"•'i
iJfl
7 .
T
Data Collection
Data collection offers one wore instance for assessing research design within each tradition of inquiry. However, l>efbre exploring this point, I find it useful to visualize the phases of data collection common to all traditions. A "circle" of interrelated activities best displays this process, a process of engaging in activities that include but go beyond collecting data.
I begin this chapter by presenting this circle of activities, briefly introducing each activity. These activities are locating a site or individual, gaining access and making rapport, sampling purposefully, collecting data, recording information, exploring field issues, and storing data. Then I explore how these activities vary by tradition of inquiry, advance ii table that summarizes these differences, and end with a few summary comments about comparing tlie data collection activities across the five traditions.
Questions for Discussion
ľ How might the data collection process and the activities in the process be visualized?
V  What are typical access and rapport issues in each tradition?
V  How docs one select people or places to study in each tradition? ▼ What type of information typically is collected in each tradition? T How is information recorded in each tradition?
T 109
.(Hit
no t
QUALITATIVE  INQUIRY
Figure 7.1     Dolo Collodion Activities
▼ What are common issues in collecting data in each tradition?
T How is information typically stored in each tradition?
T How do the íive traditions differ in the activities of data collection?
A DATA COLLECTION CIRCLE
I visualize data collection as a series of interrelated activities aimed at gathering good information to answer emerging research questions. As shown in Figure 7.1, a qualitative researcher engages in a series of activities in the process of collecting data. Although I start with locating a site or an individual to study, an investigator may begin at another entry point in the circle. Most important, I want the researcher to consider the multiple phases in collecting data, phases that extend beyond the typical reference point of conducting interviews or making observations.
An important step in the process is to find people or places to study and to gain access and establish rapport so that participants will provide good data. A closely interrelated step in the process involves determining a strategy for the purposeful sampling of individuals or sites. This is not a probability sampling so that statistical inferences can be made; rather, it is sampling so that one can best study the
Id
Data Collection                                                                             ▼ ill
problem under examination. The researcher needs to determine the
[type of purposeful sampling from the array or possibilities and present a rationale for the selected approach.
Once the investigator selects the sites or people, decisions need to be made about the most appropriate data collecnon approaches. Increasingly, a qualitative inquirer faces newer and more innovative approaches such as e-mail messages, and usually a study involves more than a single source o£ data. To collect this information, the researcher develops protocols or written forms for recording the information and needs to assess the logistics of this recording process. Also, noting and being aware of potentially difficult field issues that may compromise the data, lead to premature exit from the field or site, and/or contribute to lost information is an important consideration. Finally, an investigator must decide how he or she will store data to find them easily and to protect them from damage or loss.
I now turn to each of these data collection activities, and I address each for general procedures and approaches within each tradition of inquiry. As shown in Table 7.1, these activities are both different and similar across the five traditions of inquiry.
THE SITE OR INDIVIDUAL
In a biographical study, one needs to find an individual to study, an individual who is accessible, willing to provide information, and distinctive for her or his accomplishments and ordinariness or who sheds light on a specific phenomenon or issue being explored. Plummer (1983) recommends two sources of subjects to study. The pragmatic approach is where an individual is met on a chance encounter, a subject of interest emerges from a wider study, or an,-individual volunteers. Alternatively, one might identify a "marginal person" who lives in conflicting cultures, a "great person" who impacts the age in which he or she lives, or an "ordinary person" who provides an example of a large population.
In a phenomenologkal study, the participants may be located at a single site, although they need not be. Most important, they must be individuals who have experienced the phenomenon being explored and can articulate their conscious experiences. Likewise, in a grounded theory study, the individuals may not be located at a single site; in fact.
'HI.
kil
-I rn
S
▼   112
U-

111

l
1
I
r I
ii-i
ŕ i
o S o
íl
■ a
.n.I,
...
II   í
114 ▼
QUALITATIVE INQUIRY
if they are dispersed, then they can provide important contextual information useful in the axial coding phase or research. They need to be individuals who have taken an action or participated in a process that is central to the grounded theory study. For example, in Creswell and Brown (1992), we interview 32 department chairpersons located across the United Slates.
In an ethnographic study, however, a single site is important where an intact culture-sharing group has developed shared values, beliefs, and assumptions. The researcher needs to select a group (or an individual or individuals representative of a group) to study, preferably one to which the investigator is a "stranger" (Agar, 1986) and can gain access. For a case study, the site (or sites) also is important, but it is much more circumscribed than an entire cultural system in an ethnography. These sites may be programs, events, processes, activities, or multiple individuals. Although Stake (1995) refers to an individual as an appropriate "case," I turn to the biographical approach or the life history approach in studying a single individual. However, the study of multiple individuals, each defined as a case and considered a collective case study, seems acceptable practice.
I need to register a cautionary note, as do ethnographers such as Glesne and Peshkin (1992), ubout studying a site or people in whom one has a vested interest Glesne and Peshkin question research that examines "your own backyard—within your own institution or agency, or among friends or colleagues" (p. 21; emphasis added). This form of qualitative research attracts many students of qualitative research because studying one's own backyard provides easy access to informants and information at rninimal cost. Undoubtedly, quali-, tative researchers bring their values, biases, and understandings to a project, and intimate knowledge of a setting may be an asset. But the negatives outweigh the positives. Studying such people or sites establishes expectations for data collection that may severely compromise the value of the data; individuals might withhold information, slant information toward what they want the researcher to hear, or provide "dangerous knowledge" that is political and risky for an "inside" investigator (Glesne & Peshkin, 1992). Also, especially in ethnographic research, the investigator tracks norms and values of which participants in the culture may not be aware; being an insider may not yield this information. Unless a compelling argument can be made for
-K»
Data Collection
▼ 115
or she)... i, deeply uwolved as a participant" (p. n 7).
ACCESS AND RAPPORT
Gaining access to the site or individual^ also involves several steps Regardless of the edition of inquiry, permissions need o Ä,
coTrn ,                  ,eCte reVÍeW bo"* a P'0MSS - -hich a campu
cor™meerev,ewsresearchs.udiesfor(neirpolenHalha^1^PUc
„n íŤ!"*» 'ľ PTrntSÍ ThÍS Process <™»«* s"bmi„ingľpľ„
quahtahve stud.es are exempt from a lengthy «view (eL the expe d.tcd orfuUrev.ew,bu,studiesinvolvingUviduaisasminorsTe 1«years or under   or studies of sensitive population, (e.g., «£ pos, ,ve md,v,duals) «quire the expedited or full review a proceľs
renv™wnL   ,i,Cd' len8",r aPP'ÍCa'i0nS - - «P^ tfmľZ
ZfLfeľ     T *"*" b0ardS are more to»'"» "» 'he
fumfttta» approaches to social and human science research than
*ey are to „native approaches, the qualitative project desrtphon may need to conform to standard procedures JľÄ
Hoľ ľľ^(e*' hyPOtheSe5' "** »«*•> - weTafiZma-hon about the protect™ of human subjects. As shown in Figure 72
SSiSSÍ" **parlioipants compte fa a ***
- Their right to voluntarily withdraw from the study at any Hrne
■si::ionrposconhesmdyand,he^duresiob=-^«
■   Comments about protecting the confidentiality of the respondents.
■   ^statement aboutknownrisksassociated with participation in the
iQUi
t
.„, _                                                                  QUALITATIVO INQUIRY
"Experience» in loornlng. Quolltotivo Research: A Quolllalívo Cose Study"
THo following information n provided lor you 'o decide whether you wish lo participate in
ď« präsent iiudy. Vou should bo aware that you oro free lo decide not <o participate or to
withdraw ot any tima without ofleding your relationship with this department, the ittvtruclor,
of the Unlvôriŕty ol Nebroska-Uneoln.
Th« puiposq of this study ii to understand the process of learning qualitative research in o
dodorol-level college course. The procedure '«ill be a single, holistic cose study design. At
this, stage in tho research, process "ill be generally defined os perceptions ol the course
and moking tense cut of qualitative research ol different phases in the course.
Oato will be collected at three points—at the beginning of the course, at the midpoint, and
at the end of the course. Dalo collection will involve documents (journal entries made by
studenti and the instructor, student «valuations of the dass and the research procedure).
oudio-visuol maleriol (a videotape ol the doss), interviews (transcripts ol interviews, between
students), und clostroom observation field notes (made by students and the instructor).
Individuals involved in the doto collection will be the Instructor and tho students in the class.
Do not heiilale to osk ony questions about tho study either before participating or during tho lime thijt you arrj porticlftating. We would be happy to shorn our lindlngi with you oltor thfl research is completed. However, your name «ill no' be oisocioled with the research findings in any way, and your identity 0s a participant will bo known only to the roieorchors.
There ce no known risks and/or discomforts ossociatsd with this study.
The expected benefits associated with your participation ore the information about the
experiences in learning qualitative research, the opportunity to participate in a qualitative
research study, and co-authorship for those students who portidpate in tho detailed analysis
of the doto. II submitted for publication, o byline will indicate the. portidpation of oil students
h the class.
Please sign your consent with full knowledge ol tho nature ond purpose of the procedures.
A copy of this content form will bo given lo you to keep.
Signature ol Participant                                                                   a"                          lľ,
John W. Creswell, Ed. Psy., UNI, Principal Investigator 402-477-2248
Figure 7.2    Sample Human Sub'icds Consent to Participate Form
■   The expected benefits to accrue to the participants in the study.
■   A place for them to sign and date the form (a place for the researcher to sign and date also may be offered).
-If,
Data Collection
▼   117
For a biographical study, inquirers gain information from individuals by obtaining their permission to participate in the study. Study subjects should be appraised of the motivation of the researcher for their selection, granted anonymity (if they desire it), and told by the researcher about the purpose of the study. This'tlisclosure helps build rapport. Access to biographical documents and archives requires permission and perhaps travel to distant libraries.
In a plietiomenological study, the access issue is limited to finding individuals who have experienced the phenomenon and gaining their written permission to be studied. In Riemen'? (1986) study, for example, she found 10 nonhospitalized adults over 18 years of age who had prior interactions with registered nurses and who could articulate their experiences. Because of the in-depth nature of extensive and multiple Interviews with participants, it is convenient for the researcher lo obtain people who are easily accessible.
The access through individual permission to be studied also is part of a grounded theory study. This group needs to provide permission to be studied and needs to have rapport with the researcher to disclose detailed perspectives about responding to an action or process. The grounded theorist starts with a homogeneous sample, individuals who have commonly experienced the action or process. In an ethnography, access typically begins with a "gatckecpor," an individual who is a member of or has insider status with a cultural group. This gatekeeper is the initial contact for the researcher and leads the researcher to other informants (Hammerslcy & Atkinson, 1995). Approaching this gatekeeper and the cultural system slowly is sage advice for "strangers" studying the culture. Gaining access through the gatekeeper and establishing rapport with the case being studied also are important for a case study. For both ethnographies and case studies, gatekeepers require information about the studies that includes the following points (usually submitted in writing, as Bogdan and Biklen (1992) suggest:                                               •
■   Why was the site chosen for study?
■   What will be done at the site during the research study? (time and resources required by participants and amount of time to be spent at the site by the researcher)
■   Will the researcher's presence be disruptive?
.on*
i
118 ▼
QUALITATIVE  INQUIRY
■   How will the results be reported?
■   What will the gatekeeper gain from the study?
PURPOSEFUL SAMPLING STRATEGY
The purposeful selection of participants represents a key decision point in a qualitative study. Researchers designing qualitative studies need clear criteria in mind and need to provide rationales for their decisions. 1 recommend that qualitative researchers, regardless of tradition, examine the typology of 16 strategies for purposeful sampling advanced by Miles and Huberman (1994). As shown in Figure 7.3, the authors identify the purpose for each sampling option. I especially like when writers can identify their specific strategies, offer definitions for them, and provide brief rationales for their use. 1 illustrate this approach for each of the five traditions.
In a biographical study, the individual may be "convenient" to study because she or he is available, may be a "politically important case" who attracts attention or is marginalized, or may be a "typical" case, the ordinary person. Inquirers may select several options depending on whether the person is marginal, great, or ordinary (Plummer, 1983). Vbnnie Lee, who consented to participate and provide insightful information about the mentally retarded (Angrosino, 1994), was convenient to study but also was a "critical case" who permitted generalization and application to individuals with mental retardation.
1 find, however, a much more narrow range of sampling strategies for a phenamenotogical study. It is essential that all participants expert ence the phenomenon being studied. "Criterion" sampling works well when all individuals studied represent people who have experienced Uie phenomenon. All individuals meet this criterion. For a grounded theory study, the investigator chooses participants based on their ability to contribute to an evolving theory. Using the terms of Miles and Huberman (1994), this process is "theory based," but in grounded theory the term is ffteoretital sampling, which means that the investigator examines individuals who can contribute to the evolving theory. This begins with selecting and studying a homogeneous sample of individuals (e.g., all women who have experienced childhood abuse) and then, after developing the theory, selecting and studying a heterogeneous sample (e.g., types of support groups odier than
-If.
Data Collection                                                                             ▼ 119
Typo of Sampling	Purpose
Maximum voriallon	Documonli diverse variations and identifies important common pott ems
Homogeneous	Focuses, reduces, simplifies, and facilitated g roup interviewing
Critical case	Permits logical generalization and maximum application of information to other coses
Theory based	Find examples of a theorelicol construct and thereby eleboľote on and «»amine it
Confirming and discontinuing eases	Elaborate on initial analysis, seek exceptions, looking for variation
Snowball or chain	Identifies cases ol interest from people who know people who know what cases are information .rich
Extreme or devlonl COM»	learn from highly unusual manifestations of the phenomenon of Interest
Typical coso	Highlight* «hol Is normal or overoge
! dlon sily	Information-rich coses that manifest Ihe phenomenon intensely but not extremely
Politically important casts	Attracts desired ohention or avoids attracting undesired attenlion
Random purposeful	Adds credibility to sample when potential purposeful sample is too largo
Stratified purposeful	Illustrates subgroups ond facilitates comparisons
Criterion	All coses Ihot moot some criterion; useful for quality asiuionce
Opportunistic	Follow new leads; loklng advantage of the unexpectecd
Combination or m «od	Trkwgulotlon. llexlbllity; meels multiple inleroils and no od s
Convenienco	Saves lime, money, and effort, but at the expense of information ond credibility
Figure 7.3    Typology of Sampling Strategies in Guolitolive Inquiry SOURCE: Miles and Huberman (1994. p. 28). Reprinted with permission trom Miles, M. 8., & Huberman, A. M. (1994). Guo/<íoíŕvc dofo analysis: A sourcebook o( new methods {2nd ed.). Thousand Oaks, CA: Sage.
women who have experienced childhood abuse). The rationale for studying this heterogeneous sample is to confirm or disconfirm the conditions, both contextual and intervening, under which the model holds (see Creswell & Urbom, 1997, for this procedure).
.on.
i
67
120 ▼
QUALITATIVE INQUIRY
In an ethnography, once the investigator selects a site with a cultural group, the next decision is who and what will be studied. Thus, within-culture sampling proceeds, and several authors offer suggestions for this procedure. Fetterman (1989) recommends proceeding with the "big net approach" (p. 42), where at first the researcher mingles with everyone. Ethnographers rely on their judgment to select members of the subculture or unit based on their research questions. They take advantage of opportunities ("opportunistic" sampling [Miles & Huberman, 1994)) or establish criteria for studying select individuals (criterion sampling). The criteria for selecting who and what to study for Hammersley and Atkinson (1993) are based on gaining some perspective on time in the social life of the group, people representative of the culture-sharing group in terms of demographics, and the contexts that lead to different forms of behavior.
In a case study, the investigator might consider any one of the strategies for sampling identified by Miles and Huberman (1994)-1 prefer tů select unusual cases in collective case studies and employ "maximum variation" as a strategy to represent diverse cases to fully display multiple perspectives about the cases. "Extreme and deviant cases" may comprise my collective case study, such as our study of the campus response to the gunman (Asmussen & Creswell, 1995).
FORMS OF DATA
Although approaches to data collection continually expand in the qualitative area (see Creswell, 1994), there are four basic types of i information to collect: observations (ranging from nonparticipant to participant), interviews (ranging from aemistructured to open-ended), documents (ranging f torn private to public), and audio-visual materials (including materials such as photographs, compact disks, and videotapes). In Figure 7.4,1 present a compendium of approaches under these four categories. In recent years, new forms of data have emerged such as journaling in narrative story writing, using text from e-mail messages, and observing through videotapes and photographs. I encourage individuals designing qualitative projects to pursue information from sources generally unfamiliar to the reader. For example, I like the technique of "photo elicitation" in which participants are shown pictures (their own or those taken by the researcher)
-»/»
Data Collection
▼ 121
Observotions;
•  Gcther ticlcno'es by conducting on observation os o pantopon!.
•  Gather ficldnotcs by conducting on observation os on observer.
" Gcther ficldnotes by spending more time cs o participant ihon ps on observer.
•  Gcther tieldnotes by spending more lime cs en observe than cs c participant.
•  Gcthef ficldnotes first by observing as on "outsider* and then moving into the setting and observing as on 'insider*'
interviews;
•   Conduct an unstructured, cpen-ended interview and take interview notes.
•   Conduct en unstructured, cpen-ended interview, audiotape the interview, ond transcribe the interview.
•   Conduct a semistructured interview, audiotape the interview, ond transcribe the interview.
•   Conduct a focus group interview, audiotape the interview, and transcribe the interview. Documents:
•   Keep a Journal during the research study.
•   Have a participant keep a journal or diary during the resecrch study.
•   Collect personal letters from participants.
•  Analyze public documents [e.g., official memos, minutes, records, archival moteriol).
•   Examine autobiographies end biographies.
•   Hove infofmonts take photographs or videotapes {i.e., photo elkitotion). Audio-w'sfof rooŕeríofs.'
•   Examine physical traco evidence (e.g., footprints in the snow).
•  Videotape or film a social situation or an individual/group.
•   Examine photograph* or videotapes.
•   Collect sounds {e.g., musical sounds, a child's laughter, car horns honking).
•   Collect e-mail or electronic messages.
•   Examine possessions or ritual objects.
Figure 7.4    A Compendium ot-Data Collection Approaches in Qualitative Research
and asked by the researcher to discuss the contents of the pictures (Denzin & Lincoln, 1994). Ziller (1990), for example, handed a Polaroid camera, loaded with film, to each of 40 male and 40 female 4th graders in Florida and West Germany and asked them to take pictures of images that represented war and peace.
The tradition of inquiry directs an investigator's attention toward preferred approaches to data collection, although these approaches are not rigid guidelines. For a biography, for example, the portrait of
!»!•
122 T
QUALITATIVE INQUIRY
an individual's life is created from documents, interviews, and perhaps observations. For a life history of a living individual, Plummer (1983) discusses having the individual write down his or her history, a form of joumaling. Also, the researcher conducts unstructured, open-ended interviews, with a "mixture of participant observation and almost casual chatting with notes taken" (p. 95). The cornerstone for life history writing for Plummer, however, is the open-ended interview.
For a phenomenological study, the process of collecting information involves primarily in-depth interviews (see, e.g., the discussion about the long interview in McCracken, 1988) with as many as 10 individuals. 1 have seen the number of interviewees referenced in studies range from 1 pukes, 1984) up to 325 (Polkinghome, 1989). Dukes (1984) recommends studying 3 to 10 subjects, and the Riemen (3986) study included 10. The important point is to describe the meaning of a small number of individuals who have experienced the phenomenon. With an in-depth interview lasting as long as 2 hours (Polkinghome, 1989), 10 subjects in a study represents a reasonable size. Added to 10 in-depth interviews might be the self-reflection of the researcher as a preparatory step to interviewing (Polkinghome, 1989) or as the initial step in the analysis (Moustakas, 1994). Besides interviewing and self-reflection, Polkinghome (1989) adds gathering information from depictions of the experience outside the context of the research projects such as descriptions drawn from novelists, poets, painters, and choreographers.
Interviews play a central role in the data collection in a grounded theory study. With the intent of developing a model or theory and saturating categories, I recommend that a grounded theorist interview 20 to 30 individuals. In our study of academic chairpersons (Creswell & Brown, 1992), each of our interviews with 33 individuals lasted approximately 1 hour. Other data forms besides interviewing, such as participant observation, researcher reflection or journaling (memoing), participant journaling, and focus groups, may be used to help develop the theory (see Morrow & Smith's [1995] use of these forms in their study of women's childhood abuse). However, in my experience, these multiple data forms play a secondary role to interviewing in grounded theory studies.
In an ethnographic study, the investigator collects descriptions of behavior through observations, interviewing, documents, and arti-
-ÍA
Data Collection
▼ 123
facts (Hammersley & Atkinson, 1995; Spradley, 1980), although observing and interviewing appear to be the most popular forms of data collection. Partidpant Observation, for example, offers possibilities for the researcher on a continuum from being a complete outsider to being a complete insider (Jorgensen, 1989). The approaclvof changing roles from that of an outsider to an insider through the course of the ethnographic study is well documented in field research (Jorgensen, 1989). Wolcott's (1994b) study of the Principal Selection Committee illustrates an outsider perspective as he observed and recorded events in the process without becoming an active participant in the committee's conversations and activities.
A case study involves the widest array of data collection as the researcher attempts to build an in-depth picture of the case. I am reminded of the multiple forms of data collection recommended by Yin (1989) in his book about case studies. He refers to six forms: documents, archival records, interviews, direct observation, participant observation, and physical artifacts.
Because of the extensive data collection in our gunman case study (Asmussen & Creswell, 1995).. we presented a matrix of information sources for the reader. This matrix contains four types of data (interviews, observations, documents, and audio-visual materials) for the columns and specific forms of information {e.g., students at large, central administration) in the rows. Our intent was to convey through this matrix the depth and multiple forms of data collection, thus suggesting the complexity of our case. The use of a matrix, especially applicable in an information-rich case study, might serve the inquirer equally well in all traditions of inquiry.
In an examination of the data collection forms of the five traditions in Table 7.1, interviewing and observing are central to all traditions and deserve additional attention. Because of the extensive discussions of these topics in the literature-1 highlight only basic procedures that I recommend to prospective interviewers and observers^
Interviewing
One might view interviewing as a series of steps in a procedure:
▼   Identify interviewees based on one of 'the purposeful sampling procedures mentioned in the preceding (Miles & Huberman, 1994).
.i)n-
i
124 ▼
QUALITATIV!!   INQUIRY
T Determine what type of mtervievt is practical and will net the most useful information lo answer research questions. Assess the types available such as a telephone interview, a focus group interview, or a one-on-one interview. A telephone interview provides the best source of information when the researcher does not have direct access to individuals. The drawbacks of this approach are thai the researcher cannot see the informal communication, and the phone expenses. Focus groups are advantageous when the mir i.h lit ni among inter\ lewees will likely yield ihn* It-.I information, when Interviewees aiv similar ami eunperative wtlh each oil» i. whan time to collect information Is limited, anil when individuals interviewed one on one may he hesitant to provide information (Krueger. 1994; Morgan, 1988; Stewart tfe Sham-dasani, 1990). With this approach, however, care must be taken to encourage all participants to talk and to monitor individuals who may dominate the conversation. For one-on-one interviewing, the researcher needs individuals who are not hesitant to speak and share ideas and needs to determine a setting in which Ihis is possible. The less articulate, shy interviewee may present the n -..in Int with a challenge and less than adequate data.
▼    Whether conducting one-on-one or focus group interviews, I recommend the use of adequate recording procedures, such as a lapel mike for both the interviewer and interviewee or an adequate mike sensitive to the acoustics of the room.
T Design the interview protocol, a form about four or five pages in length, with approximately five open-ended questions and ample space between the questions to write responses to the interviewee's comments.
▼    Determine the place for conducting the interview. Find, if parable, a quiet location free from distractions. Ascertain that the plivn a I setting lends itself toaudiotaping, an essential necessity, I believe, in accurately recording information.
T    After arriving at the interview site, obtain consent from the interviewee
to jwticipate in the study. Have the interviewee complete a consent form for the human relations review board. Go over the purpose <»t the '.tudy, the amount of time that will be needed lo complete
Ihe interview, and plans fur using the remills from the interview (of lei .» ropy "I the repnrl or .in .ibstr.u I ol it to Ihe inlei\ leWee)
I-,
Data Collection
V 125
T During Ihe interview, stick to the questions, complete within the time specified Of possible), be respectful andcourteous, and offerfcxo questions and advice. This last point may be Ihe mosl important, and I am reminded how a good interviewer is ,i listener rather than a speaker during an interview. Also, leiord information ort the interview protocol in the event that the audio-recording does not work. Recognize that quickly inscribed notes may be incomplete and partial because of the difficulty of asking questions and writing answers at the same time.
Observing
Observing in a setting is a special skill that requires management of issues such as the polcnlial deception of tine people being interviewed. Impression management, and the potential marginaliiy of the researcher in a strange setting (Hammersley It Atkinson, 1995). Like interviewing, I also see observing as a series or steps:
f Select a site to be observed. Obtain the required permissions needed lo git In access to the site.
Y    At the site, identify who or iľ/m/ to obferve, wtien, and for how long. A gattknpir or key informants help in this process.
V    Determine, initially, a role as an observer. This role can range from thai ol .i complete partii ip.int (going native) to thai of ai nrnplete observer. I especially like the procedure of being an outsider initially followed by bwcuming an insider over rime.
f Design an observational protocol as a method for recordvig note* in the field Include in this protocol both descriptive and reflective notes (i.e., notes about your experiences, hunch«*, and learnings).
T   Record aspects such ai portraits of tlie informant, the physical setting,
particular events and activities, and your own reactions (Bogdan & BikltH, Í989).
▼   During the observation, have someone introduce you if you are an out/iiiler. be passive ami friendly, and start with limited objectives in the first few sessions of observation. The early observational sessions .uti.
QUALITATIV!*  INQUIRY
126 T
may bo times in which lo like few notes and confine attention lo
observing.
After observing, slowly withdruwfrom the si«, thanking the fXfUci-
Lts and informing them of the use oftl,e data and their accessibility to
puríís íitMÍ ŕ'i/< Die study,
RECORDING PROCEDURES
Forboth Irving and interviewing, data arfkdlon fom^sed in all
five .r.,1......« <>l inquiry, I mmHofl the ,,.....I |m**oK -1^
mined-heel on whichoneloe.sinformah.mle.....>ddunng<heoh, n
vation or interview. Interview protocols ťnabk- a person .c» take no es during theinterviewahou. the responsesnt «he interviewee. I l«-y also help a n-a, her organize thought, on items such a, heading« inior-„aaon about starting the mterview,concluding ideas Information on ending the Interview, and thanking the respondent. In Figure 75,1 provide the interview protocol used in our gunman case study (As-müssen & Owwdl 1995). Ik-sides the five open-ended qucshons in the gunman case study, this iorm contains several features I rtfcom-mend: Y    Utf a h#dtr to record UHttM Information about lbe projev., and as a
hMKfeuj might also include inřormalion about conhdentlality and addn-ss aspects included in the consent form. f   Place space between the autítions in lite protocol form. Recogni/.e that an individual may not always respond directly to the queens
,.,,   .a^.i-or example, while you ask Question 2, the inter-
,,«;.. may respond t o Quttlkmí ŕtP«p««dtowíH.not»ta
all oí the questions m the interviewee spaflks. r   M*nori* the questions and their order to minimi* losing eye contact.
Provide «ppropriale verhal transitions Irn.n one qiutftion I« the
next.
r    Wrde out the closing comments lhal thank the individual for the mttP view mul rcqurtt follow-up mfontmtionjf needed, from tlu-m.

V,
Data Collection                                                                             ▼ 127
Inrtrvlow Protocol
fro|*cl! Un|v«n% Rtodion lo a Ttrrorltl lncld«nl
Tim# 01 inf«/vww;
Dol#:
Plo«:
InJtMVNM
iiliMwwi
Position ol Mil«rvltw«a:
(Briefy dttfffet the pfottdj
QvMrtQMI
I,   Whol ho« ba*n rovt rok i« *• IMtdmlt
2.  Who) Kui Imppmffd «n« 'l>» ««nl that >«ľ hov* b»»n Mvolvtd mř
3. Whol hot bean *• Unpad on »• wn««ii*< commumrr o> 'h« incident?
4    Whol lofgat iimnllcö'loni, l( on« ■«til ham Ih« IncnUit)
5. lo *ho™ thou'd «* taft to And out moi» obowt oofpui '«tflon to iht incident?
(Thonfc inoVndvď Im pörhopo*ng m ihn «ife»wew A»tv» tum oc h*>» of conhoVi«ol*> of reipons»! anil potential foru« m««rvi«wi)
Figure 7.3    Sompfo Interview Protocol
.i*f.
I   'S     ▼
QUALITATIVE  INQUIRY
During <\n «bsvrv.ilinn, use an obwrvation.il protocol to record information. As shown in I'lgure 7.6. iliis prolocol records Information by one of my students on a class visit by Harry Wolcott. I provide only one page of the prolocol, but one can see thai il has a header describing information about the observational session and then includes "descriptive notes" where the researcher records a description of activities and a drawing of the physical setting. Moreover, the researcher provides "reflective notes"—notes about the process, reflections on activities, and summary conclusions about activities for later theme development. A line down the center of the page divides descriptive notes from reflective notes. A visual sketch of the setting and a header provide additional useful information.
Whether the investigator is using an observational or interview protocol, the essential process is recording information or, as Lofland and Lofland (1995) put it, "logging data" (p. 66). This process involves recording information through various forms such as observational i ii'Mnotes, interview write-ups, mapping, census taking, photographing, sound recording, and collecting and organizing documents. An Informal process may occur in recording information composed of ii nt i.i I "|nttings" (Emerson, ľretz, & Shaw, 1995). daily logs or summa-rieti, and descriptive summaries (see Snnjek, 1990, for examples of uctdnotes). These forms of recording information are popular in biographies, ethnographies, and case studies.
HELD ISSUES
■   i
Researchers engaged in studies within all five traditions face issues in tin- held when gathering data. During the last several years, I have systematically collected notes from students and ťolleagues .ibout the
Issues they encounter. Some common issues arc tlw need to change or «uljust the form of data collection unce they enter the field. An overwhelming response is surprise by beginning qualitative researchers nbout the amount of time needed to collect extensive data. For practice, I recommend limited data collection, such as one or two interviews or observations, so that nrsearchers can estimate the time needed to collect data. Along with the time issue is the concern about the .»mount ol energy ..ml fueus i.-ijuue.l to establish a substantial database.
■f/l
natal olU-ilwn
r u*
lang* o/At*wv 00 AW„
OaiBfrHnHrtti
ÄwflacirvaN.v.,
General. WW ate the experiences ol »word» In the dossrooml
"
$e* eiosiroom loyow ond coaunents obou*
phyilcol tatting at tht bottom of this page
O**«* —S Sap.: I «ortdW * *« faort
wHwwomww abh to mad it.
Apf*on*>o»V 5 17 pm. Of Creswe" •nwn Ihn hflid raor«, tntrodutes Df. Wokotl Clu» members nm njtewd.
1OvevAeod WOof not pfagoed if at thm beginning <>' the dint: I wider d 'hli writ o *tfroe*>- (.hen "too* <**„ r*»e k>
O Gaiw.ll g*., briel botka«»**.
coneanfroi|no an hi, irHernanonol mpoiloiK«! (buIuíoi o comment oboul Hi» eoWoiio-ol e*"ogn»ph, t*. Won * V» ftíncfpol". Office.
fi—ntu i/ifii imfinfiif f*ie C#Mwf<V*0* Wo'coll. S'uden's teemed o bif acinous. Maybe H had fo éo wtlŕi Ihe < Innige In storing Wne >D 5 p«. (sowe "»cr* ho»e boó 6:30 classes (C oppoi'n'nunn io get M.__________
Dr. Woko'i begins by letting ihe clou he nc1" wni»t out educononal ethnography and rMtfttSflhli tun primary o<Cupa*>on by ntfi'ionlntl two book»: I/umfening Qyofitoto* Dato and Tht Art of fretoW*.
I Of», CmiWf and Wo/coB sea/» to hovo a Hood ngpport between ihfMv jvógmg Aon. many thort axnangei mot May nod.
While Or Woltotl begins Kit praianlollon by opötog*hny to« hn waory yoke fdv* to rolling nil day, tippo-nnfl,), Or. Crafwall leaves »ho <*oisroom Io retrieve (he goesl's ourheod Sroniporrjnöei.
1   Lullt......I
chair
Seamed to ba thraa por»i io shls oít>. t| {1) tha ipeosW't ehorJange to rha dan ol delecting pi«o ethnographical molhddolo. gets. (?) tha speaker's prasa**)*»! cd she 'tr»V lhal portrays vorlaut ťolaglai and subitrolegiat tor qualitative research m adiKOtMHi, ond (3) tha raioitfd '«toV ilotai-man" lalding c<oti quatilon'i prlmorlly oboiH itvdanlt' pota^oi rataorch profach ond pnor ihrdiai Or Wilcott ťod wnfan
desk
ovarheod projeaor
:-"■)■-■ i
Saols lor panic^aMi
\
saafi
Tha (irsl question was 'How do you look at qw*«h, rwaeatAr kMowwd by Tto-doei olhnogrophy lil iní"
SWTCM Of ClASUfOOM
Figur« 7.6    Sornpl« Obwrvollonal Protocol
IM T
QUALITATIV!! INQUIRY
In Figure 7.7,1 enumerate other field issues and group them into topical areas These issues span access/site problems to observations. Interviews, document research, journals, video materials, and general ethical issues.
Issues of locating and obtaining permission to use materials present a challenge to biographical writers. The issues related to interviewing surface during phenomenological studies and grounded theory studies, whereas ethnographers struggle with access concerns and with sharing information with interviewees and participants in the cultural group. Case study writers, who gather extensive information, struggle with the time commitment and the details of interviewing.
Conducting interviews is taxing, especially for inexperienced researchers engaged in studies thai require extensive interviewing, such as phenomenology, grounded theory, and case study research. Equipment issues loom large as a problem in interviewing, and both recording equipment and transcribing equipment need to be organized in advance of the interview. The process of questioning during an interview (e.g., saying "little," handling "emotional outbursts," using "ice-breakers") includes problems that an interviewer must address. Many inexperienced researchers express surprise at the difficulty of conducting interviews and the lengthy process involved in transcribing audiotapes from the interviews. In addition, in phenomenological interviews, asking appropriate questions and relying on informants to discuss the meaning of their experiences require patience and skill on (he part of the researcher. In document research, the issues involve locating materials, often at sites far away, and obtaining permission to use the materials. For biographers, the primary form of data collection might be archival research from documents.
When the researcher asks participants in a study to keep journals, additional field issues surface. Joumaling is a popular data collection process in case studies. What instructions should be given to individuals prior to writing in their journals? Are all participants equally comfortable with joumaling? Is it appropriate, for example, with small children who express themselves well verbally but have limited writing skills? The researcher also may have difficulty reading the handwriting of participants who journal. Recording videotapes raises issues for the qualitative researcher such as keeping disturbing room sounds to a minimum, deciding on the best location for the camera,
K,
Data Collection
▼ 131
Aec#H/sife quMlfoni;
•  (Mfccufcr moling initial conto« with penon in ih« field
•   Building trust of credibility ot Ihe field tf» {(»ling like IntrvoV)
•  Go-Mta oecest "o the s*e and tndWduoli
•  Gelling people lo respond la rnqucsts Ic information
•  Deciding -f***' to eoHeci infonnobon <n it« no'urol site ■  Determining whether ona can understand u telling whan one if close to II
OrMewoficrtJ-
•   PetemifWftg white rWdnoHi am cmJMi
•  Willing down "ipHings'
•  Irtconporo-ng 'quotes' iftto UWnctes
•  Atigminr) an obtervor role and how to change roles
•   iMffWK) !»■ lo beat CO*eel information *">*» «erfy Seid WM| in cos« sh.-J.ei ■ Iteming how to 'funnel* dum food observations lo narrow ones
IWUntfrnW
-  Sflř"ng TimV during irsta/vM«
•  Moving tapes lhal will work in the transcribing machine
•   Sc^mduKng O Ume lor oll so participate In o group rterview
•  Morehing the 'level* of queitions to the ability ol informants
•   Rro'tting (he cosllu*» and teogrhy proces» of konscribieiQ da*o
•  Using an Appropriate level of questioning at the beginning ol Ihe interview
•   Interruptions during on interview
•  Difficulty scheduling on interview
•   Having confidence in interviewing skills
•  Having diSkvlry tailing notes -«■* interviewing
•   Conducting Interview with iwo or more individuals
•   f mouroging od partklpants lo lorV In a group interview
•  Aslang oppropnoto queslics
•   Leerung lo «sien rather man totting in interviews
•  Hondling «molional outbursts
•  Addressing participants who do no* won) » be auoWlaped
•  Finding a trontrrlptionisl artd Ihe rlghl typo of equipment in o cote study^pnd grounded »Seary project                                                                                                  e
•  Moving from ice-br>okert to questions in interview
•  Addressing when n Jen-tews ttroy from the interview queifcont
•  Giving the interwew quettlom lo pnniciponti before the interview e Working with the logáfei of *• lot*-recording equ-pmem
•   'Bracketing' personal bios
•  Focusing Ihe questions lo o*k m o group interview
reofflinuodj
Figur« 7.7    Field Ihuci in Quoltforive Research .1.11-
i
T                                                                  QUALITATIVE INQUIRY
DoCuro«nl růivirch:
•   Moving diKiiy'ly locoling ntoloriali
•  Moving diHicvlly obtaining petmnito" lo ula moleilaU
•  Qv»»**olng ** voloe of moWtoU
ImmA
■  Having paopl« »rtiM comploW |Oumal «nlrim
•   Moving diHicuily trading handvrilisn (ouinol*
•  Moving Inlonnonlj wpie&s <he difficulty ot |wfn«llng
t  Mavmg tnlormontt «pre» quethont oboui how or* mlghl journal
Wa*onto^"eh.
•   Hovtog d>Wv>b"ig room loMvjf
•  HovinQ problem» ndcoťsping i« o «noil n>om
•  Having, difficult!« focusing ond poitHonlng th» comtio
Ff'i.tol IMVHI
•   Protecting iho (»nonvmffy ol th* informant i
•  OndoiMie |oc itoQ <he ftvfpeuo oj *• rntofch
•  Binding ■htW {or new) lo us« uifemotun *thw*d o*> *i* nconf in on intem«w In a cow studf
•  D«ta«nining whsthei lhe tewrtw ihoMd »hau pononol wtpwiMcK
L------------.----------------------------------------------
Figur« 7,7    Continued
and determining whether to provide close-up shots versus distant
-.hoi .
Regaullcss of tradition of inquiry, a qualitativ? researcher faces many ethical issues that surface during data collection in the field and in analysis and dissemination of qualitative reports. The criteria of Ihe American Anthropological Association (sec Glesne & Peshkin, 1992) reflect appropriate standards. A researcher protects the anonymity of the informants, for example, by assigning numbers or aliases to individuals. A researcher develops case studies of individuals that represent a composite picture rather than an individual picture, furthermore, lo gain support from participants, a qualitative researcher conveys to participants thai they are participating in a study, explains the purpose of the study, and does not engage in deception about tht? nature of the study. What if the study is on a sensitive topic and the participants decline to be involved if they arc aware of the topic? This issue of disclosure of lhe researcher, widely discussed in cultural
■■Vi
Data Collection
T   133
anthropology (e.g., Hammersley & Atkinson, 1995), is handled by the researcher by presenting general information, not specific information about the study. Another issue likely to develop is when participants share information "off the record." Although in most instances this information is deleted from analysis by the researcher, the issue becomes problematic when the information harms individuals. I am reminded of a researcher who studied incarcerated Native Americans in prisons and learned aboul a potential "breakout" during one of the interviews. This researcher concluded that it would be a breach of faith with the informant if she reported lhe matter, and she kept quiel. Fortunately, the breakout was not attempted. A final ethical issue is whether the researcher shares experiences with informants in an interview setting such as in a case study, phenomenology, or ethnography. This sharing minimizes the "bracketing" that is essential to construct the meaning of participants in phenomenology and reduces information shared by informants in case studies and ethnographies.
STORING DATA
I am surprised at how little attention is given in books and articles about storing qualitative data. The approach to storage will reflect the type of information collected, which varies by tradition of inquiry, in writing a biographical life history, the researcher needs to develop a filing system for the "wad of handwritten notes or a tape" {Plummer, 1983, p. 98). Although his ideas are based on quantitative data, Davidson's (1996) suggestions about backing up information collected and noting changes made lo the database represent sound advice for qualitative researchers. With Ü\c advent of tile use of computers in qualitative research, more attention will likely be given to how qualitative data are organized and stored, whether the data are fiťldnolos, transcripts, or rough joltings. With extremely large databases being used by some qualitative researchers, this aspect assumes major importance. A computer program, such as Folio VIEWS, provides n program for organizing, sorting, and making subsets of text data. As one example of a textbase managing program. Folio VIEWS enables the investigator to search and retrieve various combinations of words, phrases, coded segments, memos, or other material (Weitzman At Miles, 1995),
nil
134 ▼
QUALITATIVE INQUIRY
Some principles about data storage and handling that are especially well suited for qualitative research include the following:
■   Always develop backup copies of computer files (Davidson, 1996).
■   Use high-quality tapes for audio-recording information during interviews. Also, make sure that the size of the tapes fits the transcriber's machine.
■   Develop a master list of types of information gathered.
■   Protect the anonymit)' of participants by masking their names in the data.
■   Convert word processing files over to ASCII files for easy entry into some qualitative computer programs (this topic will be addressed further in Chapter 8).
■   Develop a data collection matrix as a visual means of locating and identifying information for a study.
FIVE TRADITIONS COMPARED
Returning again toTable 7.1, there are both differences and similarities among the activities of data collection for the five traditions of inquiry. Turning to differences, first, the diversity of forms of data collection is great. For case studies, the researcher uses multiple forms of data to build the in-depth case. For grounded theory studies, pheno-menological projects, and biographies, investigators rely primarily on interviews. Ethnographers rely heavily on participant observation. Unquestionably, some mixing of forms occurs, but in general these patterns of collection by tradition hold true.
Second, the unit of data collection varies. Biographers, pheno-menologists, and ground theorists study individuals; case study researchers examine groups of individuals participating in an event or activity or an organization; and ethnographers study entire cultural systems orsomc subcultures of the systems. Third, I found the amount of discussion about field issues to vary. Ethnographers have written extensively about field issues (e.g., Hammersley & Atkinson, 1995), thus reflecting tine concerns of a stranger going into the field. Biographers are less specific about field issues (e.g., Denzin, 1989b), although
•if.

Data Collection                                                                            ▼ 135
concerns about sources of bias surface in discussions about the classical approach to biographical writings (Plummer, 1983).
Fourth, the traditions vary in their inlrusiveness of data collection. Conducting interviews seems less intrusive in phenomenological projects and grounded theory studies than does"ihc high level of access needed in personal biographies, the prolonged stays in the field in ethnographies, and the immersion into programs or events in case studies.
In my review of the five traditions and data collection, I found overlap in several areas. A number of these were highlighted earlier in the chapter. All qualitative studies conducted in public organizations need to be approved by a human subjects review board. Also, the use of interviews and observations is central to many of the traditions. Furthermore, the recording devices such as observational and interview protocols can be similar regardless of tradition (although specific questions on each protocol will reflect the language of the tradition). Finally, the issue of storage of information is closely related to the form of data collection, and the basic objective of researchers, regardless of tradition, is to develop some filing and storing system for organized retrieval of information.
SUMMARY
In this chapter, I addressed several components of the data collection process. The researcher attends to locating a site or person to study, gaining access and building rapport at the site or with the individual, sampling purposefully using one of the many approaches to sampling in qualitative research,collecting information throughas many as four forms (interviews, observations, documents, and audio-visual material), establishing approaches for recording information such as the use of interview or observational protocols, resolving field issues ranging from access to ethical concerns, and developing a system for storing and handling the databases. Applied to the five traditions of inquiry, the traditions differ in the diversity of information collected, the unit of study being examined, the extent of field issues, and the intrusivencss of the data collection effort. Researchers, regardless of tradition, need approval from review boards, engage in similar data collection of interviews and observations, and use similar recording protocols and forms for storing data. .ou*
i
136 T
QUALITATIVE INQUIRY
T ADDITIONAL READINGS
For a discussion about purposeful sampling strategies, I recommend Miles and Huberman (1994).
Mil», M. B., U Huberman, A. M. (1994). Qualitative data analysis: A sourcebook of new methods (2nd ed.). Thousand Oaks, CA: Sage.
For interviewing, 1 direct researchers to Kvole (1996), McCracken (1988), Rubin and Rubin (1995), Seidman (1991), and Weiss (1992).
Kvale, S. (19%). Interviews: An introduction to qualitative research
interviewing. Thousand Oak», CA: Sage. McCracken, G. (1988). The long interview. Newbury Park, CA: Sage.
Knhln, 11.1., & Rubm,I.S.(1995).yr^^
Onks, CA: Sage. Seidman, 1, E. (1991). Interviewing as qualitative research. New York:
Columbia University, Teachers College Press.
Weiss, R. S. (1992). Learning from strangers: The an and method of qualitative interview studies. New York: I'rcc Press.
e'or discussions about making observations and taking fieldnoles, I suggest several writers: Bernard (1994), Bogdewic (1992), Emerson el al. (1995), Hammeraley and Atkinson (1995), Jorgensen (1989), and Sanjek (1990).
Bernard, 11, K. (1994). Restaur method» in anthropology: Qualitative mul quantitative approaches (2nd ed.), Thousand Oaks, CA; Sage,
Bogdewic, S. P. (1992). Participant observation. In B. F. Crabtree & W. L Miller (Eds.), Doing qualitative researdt (pp. 45-69). Newbury Park, CA: Sage.
Emerson, R. M-, Fretz, R. I„ & Shaw, L. L. (1995). Writing ethnographic f ieldnotes. Chicago: University of Chicago Press.
I l.immersley, M., & Atkinson, P (1995), Venography: Principles in practice (2nd ed.). New York Rout led j-e
Jorgensen, D. L. (1989). Participant observation: A methodology for
human studies. Newbury Park, ( A: S.i^e.
».,
Data Collection
T   1.T7
Sanjek, R. (1990). Iieldiiotes: The making ofanthr,;. -,'■ , u Ithaca, NY: Cornell University I'revt
ľor a discussion of field relations and issues, see I l.itninersley and Atkinson (1995) and Lofland and Lofland (1995}.,
Hammersley, M., U Atkinson, P. (1995). Ethnography: Principles m practice (2nd ed.). New York: Routlcdge.
Lofland, J., & Lofland, L. H. (1995). Analyzing social settings: A guide to qualitative observation and analysis (3rd ed.). Belmont, CA: Wadsworth.
EXERCISES
I (iain some experience In collet Iin>; data for your projeti. ( i nullit I eilher an interview or an observation and record the informal ion on a protocol form. After this experience, identify issues that posed challenges in dala collection.
2. It is helpful to design the data collection activities for a project.
Examine Table 7.1 for the seven activities. Develop a matrix that describes data collection for all seven activities for your project.
on»
t
1