeitiiion strsngies and &ods cannot be c h w n
in a w u m . Inteasive ermihation o£ the possibb methods, Q i n g them out,
examining their potentials, and fiíí~ngthem totherwa& quation, site, and
sample are irnporfanl de& cmsi&rations. In addirim, researchad must
c o n s s d ~their own personal abilitiq in carrying out @y partrcniar overall
approaeh or method.
McrhodqfObraining Infimuii~on
7
Fnqucncy
d~s<Rbu~ah<
Incidcnts. histodes
bu1 fom>
Usnally inadcqvstc
and inefllciwt
itsclf:no1 cflln~nt
Pmolypeand
bcst fom
Irstitut~onsliud
Adequate but
Adcquatc but
n o m add siatu~s
,nefficiwn
mcffi~c~tnt,
except
foz unverbahd
norm
No( adquatc by
the use of represemtive samples. Then the researcher ~ o u l ddev~lopsur
and enumerate the findings. On the Mher hand, the f i n d k s migh\ be des(
tions. not numbers. If the research goal is description of processds, concl
categoties, and typologies. sampling and counting are merely tools 0f ar
sis, not necessarily paft of the research ñndings. The proposd alould C
onstrate that the researcher is capable of designing and selecting dafa Co
tion methods that are appropriate, well thought-out, and thorough BCC
theresearch question may change as theresearch progresses, t h e ~ e t h o d s
change and the researehet must ensure this flexibility. vignette 24 pro'
an example.
General Pnnciples for Designing
Data Collection Strategies
In the proposal, tha methods planneed for data coliection should be related
to the type af inf~qmaílonsought Zelditch's (1962) chart, reprodlrced in
Table 4.4, provides guidelinesfor thr& I a e e a ~ o r i e ofmethods:
s
enurneritting, participant Obsey~5tion.and in-depch intewewing. Haeh broad categwy
best yields a p~ticularlypeof infomation. In detemining wh'ih metbod to
use, t b e ~ e a r c h eshould
r
carefully examine the quwtimsguiding the srudy:
Many questions khat appear to be ''hbw" questions are redly "how many" .
c$uestions in disguise. For example, the selection of particípant observaúon to
help uncover hhow a prodevelopcd will no! adequately ~espoadto the
"how many" questions &a; might be embedded in it.
n i e researeher shmld delemine themost gractical, efiiiwt, feasible, and
etbical ni&a& for eollekting&ta as C h e t e s m h progmsea IEb might start
with píirtidpant obsemtion as he seeh Fo.ideRtify
p a W ~ 6and
,
. qwxtians,
.
domains. n i i s suategy may change as theresearch b e c h e s more focused and
progresses toward more specific questions and clearer concepis that suggest
Pmtatypc aod
'
A graduate student wanted to explore the'implementation of state 1
date for local school councils. Rodriguez first proposed participant obs
tion of meetings and in-depth i n t e ~ e w with
s board members.The da*
collectionplan showed a schedule for obseming tbe meetings, 608.1s fo
interviewing, and a time ailowance for analysis of data and for follow-1
r"
DESlGNlNG
QUALITATIVE RESEARCH
data collection. But in the process of initial data collection and preliminary
analysis, h e discovered that teacher resentment of the councils was creating
a pattern of unintended negative consequences. This discovery could have
importa& implications for policy development. Did Rodriguez have to stay
with the original question and data collection plan? Wouldn't a design
alteration offer important insighfs?
Rodriguez reasoned that if he could describe the processes whereby wellintended policy is thwarted, policymakers could gain insight that might
help them make timely alterations in policy development or implementation. Given this possible benefit to the study, he could choose to focus subsequent data collection on the conflicts hetween t&cher needs and the mandate to school boards that they implement councils. This would require him
to tum to additional literatures on,for example, teacher needs, teacher participation in decision making, or teacher unions. He might also need to empl6y additional data collection methods (such as surveying teacher needs,
obseming teacher union meetings, and doing historicalresearch on the reactions of teacher lobbies to mandates for school councils), or he might need
to sample additional settings or people. As the research question became
more focused, his initial research design and data collection strategy would
most likely undergo some changes.
Dnra Collcction Methods
how the data will be managed and stored, and provide some preliminary
discussion of the processes for analyzing those data. We discuss these issues
in !he next chapter.
Note
l. This vignene ir fictitiour.
Further Reading
PorhCipont Observation
Bogdan. R. C.. & Bikkn, S. K. (1998). Qualirariva rclcorch in rducotion: An inrn,dueri,m lo
rhrory ond meihods (3rd ed,).Bosran: Allyn & Bacon.
Delnmont. S. (1992). Fieldwork in educorimol scilings: Methods. pigüllr. nndprr~ecti~~es
Landon: Falmer.
Jargcnsm. D. L. (1989).Ponicipunl r~bservniion:A m~rhodologyfrirh u m srudicr. ~nvbury
Park, CA: Sage.
Pelro. P. &Pella. G. H. (1978).Anrhn>polosicalreseurch; The rrrucmre of: uz>y (2nd cd.).
'
N+ York: Cambndge Univcrsiíy Prcss.
Spradiey. J. S. (1980).Parricipant <ibsrrvuiim.New York: Holt. Rinchan & Winston.
~ o l c .&. t tF., ~(1995).
.
TheUn offieldwork. Walnut Cceek. CA: AltaMira.
*
Obsewntion
Adler. P. A.. 8: Adler. P. (1994).Obscrvational techniques. ION.K.Denzin&Y.S.Lincaln(Eds.),
In the examplein Vignette24, the resezrch proposal probably did not include
a plan for analysis of lobbying efforts or observation of collective bargaining
sessions. It would, however, be entirely appropnate-indeed recommendedfor the researcher to modify the research proposal if an exciting and significant
focus emerges from early data collection. In fact, the pnmary strength of the
qualitative approach is this very flexibility that allows, even encourages, exploration, discovery, and mativity.
Along with choosing appropriate strategies for data collection, the researcher must address the complex processes of managing, recording, and
analyzing data. Rather than discrete, sequential events, these processes oecur
dialectically throughout the conduct of a qualitative study: Analysis occurs as
themes are identifíed, as the deeper sttuctures of the social setting become
clear, and as consequent modifícations are made in the initial design. At the
proposal stage, however. the researcher should presentsome initial ideas about
Hwidbor>lr <ifyuuiiforiverrscardi (pp. 377-392).Thousand Oak. CA: Sage.
Eveifson, C.. & Gieen. J. (1983. Observarion as inquiry and mthod. In M. C. Wituaek (Ed.).
Hundbr~okrifre$rrirchm Icriching (pp. 162.223). New York: Mscmillan.
Laflnnd.1.. &Lofland.L.H.(1995).Anliiyuñ~rociulse~iinss:A~uideroquu~irarivrobsrrv~i
und nnolysir (3rd ed.). Belmont. CA: Wndswoirh.
Smith. C. D., & Karnbium. W. (Eds.).(1996). I n rhe ficld: Rcadingr on rhcfirld ressnrch
expenerice. WwtwR. CT: Racga.
Generic In-Drpth I n t e ~ i e w i n g
Fontana. A,. & Frey. J. H. (1994).Intemievring: The ort of scienoe. In N. K. Denrin & Y. S.
Lincoln (Eds.),Hmdbook of quulirative rerrarch (pp. 361-376).Thousmd Oaks. CA: Sage.
Holrtein, J. A.. & Gubrium, J. F. (1995). The active interview. Thousmd Oaks. CA: Ssgc.
McCmken. G. (1988).The lnng inuwiew. Newbury Park CA: Sage.
Parton, M. Q. (1990). QualitoriYe resanrch clnd evuiuation mthodt (2nd ed.). Newbury Pad;,
CA: Sage.
Rubin. H. 1 ,& Rubin. E. S. (1995).Qudiintive inlerviewing: Tha orrofhsonng dota Thousand
Onks. CA: S q e .
Tripp, D. H. (1983). Co-authorshipand negorialion: The intcMew ns a l of cruition. Intcr-
-
chun~a.14,3245.
Weirs. R.
.
S. (1994). Lcurningfrom
,studies. New York: F m Prcrs.
rrrengarr: The
un and method of qualirativr intorview