strategic themes in qualitative inquiry - sage publications · within a grand strat- ... that’s...

CONCEPTUAL ISSUES IN QUALITATIVE INQUIRY Strategic Themes in Qualitative Inquiry

2Strategic Themes inQualitative Inquiry

G rand strategy should guide tactical decisions. Within a grand strat-egy all manner of tactical errors may be made, and indeed, are inevi-

table, but can be corrected as long as the strategic vision remains true and fo-cused. At least that’s the theory. In practice . . . ? Try it and see.

—Halcolm

General Principles __________________________

Strategos is a Greek word meaning “the thinking and action of a general.”What it means to be strategic is epitomized by that greatest of Greek generals,Alexander. He conducted his first independent military operation in northernMacedonia at age 16. He became the ruler of Macedonia after his father, Philip,was assassinated in 336 B.C. Two years later, he embarked on an invasion of Per-sia and conquest of the known world. In the Battle of Arbela, he decisively de-feated Darius III, King of Kings of the Persian Empire, despite being outnum-bered 5 to 1 (250,000 Persians against Alexander and fewer than 50,000 Greeks).

Alexander’s military conquests are legend. What is less known and little ap-preciated is that his battlefield victories depended on in-depth knowledge of the

� 37

psychology and culture of the ordinary people and military leaders in opposingarmies. He included in his military intelligence information about the beliefs,worldview, motivations, and patterns of behavior of those he faced. Moreover,his conquests and subsequent rule were more economic and political in naturethan military. He used what we would now understand to be psychological, so-ciological, and anthropological insights. He understood that lasting victory de-pended on the goodwill of and alliances with non-Greek peoples. He carefullystudied the customs and conditions of people he conquered and adapted hispolicies—politically, economically, and culturally—to promote good conditionsin each locale so that the people were reasonably well-disposed toward his rule(Garcia 1984).

In this approach, Alexander had to overcome the arrogance and ethnocen-trism of his own training, culture, and Greek philosophy. Historian C. A. Robin-son, Jr. explained that Alexander was brought up in Plato’s theory that allnon-Greeks were barbarians, enemies of the Greeks by nature, and Aristotletaught that all barbarians (non-Greeks) were slaves by nature. But

Alexander had been able to test the smugness of the Greeks by actual contact withthe barbarians, . . . and experience had apparently convinced him of the essentialsameness of all people. (Robinson 1949:136)

In addition to being a great general and enlightened ruler, Alexander appearsto have been an extraordinary ethnographer, a qualitative inquirer par excel-lence, using observations and firsthand experience to systematically study andunderstand the peoples he encountered and to challenge his own culture’s prej-udices.

And as Halcolm finished telling the story of Alexander the Great, he re-minded those assembled that skills in observation and interviewing are lifeskills for experiencing the world. “One can say of qualitative inquiry what Mar-cel Proust said of art, ‘Thanks to this, instead of seeing one world, our own, wesee it multiplied. So many worlds are at our disposal.’ ”

—From Halcolm’s Historical Biographies

The Purpose of a Strategic Framework

P erception is strong and sight weak. In strategy it is important to seedistant things as if they were close and to take a distanced view of

close things.

—Miyamoto Musashi (1584-1645),Japanese warrior, strategist

38 � CONCEPTUAL ISSUES IN QUALITATIVE INQUIRY

D on’t mistake a clear view for a short distance.

—Grand Canyon hiking advice

E verybody has a plan until they’ve been hit.

—Old boxing saying

A well-conceived strategy, by providingoverall direction, provides a framework fordecision making and action. It permitsseemingly isolated tasks and activities to fittogether, integrating separate efforts towarda common purpose. Specific study designand methods decisions are best made withinan overall strategic framework. This chap-ter offers 12 major themes or principles ofqualitative inquiry that, taken together, con-stiitute a comprehensive and coherent stra-tegic framework for qualitative inquiry, in-cluding fundamental assumptions andepistemological ideals. Exhibit 2.1 summa-rizes those themes in three basic categories:design strategies, data collection and field-work strategies, and analysis strategies.

� Design Strategies forQualitative Inquiry

Naturalistic Inquiry

An anthropologist studies initiation ritesamong the Gourma people of Burkina Fasoin West Africa. A sociologist observes inter-actions among bowlers in their weeklyleague games. An evaluator participatesfully in a leadership training program she isdocumenting. A naturalist studies bighornsheep beneath Powell Plateau in the GrandCanyon. A policy analyst interviews peopleliving in public housing in their homes. Anagronomist observes farmers’ spring plant-ing practices in rural Minnesota. What do

these researchers have in common? They arein the field studying the real world as it un-folds.

Qualitative designs are naturalistic to theextent that the research takes place in real-world settings and the researcher does notattempt to manipulate the phenomenon ofinterest (e.g., a group, event, program, com-munity, relationship, or interaction). Thephenomenon of interest unfolds naturally inthat it has no predetermined course estab-lished by and for the researcher such aswould occur in a laboratory or other con-trolled setting. Observations take place inreal-world settings and people are inter-viewed with open-ended questions in placesand under conditions that are comfortablefor and familiar to them.

Egon Guba (1978), in his classic treatise onnaturalistic inquiry, identified two dimen-sions along which types of scientific inquirycan be described: (1) the extent to which thescientist manipulates some phenomenon inadvance in order to study it and (2) the ex-tent to which constraints are placed on out-puts, that is, the extent to which predeter-mined categories or variables are used todescribe the phenomenon under study. Hethen defined “naturalistic inquiry” as a“discovery-oriented” approach that mini-mizes investigator manipulation of thestudy setting and places no prior constraintson what the outcomes of the research will be.Naturalistic inquiry contrasts with con-trolled experimental designs where, ideally,the investigator controls study conditions

Strategic Themes in Qualitative Inquiry � 39

by manipulating, changing, or holding con-stant external influences and where a verylimited set of outcome variables is mea-sured. Open- ended, conversation-like inter-views as a form of naturalistic inquiry con-trast with questionnaires that have prede-termined response categories. It’s the differ-

ence between asking, “Tell me about yourexperience in the program” and “How satis-fied were you? Very, somewhat, little, not atall.”

In the simplest form of controlled experi-mental inquiry, the researcher enters theprogram at two points in time, pretest and


EXHIBIT 2.1 Themes of Qualitative Inquiry

Design Strategies

1. Naturalistic inquiry Studying real-world situations as they unfold naturally;nonmanipulative and noncontrolling; openness to whateveremerges (lack of predetermined constraints on findings).

2. Emergent design2. flexibility

Openness to adapting inquiry as understanding deepens and/orsituations change; the researcher avoids getting locked into rigiddesigns that eliminate responsiveness and pursues new paths ofdiscovery as they emerge.

3. Purposeful sampling Cases for study (e.g., people, organizations, communities,cultures, events, critical incidences) are selected becausethey are “information rich” and illuminative, that is, they offeruseful manifestations of the phenomenon of interest; sampling,then, is aimed at insight about the phenomenon, not empiricalgeneralization from a sample to a population.

Data Collection and Fieldwork Strategies

4. Qualitative data Observations that yield detailed, thick description; inquiry indepth; interviews that capture direct quotations about people’spersonal perspectives and experiences; case studies; carefuldocument review.

5. Personal experience5. and engagement

The researcher has direct contact with and gets close to thepeople, situation, and phenomenon under study; the researcher’spersonal experiences and insights are an important part ofthe inquiry and critical to understanding the phenomenon.

6. Empathic neutrality6. and mindfulness

An empathic stance in interviewing seeks vicarious under-standing without judgment (neutrality) by showing openness,sensitivity, respect, awareness, and responsiveness; inobservation it means being fully present (mindfulness).

7. Dynamic systems Attention to process; assumes change as ongoing whetherfocus is on an individual, an organization, a community, or anentire culture; therefore, mindful of and attentive to systemand situation dynamics.

posttest, and compares the treatment groupto some control group on a limited set ofstandardized measures. Such designs as-sume a single, identifiable, isolated, andmeasurable treatment. Moreover, such de-signs assume that, once introduced, the

treatment remains relatively constant andunchanging.

While there are some narrow, carefullycontrolled, and standardized treatmentsthat fit this description, in practice humaninterventions (programs) are often quitecomprehensive, variable, and dynamic—


Analysis Strategies

8. Unique case8. orientation

Assumes each case is special and unique; the first level ofanalysis is being true to, respecting, and capturing the detailsof the individual cases being studied; cross-case analysisfollows from and depends on the quality of individual casestudies.

9. Inductive analysis9. and creative synthesis

Immersion in the details and specifics of the data to discoverimportant patterns, themes, and interrelationships; begins byexploring, then confirming; guided by analytical principlesrather than rules; ends with a creative synthesis.

10. Holistic perspective The whole phenomenon under study is understood as a complexsystem that is more than the sum of its parts; focus on complexinterdependencies and system dynamics that cannot meaning-fully be reduced to a few discrete variables and linear, cause-effect relationships.

11. Context sensitivity Places findings in a social, historical, and temporal context;careful about, even dubious of, the possibility or meaningful-ness of generalizations across time and space; emphasizesinstead careful comparative case analyses and extrapolatingpatterns for possible transferability and adaptation in newsettings.

12. Voice, perspective,12. and reflexivity

The qualitative analyst owns and is reflective about her or hisown voice and perspective; a credible voice conveys authentic-ity and trustworthiness; complete objectivity being impossibleand pure subjectivity undermining credibility, the researcher’sfocus becomes balance—understanding and depicting the worldauthentically in all its complexity while being self-analytical,politically aware, and reflexive in consciousness.

changing as practitioners learn what doesand does not work, developing new ap-proaches and realigning priorities. This, ofcourse, creates considerable difficulty forcontrolled experimental designs that needspecifiable, unchanging treatments to relateto specifiable, predetermined outcomes.Controlled experimental evaluation designswork best when it is possible to limit pro-gram adaptation and improvement so as notto interfere with the rigor of the research de-sign.

By contrast, under real-world conditionswhere programs are subject to change andredirection, naturalistic inquiry replaces thefixed treatment/outcome emphasis of thecontrolled experiment with a dynamic, pro-cess orientation that documents actual oper-ations and impacts of a process, program, orintervention over a period of time. The eval-uator sets out to understand and documentthe day-to-day reality of participants in theprogram, making no attempt to manipulate,control, or eliminate situational variables orprogram developments, but accepting thecomplexity of a changing program reality.The data of the evaluation include whateveremerges as important to understandingwhat participants experience.

Natural experiments occur when the ob-server is present during a real-world changeto document a phenomenon before and afterthe change. Durrenberger and Erem (1999)documented “a natural experiment inthought and structure” when, because of achange at a hospital they were studying,they were able to contrast two differentstructures of leadership in a union worksite.They had already documented the degreeand nature of “union consciousness” beforethe change, so by repeating their observa-tions after the change in a hospital structure,they were able to take advantage of a natu-rally occurring experiment. Natural experi-

ments can involve comparing two groups,one of which experiences some changewhile the other doesn’t. What makes thisnaturalistic inquiry is that real-world partic-ipants direct the change, not the researcher,as in the laboratory.

However, the distinction is not as simpleas being in the field versus being in the labo-ratory; rather, the degree to which a design isnaturalistic falls along a continuum withcompletely open fieldwork on one end andcompletely controlled laboratory control onthe other end, but with varying degrees ofresearcher control and manipulation be-tween these end points. For example, thevery presence of the researcher, asking ques-tions, or as in the case of formative programevaluation, providing feedback, can be anintervention that reduces the natural un-folding of events. Unobtrusive observationsare needed as an inquiry strategy when theinquirer wants to minimize data collectionas an intervention. Nor are laboratory condi-tions found only in buildings. Field experi-ments are common in agriculture where re-searchers want to introduce a considerableamount of control, reduce variation in extra-neous variables, and focus on a limited set ofpredetermined measures, as in crop fertil-izer studies.

Let me conclude this discussion of natu-ralistic inquiry with two examples to illus-trate variations in this design strategy. Inevaluating a wilderness-based leadershiptraining program, I participated fully in the10-day wilderness experience, guided in myobservations by nothing more than the sen-sitizing concept “leadership.” The only “un-natural” elements of my participation werethat (1) everyone knew I was taking notes todocument what happened and (2) at the endof each day I conducted open-ended, con-versational interviews with staff. While thisconstitutes a relatively pure naturalistic in-


quiry strategy, my presence, note taking,and interviews must be presumed to havealtered somewhat the way the program un-folded. I know, for example, that the debrief-ing questions I asked staff in the eveningsgot them thinking about things they weredoing that led to some changes along theway in how they conducted the training.

The second example comes from thefieldwork of Beverly Strassmann among theDogon people in the village of Sangui in theSahel, about 120 miles south of Tombouctouin Mali, West Africa (Gladwell 2000). Herstudy focused on the Dogon tradition ofhaving menstruating women stay in small,segregated adobe huts at the edge of the vil-lage. She observed the comings and goingsof these women and obtained urine samplesfrom them to be sure they were menstruat-ing. The women only slept in the huts. Dur-ing the day, they went about their normal ac-

tivities. For 736 consecutive nights, Strass-mann kept track of all the women who usedthe hut. This allowed her to collect statis-tics on the frequency and length of menstru-ation among the Dogon women, but with acompletely naturalistic inquiry strategy, il-lustrating how both quantitative and quali-tative data can be collected within a natural-istic design strategy. There’s no reason tobelieve that her presence over this long pe-riod changed the women’s menstruationpatterns.

Emergent Design Flexibility

In the wilderness leadership training pro-gram I evaluated, halfway through the10-day experience the group I was with un-expectedly split into two subgroups. I had tomake an in-the-field, on-the-spot decision


©20

02M

icha

elQ

uinn

Patt

onan

dM

icha

elC

ochr

an

about which group to follow and how to getinterviews with the others at a later time.

Naturalistic inquiry designs cannot usu-ally be completely specified in advance offieldwork. While the design will specify aninitial focus, plans for observations, and ini-tial guiding interview questions, the natu-ralistic and inductive nature of the inquirymakes it both impossible and inappropriateto specify operational variables, state test-able hypotheses, or finalize either instru-mentation or sampling schemes. A natural-istic design unfolds or emerges as fieldworkunfolds.

Lincoln and Guba (1985) made an exten-sive comparison of the design characteristicsof qualitative/naturalistic inquiry in con-trast to quantitative/experimental methods.They concluded:

What these considerations add up to is that thedesign of a naturalistic inquiry (whether re-search, evaluation, or policy analysis) cannotbe given in advance; it must emerge, develop,unfold. . . . The call for an emergent design bynaturalists is not simply an effort on their partto get around the “hard thinking” that is sup-posed to precede an inquiry; the desire to per-mit events to unfold is not merely a way ofrationalizing what is at bottom “sloppy in-quiry.” The design specifications of the con-ventional paradigm form a procrustean bed ofsuch a nature as to make it impossible for thenaturalist to lie in it—not only uncomfortably,but at all. (p. 225)

Design flexibility stems from the open-ended nature of naturalistic inquiry as wellas pragmatic considerations. Being openand pragmatic requires a high tolerance forambiguity and uncertainty as well as trust inthe ultimate value of what inductive analy-sis will yield. Such tolerance, openness, andtrust create special problems for dissertationcommittees and funders of evaluation or re-

search. How will they know what will resultfrom the inquiry if the design is only par-tially specified? The answer is: They won’tknow with any certainty. All they can do islook at the results of similar qualitative in-quiries, inspect the reasonableness of theoverall strategies in the proposed design,and consider the capacity of the researcherto fruitfully undertake the proposed study.

As with other strategic themes of qualita-tive inquiry, the extent to which the design isspecified in advance is a matter of degree.Doctoral students doing qualitative disser-tations will usually be expected to presentfairly detailed fieldwork proposals and in-terview schedules so that the approvingdoctoral committee can guide the studentand be sure that the proposed work will leadto satisfying degree requirements. Manyfunders will fund only detailed proposals.As an ideal, however, the qualitative re-searcher needs considerable flexibility andopenness. The fieldwork approach of an-thropologist Brackette F. Williams repre-sents the ideal of emergence in naturalisticinquiry.

Williams has focused on issues of culturalidentity and social relationships. Her workhas included in-depth study of ritual andsymbolism in the construction of nationalidentity in Guyana (1991), and the ways thatrace and class function in the national con-sciousness of the United States. In 1997, shereceived a five-year MacArthur Fellowship,which has allowed her to pursue a trulyemergent, naturalistic design in her currentfieldwork on the phenomenon of killing inAmerica. I had the opportunity to interviewher about her work and am including sev-eral excerpts from that interview1 through-out this chapter to illustrate actual scholarlyimplementation of some of the strategic ide-als of qualitative inquiry. Here she describesthe necessity of an open-ended approach toher fieldwork because her topic is broad and


she needs to follow wherever the phenome-non takes her.

I’m tracking something—killing—that’s mov-ing very rapidly in the culture. Every time Italk to someone, there’s another set of data, an-other thing to look at. Anything that happensin America can be relevant, and that’s the ex-hausting part of it. It never shuts off. You listento the radio. You watch television. You pass abillboard with an advertisement on it. There’sno such thing as something irrelevant whenyou’re studying something like this or maybejust studying the society that you’re in. Youdon’t always know exactly how it’s going to berelevant, but somehow it just strikes you andyou say to yourself: I should document thedate of when I saw this and where it was andwhat was said because it’s data.

I don’t follow every possible lead peoplegive me. But generally, it is a matter in somesense of opportunity sampling, of serendipity,whatever you want to call it. I key into thingsthat turn out to be very important six monthslater.

I do impromptu interviews. I don’t havesome target number of interviews in mind orpredetermined questions. It depends on theperson and the situation. Airports, for exam-ple, are a good place for impromptu inter-views with people. So sometimes, instead ofusing airport time to write, I interview peopleabout the death penalty or about killing orabout death in their life. It’s called opportunitysampling. I begin with a general description.You’re such and such an age. You come fromsuch and such a place and, by the way, what doyou think about all this killing? And I sort oflaunch into a conversation. Sometimes the in-terview goes on for a couple of hours andsometimes, maybe 10 or 15 minutes. I just say,“You wouldn’t mind if I record this, wouldyou?” If they say no, I take notes.

I did a lot of that kind of impromptu inter-viewing in the first year to formulate a proto-

col of questions and issues to pursue. It wasgeneral sampling to get a sense of what Iwanted to know. At other times, it’s just to geta general opinion from John Q. Public about aquestion that I’ve gotten all kinds of official re-sponses to, but I want to know what people ingeneral think. In an airport, I may get an op-portunity to talk to 5 or 10 people. If I have sev-eral stops, I may get 15 or 20 by the time I comehome.

I fashion the research as I want to fashion itbased on what I think this week as opposed towhat I thought last week. I don’t follow someproposal. I don’t have in mind that this has tobe a book that’s going to have to come out acertain way. I’m following where the data takeme, where my questions take me.

Few qualitative studies are as fully emergentand open-ended as the fieldwork of Wil-liams. Her work exemplifies the ideal ofemergent design flexibility.

Purposeful Sampling

In 1940, eminent sociologist KingsleyDavis published what was to become a clas-sic case study, the story of Anna, a baby keptin nearly total isolation from the time of herbirth until she was discovered at age six. Shehad been deprived of human contact, hadacquired no language skills, and had re-ceived only enough care to keep her barelyalive. This single case, horrifying as was theabuse and neglect, offered a natural experi-ment to study socialization effects and therelative contributions of nature and nurtureto human development. In 1947, Davis pub-lished an update on Anna and a comparisoncase of socialization isolation, the story ofIsabelle. These two cases offered consider-able insight into the question of how long ahuman being could remain isolated before“the capacity for full cultural acquisition”was permanently damaged (Davis 1940,


1947). The cases of Anna and Isabelle are ex-treme examples of purposeful case sam-pling.

Unusual clinical cases in medicine andpsychology, instructive precisely becausethey are unusual, offer many examples ofpurposeful sampling. Neurologist OliverSacks (1985) presents a number of such casesin his widely read and influential book TheMan Who Mistook His Wife for a Hat, the verytitle of which hints at the uniqueness of thecases examined. While one cannot general-ize from single cases or very small samples,one can learn from them—and learn a greatdeal, often opening up new territory for fur-ther research, as was the case with Piaget’sdetailed and insightful observations of hisown two children.

Perhaps nowhere is the difference be-tween quantitative and qualitative methodsbetter captured than in the different strate-gies, logics, and purposes that distinguishstatistical probability sampling from quali-tative purposeful sampling. Qualitative in-quiry typically focuses on relatively smallsamples, even single cases (N = 1) such asAnna or Isabelle, selected purposefully to per-mit inquiry into and understanding of a phe-nomenon in depth. Quantitative methodstypically depend on larger samples selectedrandomly in order to generalize with confi-dence from the sample to the populationthat it represents. Not only are the tech-niques for sample selection different, but thevery logic of each approach is distinct be-cause the purpose of each strategy is differ-ent.

The logic and power of probability sam-pling derive from its purpose: generaliza-tion. The logic and power of purposefulsampling derive from the emphasis onin-depth understanding. This leads to se-lecting information-rich cases for study indepth. Information-rich cases are those from

which one can learn a great deal about issuesof central importance to the purpose of theresearch, thus the term purposeful sampling.For example, if the purpose of an evaluationis to increase the effectiveness of a programin reaching lower-socioeconomic groups,one may learn a great deal more by focusingin depth on understanding the needs, inter-ests, and incentives of a small number ofcarefully selected poor families than bygathering standardized information from alarge, statistically significant sample. Thecases sampled can be individual people,families, organizations, cultures, incidents,or activities, to mention examples. But re-gardless of the kind of unit of analysis (e.g.,an athlete or a sports team, a teacher or aclassroom), the purpose of purposeful sam-pling is to select information-rich caseswhose study will illuminate the questionsunder study.

Chapter 5 will review several differentstrategies for purposefully selecting infor-mation-rich cases. In my interview with her,Brackette F. Williams offered an example ofan information-rich case from her ongoingstudy of killing in America.

I’ve been tracking information on a serialkiller—someone who has just been identifiedas a “serial killer” in Louisiana—who’s killingyoung Black men, shooting them up withdrugs and taking one of their tennis shoes,sometimes both. Now, I’m interested in thefact that as society more and more identifiesyoung Black men as sort of the quintessentialbad guys, this serial killer picks a bad guy. Forcontrast, look at serial killers who pickedwomen at a certain period of time, about 15-20years ago, because they wore, in his estima-tion, a size 13. Now, track our obsession withobesity. How a serial killer picks his victimscan tell you something important aboutwhat’s going on in society.


� Data Collection andFieldwork: Strategiesfor Qualitative Inquiry

Qualitative Data

Qualitative data consist of quotations, ob-servations, and excerpts from documents.The first chapter provided several examplesof qualitative data. Deciding whether to usenaturalistic inquiry or an experimental ap-proach is a design issue. This is differentfrom deciding what kind of data to collect(qualitative, quantitative, or some combina-tion), although design and data alternativesare clearly related. Qualitative data can becollected in experimental designs whereparticipants have been randomly dividedinto treatment and control groups. Likewise,some quantitative data may be collected innaturalistic inquiry approaches. Neverthe-less, controlled experimental designs pre-dominantly aim for statistical analyses ofquantitative data, while qualitative data arethe primary focus in naturalistic inquiry.This relationship between design and mea-surement will be explored at greater lengthin the chapter on design.

Qualitative data describe. They take us,as readers, into the time and place of the ob-servation so that we know what it was like tohave been there. They capture and commu-nicate someone else’s experience of theworld in his or her own words. Qualitativedata tell a story. In the excerpt below, frommy interview with her, Williams tells thestory of checking out a childhood memory.This story gives us insight into the nature ofher naturalistic inquiry and open-ended in-terviewing, shows how a critical incidentcan be a purposeful sample, and, in the storyitself, offers something of the flavor of quali-tative data.

I was down in Texas interviewing last March,thinking about my research and interviewingpeople, and there was a childhood memorythat I had of an electrocution of a man that wasthe son of a woman who lived across the fieldfrom us. Now a rumor about this had alwaysbeen in the back of my mind. Whenever I’dhear about a death penalty case over the years,I would think about this man having beenelectrocuted. I thought he was electrocuted be-cause he raped this White woman. So I’m sit-ting in my cousin’s kitchen after I had donesome of these interviews and another woman,an older woman who was a relative of hers,came in and the conversation goes around. Ihappen to mention this memory of mine. Iasked, “Is that just something that I concoctedout of having read a book or something, but itnever happened?” She answered, “Oh, no, ithappened. You only have one part of the storywrong. He didn’t rape her. He looked at her.”

You know, you read about these things inhistory books and then all of a sudden, it’s likea part of a world that you existed in. Thesethings happened around you and yet some-how there was so much of a distance, youcouldn’t touch it. I knew about this man all mylife, but in all the reading and all the historybooks, I couldn’t touch that. Doing this projectthe way I’m doing it allows me to touch things thatotherwise I would never touch.

Direct Personal Experience andEngagement: Going Into the Field

The preceding quotation from Williamsexemplifies the personal nature of qualita-tive fieldwork. Getting close to her subjectmatter, including using her own experi-ences, both from childhood and day-to-dayin her adult life, illustrates the all-encom-passing and ultimately personal nature ofin-depth qualitative inquiry. Traditionally,


social scientists have been warned to staydistant from those they studied to maintain“objectivity.” But that kind of detachmentcan limit one’s openness to and understand-ing of the very nature of what one is study-ing, especially where meaning-making andemotion are part of the phenomenon. Lookclosely at what Williams says about the ef-fects of immersing herself personally in herfieldwork, even while visiting relatives:“Doing this project the way I’m doing it al-lows me to touch things that otherwise Iwould never touch.”

Fieldwork is the central activity of quali-tative inquiry. “Going into the field” meanshaving direct and personal contact with peo-ple under study in their own environments—getting close to the people and situa-tions being studied to personally under-stand the realities and minutiae of daily life,for example, life as experienced by partici-pants in a welfare-to-work program. The in-quirer gets close to the people under studythrough physical proximity for a period oftime as well as through development ofcloseness in the social sense of shared expe-rience, empathy, and confidentiality. Thatmany quantitative methodologists fail toground their findings in personal qualitativeunderstanding poses what sociologist JohnLofland (1971) called a major contradictionbetween their public insistence on the ade-quacy of statistical portrayals of other hu-mans and their personal everyday dealingswith and judgments about other humanbeings.

In everyday life, statistical sociologists, likeeveryone else, assume that they do not knowor understand very well people they do notsee or associate with very much. They assumethat knowing and understanding other peoplerequire that one see them reasonably often andin a variety of situations relative to a variety ofissues. Moreover, statistical sociologists, like

other people, assume that in order to know orunderstand others one is well-advised to givesome conscious attention to that effort inface-to-face contacts. They assume, too, thatthe internal world of sociology—or any othersocial world—is not understandable unlessone has been part of it in a face-to- face fashionfor quite a period of time. How utterly para-doxical, then, for these same persons to turnaround and make, by implication, preciselythe opposite claim about people they havenever encountered face-to-face—those peopleappearing as numbers in their tables and ascorrelations in their matrices! (Lofland 1971:3)

Qualitative inquiry means going into thefield—into the real world of programs, orga-nizations, neighborhoods, street corners—and getting close enough to the people andcircumstances there to capture what is hap-pening. To immerse oneself in naturally oc-curring complexity involves what qualita-tive methodologist Norman Denzin (1978a)has called “the studied commitment to ac-tively enter the worlds of interacting in-dividuals” (pp. 8-9). This makes possibledescription and understanding of both exter-nally observable behaviors and internalstates (worldview, opinions, values, atti-tudes, and symbolic constructs). Given thequalitative emphasis on striving for depth ofunderstanding, in context, attitude surveysand psychological tests are inadequate forrevealing inner perspectives. “The innerperspective assumes that understandingcan only be achieved by actively participat-ing in the life of the observed and gaining in-sight by means of introspection” (Bruyn1963:226).

Actively participating in the life of the ob-served means going where the action is, get-ting one’s hands dirty, participating wherepossible in actual program activities, andgetting to know program staff and partici-pants on a personal level—in other words,


getting personally engaged so as to use all ofone’s senses and capacities, including the ca-pacity to experience affect no less than cog-nition. Such engagement stands in sharpcontrast to the professional comportment ofsome in the field, for example, supposedlyobjective evaluators, who purposely projectan image of being cool, calm, external,and detached. Such detachment is pre-sumed to reduce bias. However, qualitativemethodologists question the necessity andutility of distance and detachment, assertingthat without empathy and sympathetic in-trospection derived from personal encoun-ters, the observer cannot fully understandhuman behavior. Understanding comesfrom trying to put oneself in the other per-son’s shoes, from trying to discern how oth-ers think, act, and feel.

In a classic study, educational evaluatorEdna Shapiro (1973) studied young chil-dren in classrooms in the national FollowThrough program using both quantitativeand qualitative methods. It was her close-ness to the children in those classrooms thatallowed her to see that something was hap-pening that was not captured by standard-ized tests. She could see differences in chil-dren, observe their responses to diversesituations, and capture the varying mean-ings they attached to common events. Shecould feel their tension in the testing situa-tion and their spontaneity in the more natu-ral classroom setting. Had she worked solelywith data collected by others or only at a dis-tance, she would never have discovered thecrucial differences in the classroom settingsshe studied—differences that actually al-lowed her to evaluate the innovative pro-gram in a meaningful and relevant way.Where standardized tests showed no differ-ences between classrooms using differentapproaches, her direct observations docu-mented important and significant programimpacts.

It is important to note that the admonitionto get close to the data is in no way meant todeny the usefulness of quantitative meth-ods. Rather, it means that statistical portray-als must always be interpreted and givenhuman meaning. I once interviewed an eval-uator of federal health programs who ex-pressed frustration at trying to make senseout of statistical data from over 80 projectsafter site visit funds had been cut out of theevaluation: “There’s no way to evaluatesomething that’s just data. You know, youhave to go look.”

Going into the field and having personalcontact with program participants is not theonly legitimate way to understand humanbehavior. For certain questions and for situa-tions involving large groups, distance is in-evitable, perhaps even helpful, but to get atdeeper meanings and preserve context,face-to-face interaction is both necessaryand desirable. This returns us to a recurrenttheme of this book: matching research meth-ods to the purpose of a study, the questionsbeing asked, and the resources available.

In thinking about the issue of closeness tothe people and situations being studied, it isuseful to remember that many major contri-butions to our understanding of the worldhave come from scientists’ personal experi-ences. One finds many instances where close-ness to sources of data made key insightspossible— Piaget’s closeness to his children,Freud’s proximity to and empathy with hispatients, Darwin’s closeness to nature, andeven Newton’s intimate encounter with anapple. In short, closeness does not make biasand loss of perspective inevitable; distanceis no guarantee of objectivity.

Empathic Neutrality

If, as the previous section has discussed,naturalistic inquiry involves fieldwork thatputs one in close contact with people and


their problems, what is to be the researcher’scognitive and emotional stance towardthose people and problems? No universalprescription can capture the range of possi-bilities, for the answer will depend on thesituation, the nature of the inquiry, and theperspective of the researcher. But thinkingstrategically, I offer the phrase “empathicneutrality” as a point of departure. It sug-gests that there is a middle ground betweenbecoming too involved, which can cloudjudgment, and remaining too distant, whichcan reduce understanding. What is em-pathic neutrality? Consider this anecdote byway of illustration.

Pragmatist philosopher William James,also a scholar of anatomy and psychology,had a great capacity for empathy, as dis-played in his classic study The Varieties of Re-ligious Experience ([1902] 1999). Editor Clif-ton Fadiman (1985:305) recounts that whilehe was teaching at Radcliffe, Gertrude Steintook a course from him in which, having at-tended the opera and then partied into thewee hours the night before an exam, shewrote, “Dear Professor James, I am so sorrybut I do not feel a bit like writing an exami-nation paper today.” James is said to havewritten back: “Dear Miss Stein, I understandperfectly. I often feel like that myself.” Hadhe added, but the exam is still due, instead ofordinary sympathy he would have dis-played extraordinary empathic neutrality.

Methodologists and philosophers of sci-ence debate what the researcher’s stanceshould be vis-à-vis the people being studied.Critics of qualitative inquiry have chargedthat the approach is too subjective, in largepart because the researcher is the instrumentof both data collection and data interpreta-tion and because a qualitative strategy in-cludes having personal contact with andgetting close to the people and situation un-der study. From the perspective of advocates

of a supposedly value-free social science,subjectivity is the very antithesis of scientificinquiry.

Objectivity has been considered thestrength of the scientific method. The pri-mary methods for achieving objectivity inscience have been conducting blind experi-ments and quantification. “Objective tests”gather data through instruments that, inprinciple, are not dependent on human skill,perception, or even presence. Yet, it is clearthat tests and questionnaires are designedby human beings and therefore are subject tothe intrusion of the researcher’s biases bythe very questions asked. Unconscious biasin the skillful manipulation of statistics toprove a hypothesis in which the researcherbelieves is hardly absent from hypothetical-deductive inquiry.

Part of the difficulty in thinking about thefieldwork stance of the qualitative inquireris that the terms objectivity and subjectivityhave become so loaded with negative con-notations and subject to acrimonious debate(e.g., Scriven 1972a; Borman and Goetz 1986;Krenz and Sax 1986; Guba 1991) that neitherterm any longer provides useful guidance.These terms have been politicized beyondutility. To claim the mantle of “objectivity” inthe postmodern age is to expose oneself asembarrassingly naive. The ideals of absoluteobjectivity and value-free science are impos-sible to attain in practice and are of question-able desirability in the first place since theyignore the intrinsically social nature and hu-man purposes of research. On the otherhand, subjectivity has such negative connota-tions in the public mind that to admit beingsubjective may undermine one’s credibilitywith audiences unsophisticated aboutphenomenological assumptions and nu-ances. In short, the terms objectivity and sub-jectivity have become ideological ammuni-tion in the methodological paradigms


debate. My pragmatic solution is to avoidusing either word and to stay out of futile de-bates about subjectivity versus objectivity.Qualitative research in recent years hasmoved toward preferring such language astrustworthiness and authenticity. Evaluatorsaim for “balance,” “fairness,” and “com-pleteness” (Patton 1997a:282). Chapter 9 willdiscuss these terms and the stances they im-ply at greater length. At this point, I simplywant to note the strategic nature of the issueof inquirer stance and add empathic neutral-ity to the emerging lexicon that attempts tosupersede the hot button term objective andthe epithet subjective.

Any research strategy ultimately needscredibility to be useful. No credible researchstrategy advocates biased distortion of datato serve the researcher’s vested interests andprejudices. Both qualitative/naturalistic in-quiry and quantitative/experimental in-quiry seek honest, meaningful, credible, andempirically supported findings. Any credi-ble research strategy requires that the inves-tigator adopt a stance of neutrality with re-gard to the phenomenon under study. Thissimply means that the investigator does notset out to prove a particular perspective ormanipulate the data to arrive at predisposedtruths. The neutral investigator enters the re-search arena with no ax to grind, no theoryto prove (to test but not to prove), and nopredetermined results to support. Rather,the investigator’s commitment is to under-stand the world as it unfolds, be true to com-

plexities and multiple perspectives as theyemerge, and be balanced in reporting bothconfirmatory and disconfirming evidencewith regard to any conclusions offered.

Neutrality is not an easily attainablestance, so all credible research strategies in-clude techniques for helping the investiga-tor become aware of and deal with selectiveperception, personal biases, and theoreticalpredispositions. Qualitative inquiry, be-cause the human being is the instrument ofdata collection, requires that the investigatorcarefully reflect on, deal with, and report po-tential sources of bias and error. Systematicdata collection procedures, rigorous train-ing, multiple data sources, triangulation, ex-ternal reviews, and other techniques to bediscussed in this book are aimed at produc-ing high-quality qualitative data that arecredible, trustworthy, authentic, balancedabout the phenomenon under study, andfair to the people studied.

The livelihood of evaluators and re-searchers depends on their integrity andcredibility. Independence and neutrality,then, are serious issues.

However, neutrality does not mean de-tachment. It is on this point that qualitativeinquiry makes a special contribution. Quali-tative inquiry depends on, uses, and en-hances the researcher’s direct experiences inthe world and insights about those ex-periences. This includes learning throughempathy.

Empathy and Insight

The idea of acquiring an “inside” understanding—the actors’ definitionsof the situation—is a powerful central concept for understanding the

purpose of qualitative inquiry.

—Thomas A. Schwandt (2000:102)


Empathy develops from personal contactwith the people interviewed and observedduring fieldwork. Empathy involves beingable to take and understand the stance, posi-tion, feelings, experiences, and worldviewof others. Put metaphorically, empathy is“like being able to imagine a life for a spider,a maker’s life, or just some aliveness in itswide abdomen and delicate spinnerets soyou take it outside in two paper cups insteadof stepping on it” (Dunn 2000:62). Empathycombines cognitive understanding with af-fective connection, and in that sense differsfrom sympathy, which is primarily emo-tional (Wispé 1986).

The value of empathy is emphasized inthe phenomenological doctrine of Verstehenthat undergirds much qualitative inquiry.Verstehen means “understanding” and refersto the unique human capacity to make senseof the world. This capacity has profound im-plications for how one studies human be-ings. The Verstehen doctrine presumes thatsince human beings have a unique type ofconsciousness, as distinct from other formsof life, the study of human beings will be dif-ferent from the study of other forms of lifeand nonhuman phenomena. The capacityfor empathy, then, is one of the major assetsavailable for human inquiry into human af-fairs.

The Verstehen premise asserts that humanbeings can and must be understood in amanner different from other objects of studybecause humans have purposes and emo-tions; they make plans, construct cultures,and hold values that affect behavior. Theirfeelings and behaviors are influenced byconsciousness, deliberation, and the capac-ity to think about the future. Human beingslive in a world that has special meaning tothem, and because their behavior has mean-ing, “human actions are intelligible in waysthat the behavior of nonhuman objects isnot” (Strike 1972:28). Human and social sci-

ences need methods different from thoseused in agricultural experimentation andphysical sciences because human beings aredifferent from plants and nuclear particles.The Verstehen tradition stresses understand-ing that focuses on the meaning of humanbehavior, the context of social interaction, anempathic understanding based on personalexperience, and the connections betweenmental states and behavior. The tradition ofVerstehen places emphasis on the human ca-pacity to know and understand othersthrough empathic introspection and reflec-tion based on direct observation of and in-teraction with people. “Verstehen thus en-tails a kind of empathic identification withthe actor. It is an act of psychologicalreenactment— getting inside the head of anactor to understand what he or she is up to interms of motives, beliefs, desires, thoughts,and so on” (Schwandt 2000:192).

Max Weber brought the term empathy intosocial science to emphasize the importanceof comprehending the motives and feelingsof people in a social-cultural context.


The following story is passed around amongmanagement consultants. It seems thatvery late at night the president of a multi-national corporation was standing in frontof a paper shredder, trying to figure out howto turn it on. The building was deserted ex-cept for the cleaning staff, one of whomhappened by. The president asked for help,explaining that what he was working onwas very important and couldn’t wait untilmorning. The cleaning person was glad to beof help, turned on the machine, took thesheet of paper from the president, and fed itinto the shredder just as the president said,“I only need you to make one copy.”

Both Verstehen and empathy depend largelyon qualitative data. Verstehen is an attempt to“crack the code” of the culture, that is, detectthe categories into which a culture codes ac-tions and thoughts. . . . Empathy in evaluationis the detection of emotions manifested in theprogram participants and staff, achieved byevaluators’ becoming aware of similar or com-plementary emotions in themselves. (Meyers1981:180)

A qualitative strategy of inquiry proposesan active, involved role for the social scien-tist. “Hence, insight may be regarded as thecore of social knowledge. It is arrived at bybeing on the inside of the phenomena to beobserved. . . . It is participation in an activitythat generates interest, purpose, point ofview, value, meaning, and intelligibility, aswell as bias” (Wirth 1949:xxii). This is a quitedifferent scientific process from that envi-sioned by the classical, experimental ap-proach to science, but it is still an empirical,(i.e., data-based), scientific perspective. Thequalitative perspective “in no way suggeststhat the researcher lacks the ability to be sci-entific while collecting the data. On the con-trary, it merely specifies that it is crucial forvalidity—and, consequently, for reliability—to try to picture the empirical social worldas it actually exists to those under investiga-tion, rather than as the researcher imagines itto be” (Filstead 1970:4), thus the importanceof such qualitative approaches as partici-pant observation, depth interviewing, de-tailed description, and case studies.

These qualitative inquiry methods pro-vide opportunities to achieve empathy andgive the researcher an empirical basis fordescribing the perspectives of others. Chap-ter 1 cited the framework of humanistic psy-chologist Clark Moustakas, who has de-scribed this nonjudgmental empathic stanceas “Being-In” another’s world—immers-ing oneself in another’s world by listening

deeply and attentively so as to enter into theother person’s experience and perception.

I do not select, interpret, advise, or direct. . . .Being-In the world of the other is a way of go-ing wide open, entering in as if for the firsttime, hearing just what is, leaving out my ownthoughts, feelings, theories, biases. . . . I enterwith the intention of understanding and ac-cepting perceptions and not presenting myown view or reactions. . . . I only want to en-courage and support the other person’s ex-pression, what and how it is, how it came to be,and where it is going. (Moustakas 1995:82-83)

At first, the phrase “empathic neutrality”may appear to be an oxymoron, combiningcontradictory ideas. Empathy, however, de-scribes a stance toward the people one en-counters—it communicates understanding,interest, and caring. Neutrality suggests astance toward their thoughts, emotions, andbehaviors—it means being nonjudgmental.Neutrality can actually facilitate rapportand help build a relationship that supportsempathy by disciplining the researcher to beopen to the other person and nonjudgmentalin that openness. Rapport and empathy,however, must not be taken for granted, asRadhika Parameswaran (2001) found in do-ing fieldwork among young middle-classwomen in urban India who read Western ro-mance fiction.

Despite their eventual willingness to sharetheir fears and complaints about gendered so-cial pressures, I still wonder whether theseyoung women would have been more openabout their sexuality with a Westerner whomight be seen as less likely to judge thembased on cultural expectations of women’s be-havior in Indian society. The well-knownword rapport, which is often used to signifyacceptance and warm relationships betweeninformants and researchers, was thus some-


thing I could not take for granted despite be-ing an insider; all I could claim was animperfect rapport. (Parameswaran 2001:69)

Evaluation presents special challengesfor rapport and neutrality as well. Afterfieldwork, an evaluator may be called on torender judgments about a program as part of

data interpretation and formulating recom-mendations, but during fieldwork, the focusshould be on rigorously observing and inter-viewing to understand the people and situa-tion being studied. This nuanced relation-ship between neutrality and empathy willbe discussed further in both the data collec-tion and analysis chapters.

A Dynamic, Developmental Perspective

There is nothing permanent except change.

—Heraclitus (Ancient Greece)

A questionnaire is like a photograph. Aqualitative study is like a documentary film.Both offer images. One, however— the pho-tograph—captures and freezes a moment intime, like recording a respondent’s answerto a survey question at a moment in time.The other—the film— offers a fluid sense ofdevelopment, movement, and change.

Qualitative evaluation researchers, forexample, conceive of programs as dynamicand developing, with “treatments” chang-ing in subtle but important ways as stafflearns what does and doesn’t work, as cli-ents move in and out, and as conditions ofdelivery are altered. A primary challenge,then, becomes describing and understand-ing these dynamic program processes andtheir holistic effects on participants so as toprovide information for program improve-ment. In contrast, an experimental designfor an evaluation typically conceives of theprogram as a fixed thing, like a measuredamount of fertilizer applied to a crop—atreatment, an intervention—that has prede-termined, measurable outcomes. Inconsis-tency in the treatment, instability in theintervention, changes in the program, vari-ability in program processes, and diversity

in participants’ experiences undermine thelogic of an experimental design becausethese developments—all natural, even inev-itable, in real-world programs—call intoquestion what the “treatment” or experi-ment actually is.

Naturalistic inquiry assumes the ever-changing world posited by the observationin the ancient Chinese proverb that onenever steps into the same river twice.Change is a natural, expected, and inevita-ble part of human experience, and docu-menting change is a natural, expected, andintrinsic part of fieldwork. Rather than try-ing to control, limit, or direct change, natu-ralistic inquirers expect change, anticipatethe likelihood of the unanticipated, and areprepared to go with the flow of change. Onegets this sense of pursuing change in thecomment by anthropologist Williams citedearlier: “I’m tracking something—killing—that’s moving very rapidly in the culture.”Part of her inquiry task is to track culturalchanges the way an epidemiologist tracks adisease. As a result, reading a good qualita-tive case study gives the sense of reading agood story. It has a beginning, a middle, andan ending—though not necessarily an end.


� Analysis Strategiesfor Qualitative Inquiry

Unique Case Orientation

“Six windows on respect” is how Har-vard sociologist Sara Lawrence-Lightfoot(2000:13) describes the six detailed casestudies, each a full chapter, she presents inher book Respect. The cases offer differentangles on the meaning and experience of re-spect in modern society as illuminated by anurse-midwife, a pediatrician, a teacher, anartist, a law school professor, and a pastoraltherapist/AIDS activist. Before drawingthemes and contrasts from this small, pur-poseful sample, and before naming the sixangles they represent, Lawrence-Lightfoothad the task of constructing the unique casesto tell these distinct stories. Her first task,then, was to undertake the “art and scienceof portraiture” (Lawrence-Lightfoot andDavis 1997). From these separate portraits,she fashions a stained glass mosaic that de-picts and illuminates Respect.

I undertook a study of a national fellow-ship award program that had had more than600 recipients over a 20-year period. A sur-vey had been done to get the fellows’ opin-ions about select issues, but the staff wantedmore depth, richness, and detail to really un-derstand patterns of fellowship use and im-pact. With a team of researchers, we con-ducted 40 in-depth, face-to-face interviewsand wrote case studies. Through inductiveanalysis, we subsequently identified dis-tinct enabling processes and impacts, andwe created a framework that depicted rela-tionships between status at the time of theaward, enabling processes, and impacts. Theheart of the study remained the 40 case stud-ies. To read only the framework analysiswithout reading the case studies would be tolose much of the richness, depth, meaning,and contribution of qualitative research and

evaluation. That is what is meant by the“unique case orientation” of qualitative in-quiry.

Case studies are particularly valuable inprogram evaluation when the program is in-dividualized, so the evaluation needs to beattentive to and capture individual differ-ences among participants, diverse experi-ences of the program, or unique variationsfrom one program setting to another. Asnoted earlier, a case can be a person, anevent, a program, an organization, a timeperiod, a critical incident, or a community.Regardless of the unit of analysis, a qual-itative case study seeks to describe that unitin depth and detail, holistically, and incontext.

Inductive Analysisand Creative Synthesis

Benjamin Whorf’s development of the fa-mous Whorf hypothesis—that languageshapes our experience of the environmentand that words shape perceptions and ac-tions, a kind of linguistic relativity theory(Schultz 1991)—provides an instructive ex-ample of inductive analysis. Whorf was aninsurance investigator assigned to look intoexplosions in warehouses. He discoveredthat truck drivers were entering “empty”warehouses smoking cigarettes and cigars.The warehouses, it turned out, often con-tained invisible, but highly flammablegases. He interviewed truckers and foundthat they associated the word empty withharmless and acted accordingly. From thesespecific observations and findings, he induc-tively formulated his general theory aboutlanguage and perception that has informeda half-century of communications scholar-ship (Lee 1996).

Qualitative inquiry is particularly ori-ented toward exploration, discovery, and in-ductive logic. Inductive analysis begins with


specific observations and builds towardgeneral patterns. Categories or dimensionsof analysis emerge from open-ended obser-vations as the inquirer comes to understandpatterns that exist in the phenomenon beinginvestigated.

Inductive analysis contrasts with the hy-pothetical-deductive approach of experi-mental designs that require the specificationof main variables and the statement of spe-cific research hypotheses before data collec-tion begins. A specification of research hy-potheses based on an explicit theoreticalframework means that general constructsprovide the framework for understandingspecific observations or cases. The investiga-tor must then decide in advance what vari-

ables are important and what relationshipsamong those variables can be expected.

The strategy of inductive designs is to al-low the important analysis dimensions toemerge from patterns found in the cases un-der study without presupposing in advancewhat the important dimensions will be. Thequalitative analyst seeks to understand themultiple interrelationships among dimen-sions that emerge from the data withoutmaking prior assumptions or specifying hy-potheses about the linear or correlative rela-tionships among narrowly defined, opera-tionalized variables. For example, an induc-tive approach to program evaluation meansthat understanding the nature of the “inter-vention” emerges from direct observationsof program activities and interviews withparticipants. In general, theories about whatis happening in a setting are grounded inand emerge from direct field experiencerather than being imposed a priori as is thecase in formal hypothesis and theory testing.

The straightforward contrast betweenclosed-ended questionnaires and open-ended interviews in Chapter 1 illustratedthe difference between deductive and in-ductive approaches at the simplest level. Astructured, multiple-choice questionnairerequires a deductive approach becauseitems must be predetermined based on sometheory or preordinate criteria, for example,program goals about what is important tomeasure. An open-ended interview, by wayof contrast, permits the respondent to de-scribe what is meaningful and salient with-out being pigeon holed into standardized cat-egories.

In practice, these approaches are oftencombined. Some evaluation or researchquestions may be determined deductively,while others are left sufficiently open topermit inductive analyses based on directobservations. While the quantitative/exper-imental approach is largely hypothetical-


QUALITATIVECOMPARATIVE ANALYSIS

Understanding unique cases can be deep-ened by comparative analysis. Indeed, con-stant comparative analysis is a central ana-lytical approach in grounded theory, one ofthe major schools of qualitative inquiry (dis-cussed in Chapters 3 and 8). Comparisonscan also be important in illuminating differ-ences between programs in evaluation. In-deed, evaluation theorist Michael Scriven(1993) has asserted in his monographHard-Won Lessons in Program Evaluationthat “noncomparative evaluations arecomparatively useless” (p. 58).

Leonardo da Vinci’s (ca. 1519) insightinto color contrasts offers a metaphor forqualitative comparative analysis: “In orderto attain a color of the greatest possibleperfection, one has to place it in the neigh-borhood of the directly contrary color: thusone places black with white, yellow withblue, green with red” (cited in Boring 1942).

deductive and the qualitative/naturalisticapproach is largely inductive, a study can in-clude elements of both strategies. Indeed,over a period of inquiry, an investigationmay flow from inductive approaches, to findout what the important questions and vari-ables are (exploratory work), to deductivehypothesis-testing or outcome measure-ment aimed at confirming and/or generaliz-ing exploratory findings, then back again toinductive analysis to look for rival hypo-theses and unanticipated or unmeasuredfactors.

The precise nature of inductive analysisdepends, in part, on the purpose of the anal-ysis and the number and types of cases in astudy. Where there are several cases to becompared and contrasted, an inductive ap-proach begins by constructing individualcases, without pigeon holing or categorizingthose cases. That is, the first task is to do acareful job independently writing up the

separate cases. Once that is done, cross-caseanalysis can begin in search of patterns andthemes that cut across individual experi-ences. The initial focus is on full understand-ing of individual cases before those uniquecases are combined or aggregated themati-cally. This helps ensure that emergent cate-gories and discovered patterns are groundedin specific cases and their contexts (Glaserand Strauss 1967).

Just as writers report different creativeprocesses, so too qualitative analysts havedifferent ways of working. Although soft-ware programs now exist to facilitate work-ing with large amounts of narrative data andsubstantial guidance can be offered aboutthe steps and processes of content analysis,making sense of multiple interview tran-scripts and pages of field notes cannot be re-duced to a formula or even a standard seriesof steps. There is no equivalent of a statisticalsignificance test or factor score to tell the an-


©20

02M

icha

elQ

uinn

Patt

onan

dM

icha

elC

ochr

an

alyst when results are important or whatquotations fit together under the sametheme. Finding a way to creatively synthesizeand present findings is one of the challengesof qualitative analysis, a challenge that willbe explored at length in Part 3 of this book.For the moment, I can offer a flavor of thatchallenge with another excerpt from my in-terview with anthropologist Williams. Hereshe describes part of her own unique ana-lytic process.

My current project follows up work that I havealways done, which is to study categories andclassifications and their implications. Rightnow, as I said, the focus of my work is on kill-ing and the desire to kill and the categoriespeople create in relation to killing. Part of itright now focuses on the death penalty, butmainly on killing. My fascination is with thelinks between category distinctions, commit-ments, and the desire to kill for those commit-ments. That’s what I study.

I track categories, like “serial killers” or“death row inmates.” The business of con-stantly transforming people into acts and actsinto people is part of the way loyalties, com-mitments, and hatreds are generated. So I’m aclassifier. I study classification—theories ofclassification. A lot of categories have to dowith very abstract things; others have to dowith very concrete things like skin color. Butultimately, the classification of a kill is whatI’m focusing on now. I’ve been asking myselflately, for the chapter I’ve been working on, “Isthere a fundamental difference, for example,in the way we classify to kill?” Considerthe percentage of people classified as “deathworthy”—the way we classify to justify thedeath penalty.

As I write, moving back and forth betweenmy tapes and my interviews, I don’t feel that Ihave to follow some fixed outline or that Ihave to code things to come out a certain way.Sometimes I listen to a tape and I start to think

that I should rewrite this part of this chapter. Ihad completely forgotten about this tape. Itwas done in early ‘98 or late ‘97 and maybe Ihadn’t listened to it or looked at the transcriptfor a while, and I’ve just finished a chapter orsection of a chapter. I pull that tape off theshelf. I listen to it. I go back to the transcriptand I start writing again. I start revising inways that it seems to me that tape demands.

As Williams describes her analysis andwriting process, she offers insight into whatit means when qualitative researchers saythey are “working to be true to the data” orthat their analytical process is “data driven.”Williams says, “I start revising in ways that itseems to me that tape demands.” It is com-mon to hear qualitative analysts say that, asthey write their conclusions, they keep go-ing back to the cases; they reread field notes;and they listen again to interviews. Induc-tive analysis is built on a solid foundationof specific, concrete, and detailed observa-tions, quotations, documents, and cases.As thematic structures and overarchingconstructs emerge during analysis, thequalitative analyst keeps returning to field-work observations and interview tran-scripts, working from the bottom up, stay-ing grounded in the foundation of casewrite-ups, and thereby examining emergentthemes and constructs in light of what theyilluminate about the case descriptions onwhich they are based. That is inductiveanalysis.

Holistic Perspective

Holography is a method of photographyin which the wave field of light scattered byan object is captured as an interference pat-tern. When the photographic record—thehologram—is illuminated by a laser, a three-dimensional image appears. Any piece of ahologram will reconstruct the entire image.


This has become a metaphor for thinking innew ways about the relationships betweenparts and wholes. The interdependence offlora, fauna, and the physical environmentin ecological systems offers another meta-phor for what it means to think and analyzeholistically.

Researchers and evaluators analyzingqualitative data strive to understand a phe-nomenon or program as a whole. Thismeans that a description and interpretationof a person’s social environment, or an orga-nization’s external context, is essential foroverall understanding of what has been ob-served during fieldwork or said in an inter-view. This holistic approach assumes thatthe whole is understood as a complex sys-tem that is greater than the sum of its parts.The analyst searches for the totality or unify-ing nature of particular settings—the gestalt.Psychotherapist Fritz Perls (1973) made theterm gestalt equivalent with a holistic per-spective in psychology. He used the exampleof three sticks that are just three sticks untilone places them together to form a triangle.Then they are much more than the three sep-arate sticks combined: They form a newwhole.

A gestalt may be a tangible thing, such as a tri-angle, or it may be a situation. A happeningsuch as a meeting of two people, their conver-sation, and their leave-taking would consti-tute a completed situation. If there were aninterruption in the middle of the conversation,it would be an incomplete gestalt. (Brown1996:36)

The strategy of seeking gestalt units andholistic understandings in qualitative analy-sis contrasts with the logic and proceduresof evaluation studies conducted in the ana-lytical tradition of “let’s take it apart and seehow it works.” The quantitative-experimen-tal approach to evaluation, for example, re-

quires operationalization of independentand dependent variables with a focus ontheir statistical covariance. Outcomes mustbe identified and measured as specific vari-ables. Treatments and programs must alsobe conceptualized as discrete, independentvariables. The characteristics of programparticipants are also described by standard-ized, quantified dimensions. Sometimes thevariables of interest are derived from pro-gram goals, for example, student achieve-ment test scores, recidivism statistics for agroup of juvenile delinquents, sobriety ratesfor participants in chemical dependencytreatment programs. At other times, thevariables measured are indicators of a largerconstruct. For example, community well-being may be measured by such rates for de-linquency, infant mortality, divorce, unem-ployment, suicide, and poverty (Brock,Schwaller, and Smith 1985). These variablesare statistically manipulated or added to-gether in some linear fashion to test hypoth-eses and draw inferences about the relation-ships among separate indicators, or thestatistical significance of differences be-tween measured levels of the variables fordifferent groups. The essential logic of thisapproach is as follows: (1) Key program out-comes and processes can be represented byseparate independent variables, (2) thesevariables can be quantified, and (3) relation-ships among these variables are best por-trayed statistically.

The primary critique of this logic by qual-itative-naturalistic evaluators is that such anapproach (1) oversimplifies the complexitiesof real-world programs and participants’ ex-periences, (2) misses major factors of impor-tance that are not easily quantified, and (3)fails to portray a sense of the program and itsimpacts as a whole. To support holistic anal-ysis, the qualitative inquirer gathers data onmultiple aspects of the setting under studyto assemble a comprehensive and complete


picture of the social dynamic of the particu-lar situation or program. This means that atthe time of data collection, each case, event,or setting under study, though treated as aunique entity with its own particular mean-ing and its own constellation of relation-ships emerging from and related to the con-text within which it occurs, is also thought ofas a window into the whole. Thus capturingand documenting history, interconnections,and system relationships are part of field-work.

The advantages of using quantitativevariables and indicators are parsimony, pre-cision, and ease of analysis. Where key ele-ments can be quantified with validity, reli-ability, and credibility, and where necessarystatistical assumptions can be met (e.g., lin-earity, normality, and independence of mea-

surement), then statistical portrayals can bequite powerful and succinct. The advan-tages of qualitative portrayals of holistic set-tings and impacts are that greater attentioncan be given to nuance, setting, interdepen-dencies, complexities, idiosyncrasies, andcontext. John Dewey (1956) articulated whata holistic approach means for both teachingand research if one wants to gain insight intoand understand the world of the child:

The child’s life is an integral, a total one. Hepasses quickly and readily from one topic toanother, as from one spot to another, but is notconscious of transition or break. There is noconscious isolation, hardly conscious distinc-tion. The things that occupy him are held to-gether by the unity of the personal and socialinterests which his life carries along. . . . [His]


HOLISTIC UNDERSTANDING IN GENETICS

The work of geneticist Barbara McClintock,winner of the Nobel prize, illustrates holisticinquiry. She sought to understand the organi-zation and functions of genes in relation tothe rest of the cell, within the organism as awhole:

She sought a feel for the whole. Ratherthan dismissing exceptional cases as irrel-evant to general theory, she focused onanomalous pigmentation of individualplants. Instead of starting with an hypoth-esis prescribing what she expected andframing the questions for the material toanswer, as in most controlled experiments,she felt the need to “let the experiment tellyou what it wants to do” and to “listen tothe material.”

She followed each unique seedlingthrough its life in the field. . . . At times, shebecame so engrossed in examining indi-vidual cells in a grain of corn through hermicroscope that she felt as if she weredown there within the cell, the same size as

the chromosomes, and could see how they wereinteracting. In such ways, she developed whatshe called a “feeling for the organism.”

Her unique understanding let her to questionthe genetic theory of Watson and Crick: thatDNA contained a cell’s vital information, whichwas copied onto the RNA and acted as a blue-print for genetic traits. She thought that this“master molecule theory” claimed to explain toomuch and did not acknowledge the differencesbetween small simple organisms and largecomplex multicellular ones. More important, ittreated DNA as a central autonomous actor,sending out information one way, through a ge-netic organization structure hierarchically, likea classic bureaucracy. McClintock showed thatgenetic organization is more complex and in-terdependent. The DNA itself adapts to outsidefactors and can be reprogrammed by signalsfrom the environment to meet the survivalneeds of the organism. In essence, informationflows both ways. (Schmidt 1993: 528)

universe is fluid and fluent; its contents dis-solve and reform with amazing rapidity. Butafter all, it is the child’s own world. It has theunity and completeness of his own life.(pp. 5-6)

Qualitative sociologist Irwin Deutscher(1970) commented that despite the totality ofour personal experiences as living, workinghuman beings, social scientists have tendedto focus their research on parts to the virtualexclusion of wholes:

We knew that human behavior was rarely ifever directly influenced or explained by anisolated variable; we knew that it was impos-sible to assume that any set of such variableswas additive (with or without weighting); weknew that the complex mathematics of the in-teraction among any set of variables was in-comprehensible to us. In effect, although weknew they did not exist, we defined them intobeing. (p. 33)

While many would view this intense cri-tique of variable analysis as too extreme, thereaction of many program staff to scientificresearch is like the reaction of Copernicus tothe astronomers of his day: “With them,” heobserved, “it is as though an artist were togather the hands, feet, head, and other mem-bers for his images from diverse models,each part excellently drawn, but not relatedto a single body, and since they in no waymatch each other, the result would be mon-ster rather than man” (from Kuhn 1970:83).

How many program staffs have complainedof the evaluation research monster?

It is no simple task to undertake holisticanalysis. The challenge is “to seek the es-sence of the life of the observed, to sum up,to find a central unifying principle” (Bruyn1966:316). Again, Shapiro’s (1973) work inevaluating innovative Follow Through class-rooms is instructive. She found that stan-dardized test results could not be inter-preted without understanding the largercultural and institutional context in whichthe individual child is situated. Taking con-text seriously, the topic of the next section, isan important element of holistic analysis.

An illuminative example of holistic think-ing came to me from a Portuguese colleague.He told of driving in a remote area of hiscountry when he came upon a sizable herdof sheep being driven along the road by ashepherd. Seeing that he would be delayeduntil the sheep could be turned off the road,he got out of the car and struck up a conver-sation with the shepherd.

“How many sheep do you have?” heasked.

“I don’t know,” responded the youngman.

Surprised at this answer, the travelerasked, “How do you keep track of the flock ifyou don’t know how many sheep there are?How would you know if one was missing?”

The shepherd seemed puzzled by thequestion. Then he explained, “I don’t needto count them. I know each one and I knowthe whole flock. I would know if the flockwas not whole.”

Context Sensitivity

A ny single act from any single person, put out of context, isdamnable.

—Actor Kevin Spacey accepting the 2000 Academy Awardfor Best Performance by an Actor in the film

American Beauty, explaining the film’s message


Let’s move, now, from sheep to elephants.One of the classic tales used to illustrate therelationship between parts and wholes is thestory of the nine blind people and the ele-phant. Each person touches only one part ofthe elephant and therefore knows only thatpart. The person touching the ears thinks anelephant is like a large, thin fan. The persontouching the tail thinks the elephant is like arope. The person touching the truck thinksof a snake. The legs feel like tree trunks, theelephant’s side like a tall wall. And so itgoes. The holistic point is that one must putall of these perspectives together to get a fullpicture of what an elephant actually lookslike.

But such a picture will still be limited,even distorted, if the only place one sees theelephant is in the zoo or at the circus. To un-derstand the elephant—how it developed,how it uses its trunk, why it is so large—onemust see it on the African savanna or in theAsian jungle. In short, one must see it in con-

text and as part of an ecological system in re-lation to other flora and fauna, in its naturalenvironment.

When we say to someone, “You’ve takenmy comment out of context,” we are saying,You have distorted what I said, changed itsmeaning by omitting critical context.

In Victor Hugo’s great classic LesMisérables, we first encounter Jean Valjean asa hardened criminal and common thief; thenwe learn that he was originally sentenced tofive years in prison for stealing a loaf ofbread for his sister’s starving family. Thatadded context for his “crime” changes ourunderstanding. The battle over standard-ized sentencing guidelines in the criminaljustice system is partly a debate about howmuch to allow judges sway in taking into ac-count context and individual circumstancesin pronouncing sentences.

Naturalistic inquiry preserves naturalcontext. Social psychology experiments un-der laboratory conditions strip observed ac-


©20

02M

icha

elQ

uinn

Patt

onan

dM

icha

elC

ochr

an

tions from context. But that is the point ofsuch laboratory experiments—to generatefindings that are context free. The scientificideal of generalizing across time and spaceis the ideal of identifying principles that donot depend on context. In contrast, qual-itative inquiry elevates context as critical tounderstanding. Portraitist Sara Lawrence-Lightfoot (1997) explains why she finds con-text “crucial to the documentation of humanexperience and organizational culture”:

By context, I mean the setting—physical, geo-graphic, temporal, historical cultural, aesthetic—within which action takes place. Context be-comes the framework, the reference point, themap, the ecological sphere; it is used to placepeople and action in time and space and as aresource for understanding what they say anddo. The context is rich in clues for interpretingthe experience of the actors in the setting. Wehave no idea how to decipher or decode anaction, a gesture, a conversation, or an excla-mation unless we see it embedded in context.(p. 41)

Voice and Perspective:Reflexivity

ABSTRACT OF A SCHOOLACHIEVEMENT STUDY

This study will delineate the major factors thataffect school achievement. Instruments wereselected to measure achievement based on va-lidity and reliability criteria. Decisions weremade about administering the tests in con-junction with administrators taking into ac-count time and resource constraints. Aregression model was constructed to test rela-tionships between various background vari-ables and demonstrated achievement. Schoolrecords were reviewed and coded to ascertainstudents’ background characteristics. Data

were obtained on 120 students from fourclassrooms. The extraction of significant pre-dictor variables is the purpose of the finalanalysis. Interviews were conducted withteachers and principals to determine how testscores were used. The analysis concludes withthe researcher’s interpretations. The research-er wishes to thank those who cooperated inthis study.

This journal article abstract represents ac-ademic writing as I was taught to do it ingraduate school. This writing style still pre-dominates in scholarly journals and books.No human being is visible in this writing.The passive voice reigns. Instruments wereselected; decisions were made; a model wasconstructed; records were reviewed andcoded; data were obtained; predictor vari-ables were extracted; interviews were con-ducted. The warmth of thanks is extendedby a role, the researcher: “The researcherwishes to thank those who cooperated.” Thethird-person, passive voice communicates amessage: This work is about procedures notpeople. This academic style is employed toproject a sense of objectivity, control, and au-thority. The overall impression is mechani-cal, robotlike, distant, detached, systematic,and procedural. The research is the object ofattention. Any real, live human being, sub-ject to all the usual foibles of being human, isbarely implied, generally disguised, hiddenaway, and kept in the background.

Contrast that academic voice with myexplanation of how I analyzed a 10-daycoming-of-age experience with my son inthe Grand Canyon. (I presented part of theanalysis of that experience as Exhibit 1.3 inthe first chapter.) Here’s an excerpt in whichI describe the analytical process.

I’m not sure when the notion first took hold ofme that articulating alternative coming of ageparadigms might help elucidate our Canyon


experience. Before formally conceptualizingcontrasting paradigm dimensions, I experi-enced them as conflicting feelings that ema-nated from my struggle to sort out what Iwanted my son’s initiation to be, while alsograppling with defining my role in the pro-cess. I suppose the idea of alternative para-digms first emerged the second night as Ipaced the narrow beach where White Creekintersects Shinumo and pondered the GreatUnconformity [a geologic reference] as meta-phor for the gap between tribal approaches toinitiation and coming of age for contemporaryyouth. In the weeks and months after our Can-yon experience, far from languishing in thethroes of retox as I expected, the idea of con-trasting paradigms stayed with me, as did theCanyon experience. I started listing themesand matching them with incidents and turn-ing points along the way. The sequence of inci-dents became this book and the contrastingthemes became the basis for this closing chap-ter, a way for me to figure out how whatstarted out as an initiation become a humanistcoming of age celebration. (Patton 1999a:332)

The contrast between the traditional aca-demic voice and the personal voice of quali-tative analysis recalls philosopher and theo-logian Martin Buber’s (1923) influentialdistinction between “I-It” and “I-Thou” re-lationships. An I-It relationship regardsother human beings from a distance, from asuperior vantage point of authority, as ob-jects or subjects, things in the environmentto be examined and placed in abstractcause-effect chains. An I-Thou perspective,in contrast, acknowledges the humanity ofboth self and others and implies relation-ship, mutuality, and genuine dialogue.

The perspective that the researcher bringsto a qualitative inquiry is part of the contextfor the findings. A human being is the in-strument of qualitative methods. A real, liveperson makes observations, takes field

notes, asks interview questions, and inter-prets responses. Self-awareness, then, can bean asset in both fieldwork and analysis. De-veloping appropriate self-awareness can bea form of “sharpening the instrument”(Brown 1996:42). The methods section of aqualitative study reports on the researcher’straining, preparation, fieldwork procedures,and analytical processes. This is both thestrength and weakness of qualitative meth-ods, the strength in that a well-trained, expe-rienced, and astute observer adds value andcredibility to the inquiry, while an ill-pre-pared, inexperienced, and imperceptive ob-server casts doubt on what is reported. Judg-ments about the significance of findings arethus inevitably connected to the researcher’scredibility, competence, thoroughness, andintegrity. Those judgments, precisely be-cause they are acknowledged as inevitablypersonal and perspective dependent, at leastto some extent, invite response and dia-logue, rather than just acceptance or rejec-tion.

Reflexivity has entered the qualitative lex-icon as a way of emphasizing the impor-tance of self-awareness, political/culturalconsciousness, and ownership of one’s per-spective.

In the rush of interest in qualitative research inthe past 15 years, few topics have developedas broad a consensus as the relevance of ana-lytic “reflexivity.” By most accounts, reflexivi-ty is a deconstructive exercise for locating theintersections of author, other, text, and world,and for penetrating the representational exer-cise itself. (MacBeth 2001:35)

Being reflexive involves self-questioningand self-understanding, for “all under-standing is self-understanding” (Schwandt1997a:xvi). To be reflexive, then, is to under-take an ongoing examination of what I knowand how I know it, “to have an ongoing con-


versation about experience while simulta-neously living in the moment” (Hertz 1997:viii). Reflexivity reminds the qualitative in-quirer to be attentive to and conscious of thecultural, political, social, linguistic, andideological origins of one’s own perspectiveand voice as well as the perspective andvoices of those one interviews and those towhom one reports. Exhibit 2.2 depicts thisreflexive triangulation.

Writing in the first-person, active voicecommunicates the inquirer’s self-aware rolein the inquiry: “I started listing themes andmatching them with incidents and turningpoints along the way.” The passive voicedoes not: “Themes were listed and matchedto incidents and turning points along theway.” Judith Brown (1996) captured the im-portance of the first-person voice in the titleof her book The I in Science: Training to UtilizeSubjectivity in Research. By subjectivity shemeans “the domain of experiential self-knowledge” (p. 1). Voice reveals and com-municates this domain.

But voice is more than grammar. A credi-ble, authoritative, authentic, and trustwor-thy voice engages the reader through richdescription, thoughtful sequencing, appro-priate use of quotes, and contextual clarityso that the reader joins the inquirer in thesearch for meaning. And there are choices ofvoice: the didactic voice of the teacher; thesearching, logical voice of the sleuth; the nar-rator voice of the storyteller; the personalvoice of the autoethnographer; the doubtingvoice of the skeptic; the intimacy of the in-sider’s voice; the detachment of the out-sider’s voice; the searching voice of uncer-tainty; and the excited voice of discovery, tooffer but a few examples. Just as point ofview and voice have become focal points ofwriting good fiction and nonfiction, as inNancy Mairs’s (1997) Voice Lessons: On Be-coming a (Woman) Writer, so too qualitativeanalysts are having to learn about, take into

account, and communicate perspective andvoice. Balancing critical and creative analy-ses, description and interpretation, or directquotation and synopsis also involves issuesof perspective, audience, purpose, andvoice. No rules or formula can tell a qualita-tive analyst precisely what balance is rightor which voice to use, only that finding bothbalance and voice is part of the work andchallenge of qualitative inquiry, what Lewis(2001) has acknowledged as “the difficultyof trying to situate the I in narrative re-search” (p. 109).

In addition to finding voice, the critical andcreative writing involved in qualitativeanalysis and synthesis challenge the in-quirer to own one’s voice and perspective. Here,we owe much to feminist theory for high-lighting and deepening our understandingof the intricate and implicate relationshipsbetween language, voice, and consciousness(e.g., Gilligan 1982; Minnich 1990). We arechallenged by postmodern critiques ofknowledge to be clear about and own ourauthorship of whatever we propound, to beself-reflective, to acknowledge biases andlimitations, and to honor multiple perspec-tives (Greene 1998a, 1998b; Mabry 1997)while “accepting incredulity and doubt asmodal postmodern responses to all at-tempts to explain ourselves to ourselves”(Schwandt 1997b:102). From struggles to lo-cate and acknowledge the inevitably politi-cal and moral nature of evaluative judg-ments, we are challenged to connect voiceand perspective to praxis—acting in theworld with an appreciation for and recog-nition of how those actions inherently ex-press social, political, and moral values(Schwandt 1989, 2000) and to personalizeevaluation (Kushner 2000), both by owningour own perspective and by taking seriouslythe responsibility to communicate authenti-cally the perspectives of those we encounterduring our inquiry. These represent some


of the more prominent contextual forcesthat have elevated the importance of own-ing voice and perspective in qualitativeanalysis.

It takes no great self-awareness or self-confidence to report a statistically signifi-cant t test with confidence intervals based ona formula and calculations easily replicatedand confirmed. It can take considerableself-awareness and confidence to report: Icoded these 40 interviews, these are thethemes I found, here is what I think theymean, and here is the process I undertook toarrive at those meanings. The latter state-ment calls for, even demands, a sense ofvoice and perspective.

� From Strategic Idealsto Practical Choices

The 12 themes of qualitative inquiry re-viewed in this chapter are strategic ideals:

real-world observations through naturalis-tic inquiry; openness, responsiveness, andflexibility through emergent designs; focusthrough purposeful sampling; richness anddepth through qualitative data; use of all ofone’s capacities through personal experi-ence and engagement; balancing the criticaland creative through a stance of empathicneutrality; sensitivity to dynamic processesand systems; appreciation of idiosyncrasiesthrough a unique case orientation; insightand understanding through inductive anal-ysis, contextual sensitivity, and a holisticperspective; and authenticity and trustwor-thiness through ownership of voice and per-spective. These are not absolute and univer-sal characteristics of qualitative inquiry, butrather strategic ideals that provide a direc-tion and framework for developing specificdesigns and concrete data collection tactics.

Ideally, a pure qualitative inquiry strat-egy includes all the themes and dimensionsidentified in this chapter. For example, in an


EXHIBIT 2.2 Reflexive Questions: Triangulated Inquiry

Those studied(participants):

How do they know whatthey know? What shapes and

has shaped their worldview?How do they perceive me?

Why? How do I know?How do I perceive them?

Reflexive screens:Culture, age, gender, class,social status, education,family, political praxis,

language, values

Those receiving thestudy (audience):

How do they makemake sense of what I give

them? What perspectivesdo they bring to the findings

I offer? How do they perceiveme? How do I perceive them?

Myself:(as qualitative inquirer):

What do I know?How do I know what I know?

What shapes and has shaped my perspective?With what voice do I share my perspective?

What do I do with what I have found?

ideal naturalistic/inductive inquiry the re-searcher neither manipulates the setting un-der study nor predetermines what variablesor categories are worth measuring. In prac-tice, however, it is important to recognizethat actually conducting holistic-inductiveanalysis and implementing naturalistic in-quiry are always a matter of degree. In mak-ing this point, Guba (1978) has depicted thepractice of naturalistic inquiry as a wave onwhich the investigator moves from varyingdegrees of a “discovery mode” to varyingemphasis of a “verification mode” in at-tempting to understand the real world. Asfieldwork begins, the inquirer is open towhatever emerges from the data, a discov-ery or inductive approach. Then, as the in-quiry reveals patterns and major dimen-sions of interest, the investigator will beginto focus on verifying and elucidating whatappears to be emerging—a more deductiveapproach to data collection and analysis. Inessence, what is discovered may be verifiedby going back to the world under study andexamining the extent to which the emergentanalysis fits the phenomenon and works toexplain what has been observed. Glaser andStrauss (1967), in their classic framing ofgrounded theory, described what it meansfor results to fit and work: “By ‘fit’ we meanthat the categories must be readily (not forc-ibly) applicable to and indicated by the dataunder study; by ‘work’ we mean that theymust be meaningfully relevant to and beable to explain the behavior under study”(p. 3). Discovery and verification mean mov-ing back and forth between induction anddeduction, between experience and reflec-tion on experience, and between greater de-grees and lesser degrees of naturalistic in-quiry.

In program evaluation in particular, theevaluator may, through feedback of initialfindings to program participants and staff,

begin to affect the program quite directlyand intentionally (given the job of helpingimprove the program), thus moving awayfrom a purely naturalistic approach. Asevaluative feedback is used to improve theprogram, the evaluator may then move backinto a more naturalistic stance to observehow the feedback-induced changes in theprogram unfold.

In the same vein, the attempt to under-stand a program or treatment as a wholedoes not mean that the investigator neverbecomes involved in component analysis orin looking at particular variables, dimen-sions, and parts of the phenomenon understudy. Rather, it means that the qualitativeinquirer consciously works back and forthbetween parts and wholes, separate vari-ables, and complex, interwoven constella-tions of variables in a sorting-out thenputting-back-together process. While stay-ing true to a strategy that emphasizes the im-portance of a holistic picture of the program,the qualitative evaluator recognizes thatcertain periods of fieldwork may focus oncomponent, variable, and less-than-the-whole kinds of analysis.

The practice and practicalities of field-work also mean that the strategic mandateto “get close” to the people and setting un-der study is neither absolute nor fixed.Closeness to and involvement with the peo-ple under study are most usefully viewed asvariable dimensions. The personal stylesand capabilities of evaluators will permitand necessitate variance along these dimen-sions. Variations in types of programs andevaluation purposes will affect the extent towhich an evaluator can or ought to get closeto the program staff and participants. More-over, closeness is likely to vary over thecourse of an evaluation. At times the eval-uator may become totally immersed in theprogram experience. These periods of im-


mersion may be followed by times ofwithdrawal and distance (for personal aswell as methodological reasons), to be fol-lowed still later by new experiences of im-mersion in and direct experience with theprogram.

Nor is it necessary to be a qualitativemethods purist. Qualitative data can becollected and used in conjunction withquantitative data. Today’s evaluator mustbe sophisticated about matching researchmethods to the nuances of particular evalua-tion questions and the idiosyncrasies of spe-cific stakeholder needs. Such an evaluatorneeds a large repertoire of research methodsand techniques to use on a variety of prob-lems. Thus, an evaluator may be called on touse any and all social science research meth-ods, including analyses of quantitative data,questionnaires, secondary data analysis,cost-benefit and cost-effectiveness analyses,standardized tests, experimental designs,unobtrusive measures, participant observa-tion, and in-depth interviewing. The evalua-tion researcher works with intended users ofthe findings to design an evaluation that in-cludes any and all data that will help shedlight on important evaluation questions,given constraints of resources and time.Such an evaluator is committed to researchdesigns that are relevant, meaningful, un-derstandable, and able to produce useful re-sults that are valid, reliable, and believable.On many occasions a variety of data collec-tion techniques and design approaches maybe used together. Multiple methods and avariety of data types can contribute to meth-odological rigor. The ideal in evaluation de-signs is methodological appropriateness,design flexibility, and situational respon-siveness in the service of utility (Patton1997a)—not absolute allegiance to someideal standard of paradigm purity andmethodological orthodoxy.

Beyond CompetingInquiry Paradigms

Having presented the strategic ideals ofqualitative inquiry and noted variations intheir practical implementation and attain-ment, before closing this chapter I want toacknowledge and comment on the contro-versy that sometimes engulfs qualitativemethods. Students attempting to do qualita-tive dissertations can get caught up in andmay have to defend, philosophically as wellas methodologically, the use of qualitativeinquiry to skeptical committee memberswho define doctoral-level work as rigoroushypothesis testing. Evaluators may encoun-ter policymakers and funders who dismissqualitative data as mere anecdote. The sta-tistically addicted may poke fun at whatthey call the “softness” of qualitative data.(In Western society, where anything can beand often is sexualized, the distinction be-tween “hard” data and “soft” data has addi-tional nuances of meaning and innuendo.)Such encounters derive from a long-stand-ing methodological paradigms war. Thoughmany have pronounced the war and eventhe debate over (cf. Cook 1995; Greene1998a:36; Patton 1997a: 290-95), not every-one has adopted a stance of methodologicalenlightenment and tolerance, namely, thatmethodological orthodoxy, superiority, andpurity should yield to methodological ap-propriateness, pragmatism, and mutual re-spect. Therefore, a brief review of the para-digms debate is in order. (Elsewhere I haveprovided a more extensive review of themethodological paradigms debate; Patton1997a: 265-99, 1988c.)

Philosophers of science and method-ologists have been engaged in a long-stand-ing epistemological debate about the natureof “reality” and knowledge. That philosoph-ical debate finds its way into research and


evaluation in arguments over the goals ofempirical studies and differences of opinionabout what constitutes “good” research. Inits simplest and most strident formulation,this debate has centered on the relative valueof two different and competing inquiry par-adigms: (1) using quantitative and experi-mental methods to generate and test hypo-thetical-deductive generalizations versus(2) using qualitative and naturalistic ap-proaches to inductively and holistically un-derstand human experience and con-structed meanings in context-specificsettings. For example, Taylor and Bogdan(1984) contrast the Verstehen tradition,rooted in qualitative phenomenology, tomeasurement-oriented positivism as fol-lows:

Two major theoretical perspectives have dom-inated the social science scene. The first, posi-tivism, traces its origins in the social sciences tothe great theorists of the nineteenth and earlytwentieth centuries and especially to AugusteComte and Emile Durkheim. The positivistseeks the facts or causes of social phenomenaapart from the subjective states of individuals.Durkheim told the social scientist to considersocial facts, or social phenomena, as “things”that exercise an external influence on people.

The second theoretical perspective, which,following the lead of Deutscher, we will de-scribe as phenomenological, has a long history inphilosophy and sociology. The phenomen-ologist is committed to understanding socialphenomena from the actor’s own perspective.He or she examines how the world is experi-enced. The important reality is what peopleperceive it to be. (pp. 1-2)

Debate about these contrasting and com-peting perspectives has been an importantpart of the history of research and evalua-tion, but, as Chapters 3, 4, and 9 will show,

the variety of inquiry approaches has ex-panded well beyond the simplisticdichotomy between quantitative and quali-tative paradigms. In contrast to these twoclassically opposed orthodoxies, this bookoffers a pragmatic strategy of matching con-crete methods to specific questions, includ-ing the option of tactically mixing methodsas needed and appropriate. My practical(and controversial) view is that one can learnto be a good interviewer or observer, andlearn to make sense of the resulting data,without first engaging in deepepistemological reflection and philosophi-cal study. Such reflection and study can behelpful to those so inclined, but it is not aprerequisite for fieldwork. Indeed, it can bea hindrance. Getting some field experiencefirst, then studying philosophy of science,has much to recommend it as a learningstrategy. Otherwise, it’s all abstractions.Still, the paradigms debate is part of ourmethodological heritage and knowing a bitabout it, and its distortions (Shadish 1995b,1995c), may deepen appreciation for the im-portance of a strategic approach to methodsdecision making.

A paradigm is a worldview—a way ofthinking about and making sense of thecomplexities of the real world. As such, par-adigms are deeply embedded in the social-ization of adherents and practitioners. Para-digms tell us what is important, legitimate,and reasonable. Paradigms are also norma-tive, telling the practitioner what to do with-out the necessity of long existential orepistemological consideration. But it is thisaspect of paradigms that constitutes bothstrength and weakness—a strength in that itmakes action relatively easy, a weakness inthat the very reason for action is hidden inthe unquestioned assumptions of the para-digm.


Scientists work from models acquired througheducation and through subsequent exposureto the literature often without quite knowingor needing to know what characteristics havegiven these models the status of communityparadigms. . . . That scientists do not usually

ask or debate what makes a particular prob-lem or solution legitimate tempts us to sup-pose that, at least intuitively, they know theanswer. But it may only indicate that neitherthe question nor the answer is felt to be rele-vant to their research. Paradigms may be prior


Are there paradigms after death?

©20

02M

icha

elQ

uinn

Patt

onan

dM

icha

elC

ochr

an

to, more binding, and more complete than anyset of rules for research that could be unequiv-ocally abstracted from them. (Kuhn 1970:46)

But what does all this matter to the stu-dent interested in pursuing some research orevaluation question? It matters becauseparadigm-derived biases are the source ofthe distinctions mentioned earlier between“hard” data and “soft” data, empirical stud-ies versus “mere anecdotes,” and “objec-tive” research versus “subjective” studies.These labels reveal value-laden prejudicesabout what constitute credible and valuablecontributions to knowledge. Such preju-dices and paradigmatic blinders limit meth-odological choices, flexibility, and creativity.Adherence to a methodological paradigmcan lock researchers into unconscious pat-terns of perception and behavior that dis-guise the biased, predetermined nature oftheir methods “decisions.” Methods deci-sions tend to stem from disciplinary pre-scriptions, concerns about scientific status,old methodological habits, and comfortwith what the researcher knows best.Training and academic socialization tend to

make researchers biased in favor of andagainst certain approaches.

While one may still encounter peoplewho rigidly confess allegiance to only quan-titative or qualitative methods, most practi-tioners appear to have become eclectic andpragmatic. Looking back, we can now seethat the qualitative-quantitative debateoversimplified and often confused method-ological and philosophical issues. For exam-ple, the notion that some combinations ofmethods and philosophy ever constitutedconsistent, coherent, and stable paradigmshas proved problematic. Shadish (1995c), forexample, in introducing an important set ofarticles aimed at “de-Kuhnifying” the de-bate, concluded that “there is little empiricalevidence in support of such a Kuhnian para-digm portrayal. . . . [T]he relevant concep-tual and philosophical issues are far morecomplex than the simple quantitative-quali-tative dichotomy implies” (p. 48). Chapter 9will revisit the quantitative-qualitative para-digms debate in more depth as part of ourexamination of issues that affect judgmentsabout the quality and credibility of qualita-tive methods.

Pragmatism

Every thinker puts some portion of an apparently stable world in peril.

—John Dewey (1929)

While a paradigm offers a coherentworldview, an anchor of stability and cer-tainty in the real world sea of chaos, operat-ing narrowly within any singular paradigmcan be quite limiting. As a pragmatist, I takeissue as much with the purist, one-sided ad-vocacy of Lincoln and Guba (1985), who be-lieve that naturalistic inquiry is the onlyvalid and meaningful way to study humanbeings, as I do with the narrow, intolerantstance of Boruch and Rindskopf (1984), who

assert that randomized experiments are “thestandard against which other designs forimpact evaluation are judged” (p. 21). Mypragmatic stance aims to supersede one-sided paradigm allegiance by increasing theconcrete and practical methodological op-tions available to researchers and evalua-tors. Such pragmatism means judging thequality of a study by its intended purposes,available resources, procedures followed,and results obtained, all within a particular


context and for a specific audience. When anew drug is tested before being made avail-able to the general population, a dou-ble-blind randomized experiment to deter-mine efficacy is the design of choice, withcareful attention to controlled and carefullymeasured dosage and outcome interactions,including side effects. But if the concern iswhether people take the new drug appropri-ately, and one wants to know what people ina group think about the new drug (e.g., anantidepressant), how they make sense oftaking or not taking it, what they believeabout themselves as a result of experiencingthe drug, and how those around them dealwith it, then in-depth interviews and ob-servations are the place to start. The impor-tance of understanding alternative researchparadigms is to sensitize researchers andevaluators to the ways in which their meth-odological prejudices, derived from theirdisciplinary socialization experiences, mayreduce their methodological flexibility andadaptability.

I reiterate: Being pragmatic allows one toeschew methodological orthodoxy in favorof methodological appropriateness as the pri-mary criterion for judging methodologicalquality, recognizing that different methodsare appropriate for different situations. Situ-ational responsiveness means designing astudy that is appropriate for a specific in-quiry situation or interest. A major purposeof this book, and the focus of Chapter 4, is toidentify the kinds of research questions andprogram evaluation situations for whichqualitative inquiry is the appropriate meth-od of choice.

Paradigms are really about epistemology,ontology, and philosophy of science. Assuch, paradigms are important theoreticalconstructs for illuminating fundamental as-sumptions about the nature of reality. But atthe pragmatic level of making concretemethods decisions, this chapter’s emphasis

on strategic choices has conveyed, I hope,the idea that a wide range of possibilities ex-ists when selecting methods. The point is todo what makes sense, report fully on whatwas done, why it was done, and what theimplications are for findings. Chapter 5 isdevoted to design issues, including designflexibility, using multiple methods, andmaking practical decisions.

A Sufi story about the wise fool MullaNasrudin illustrates the importance of un-derstanding the connections between strate-gic ideals and practical tactics in real-worldsituations. Real-world situations seldom re-semble the theoretical ideals taught in theclassroom.

Ideal Conditions forResearch: A Cautionary Tale

In his youth, Nasrudin received trainingin a small monastery noted for its excellencein the teaching of martial arts. Nasrudin be-came highly skilled in self-defense and aftertwo years of training both his peers and histeachers recognized his superior abilities.

Each day, it was the responsibility of oneof the students to go to the village market tobeg for alms and food. It happened that asmall band of three thieves moved into thearea. They observed how the monastery ob-tained food daily and began hiding alongthe path the students had to take back to themonastery. As a returning student returned,laden with food and alms, the thieves wouldattack. After three days of such losses, themonastery’s few supplies were exhausted. Itwas Nasrudin’s turn to go to the village mar-ket. His elders and peers were confident thatNasrudin’s martial arts skills were morethan sufficient to overcome the small bandof thieves.

At the end of the day, Nasrudin returnedragged, beaten, and empty-handed. Every-one was amazed. Nasrudin was taken im-


mediately before the Master. “Nasrudin,” heasked, “how is it that with all your skill inour ancient arts of defense, you were over-come?”

“But I did not use the ancient arts,” re-plied Nasrudin.

All present were dumbfounded. An ex-planation was demanded.

“All of our competitions are preceded bygreat and courteous ceremony,” Nasrudinexplained. “We have learned that the open-ing prayers, the ceremonial cleansing, thebow to the East—these are essential to theancient ways. The ruffians seemed not to un-derstand the necessity for these things. I did-n’t find the situation ideal enough to use themethods you have taught us, Master.”

On more than one occasion, researchersor evaluators have told me of their belief inthe potential usefulness of qualitative meth-ods, but they tell me, “I just haven’t foundthe ideal situation in which to use them.”

Ideal situations are rare, but we will con-sider throughout this book the questions

and conditions in which qualitative strate-gies and methods offer advantages. InChapter 4, I will present a range and varietyof situations and inquiry problems that par-ticularly lend themselves to qualitative in-quiry. Chapter 5 will then discuss in moredetail some of the methodological trade-offsinvolved in adapting the strategic ideals ofqualitative methods to the practical realitiesof conducting research and evaluation in thefield. Chapters devoted to observation, in-terviewing, analysis, and enhancing thequality and credibility of qualitative studiesfollow that design chapter. To lay thegroundwork for in-depth review of applica-tions and methods, the next chapter exam-ines alternative theoretical frameworks thatare closely associated with and used toguide qualitative inquiry.

� Note

1. Excerpts in this chapter from the interviewwith Brackette F. Williams are used with her per-mission.


strategic themes in qualitative inquiry - sage publications · within a grand strat- ... that’s...

Documents