Chapter 5 Qualitative Research Methods PDF

CHAPTER 5 QUALITATIVE RESEARCH METHODS C H A P T E R OU T L I N E Aims and Philosophy Ethnography Mixed...

CHAPTER 5 QUALITATIVE RESEARCH METHODS C H A P T E R OU T L I N E Aims and Philosophy Ethnography Mixed Methods Research Writing the Qualitative Research Report Data Analysis in Qualitative Research Summary In-Person vs. Online Key Terms Field Observation Using the Internet Focus Groups Questions and Problems for Further Intensive Interviews Investigation Case Studies References and Suggested Readings 116 Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 5 Qualitative Research Methods 117 Part Two proceeds from a general discussion procedures, and assumptions about how of research to specific research techniques. researchers look at the world. Paradigms Chapter 5 discusses qualitative analysis, are based on axioms, or statements that are which relies mainly on the analysis of visual universally accepted as true. Paradigms are data (observations) and verbal data (words) important because they are related to the that reflect everyday experience. Chapter 6 selection of research methodologies. discusses content analysis, which focuses The positivist paradigm is the oldest and on words and other message characteristics still the most widely used in mass media but is conducted in a more systematic and research. Derived from the writings of philoso- measured way. Chapter 7 discusses survey phers such as Comte and Mill, positivism is the research, which relies on greater quantification paradigm most used in the natural sciences. and greater measurement sophistication than When the social sciences developed, research- either qualitative research or content analysis. ers modified this technique for their own pur- However, this sophistication comes with a poses. The positivist paradigm involves such price: Increasing quantification narrows the concepts as quantification, hypotheses, and types of research questions that can be objective measures. The positivist paradigm addressed. That is, research depth is sacrificed is the one that underlies the approach of this to gain research breadth. Chapter 8 discusses book. longitudinal research, and, finally, Chapter 9 Interpretive social science traces its roots concludes Part Two with a discussion of to Max Weber and Wilhelm Dilthey. The experimental methods, which are among the aim of the interpretive paradigm is to under- most precise, complex, and intricate of stand how people in everyday natural set- methodologies. tings create meaning and interpret the events of their world. This paradigm became popular in mass media research during the AIMS AND PHILOSOPHY 1970s and 1980s and gained added visibility Discussing the qualitative research approach in the 1990s and the new century. can be confusing because there is no commonly The critical paradigm draws on analysis accepted definition of the term qualitative. In models used in the humanities. Critical fact, some qualitative researchers resist defining researchers are interested in such concepts the term at all for fear of limiting the technique. as the distribution of power in society and The task is further complicated because of the political ideology. Though useful in many several levels of reference connected with the cases, a consideration of the critical para- term. The word qualitative has been used to digm is beyond the scope of this book. Inter- refer to (1) a broad philosophy and approach ested readers should consult Hall (1982). At to research, (2) a research methodology, and the risk of oversimplification, in the rest of (3) a specific set of research techniques. To bet- this section we compare the positivist and ter understand this area, it is helpful to step interpretive paradigms. back and examine some general considerations The positivist paradigm differs from the related to social science research. interpretive paradigm along three main Neuman (1997) and Blaikie (1993) sug- dimensions. First, the two approaches have gest that there are three distinct approaches a different philosophy of reality. For the pos- to social science research: positivist (or objec- itivist researcher, reality is objective; it exists tivist), interpretive, and critical. Each of these apart from researchers and can be seen by represents a model or a paradigm for all. In other words, it is out there. For the research—an accepted set of theories, interpretive researcher, there is no single Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  118 Part Two Research Approaches reality. Each observer creates reality as part of 2. Design. For a positivist, the design of a the research process. It is subjective and exists study is determined before it begins. In inter- only in reference to the observer. Perhaps a pretive research, the design evolves during classic example will help here. If a tree falls the research and can be adjusted or changed in a forest and there is no one there to hear as the research progresses. it, does it make a sound? On the one hand, a 3. Setting. The positivist researcher tries to positivist would answer yes—reality doesn’t limit contaminating and confounding variables depend on an observer; it exists independently. by conducting investigations in controlled set- On the other hand, an interpretive researcher tings. The interpretive researcher conducts stud- would say no sound was made—reality exists ies in the field, in natural surroundings, trying only in the observer. Furthermore, the positiv- to capture the normal flow of events without ist researcher believes that reality can be controlling extraneous variables. divided into component parts, and knowledge 4. Measurement instruments. In positivist of the whole is gained by looking at the parts. research, measurement instruments exist In contrast, the interpretive researcher exam- apart from the researcher; another party ines the entire process, believing that reality is could use the instruments to collect data in holistic and cannot be subdivided. the researcher’s absence. In interpretive Second, the two approaches have differ- research, the researcher is the instrument; ent views of the individual. The positivist no other individual can substitute. researcher believes that all human beings 5. Theory building. Where the positivist are basically similar and looks for general researcher uses research to test, support, or categories to summarize their behaviors or reject theory, the interpretive researcher feelings. The interpretive researcher believes develops theories as part of the research that human beings are fundamentally differ- process—theory is “data driven” and ent and cannot be pigeonholed. emerges as part of the research process, Third, positivist researchers aim to gener- evolving from the data as they are collected. ate general laws of behavior and explain many things across many settings. In con- A researcher’s paradigm has a great influ- trast, interpretive researchers attempt to pro- ence on the specific research methods the duce a unique explanation about a given researcher uses. As Potter (1996) explains, situation or individual. Whereas positivist “Two scholars who hold different beliefs [para- researchers strive for breadth, interpretive digms] … may be interested in examining the researchers strive for depth. same phenomenon but their beliefs will lead The practical differences between these them to set up their studies very differently approaches are perhaps most apparent in because of their differing views of evidence, the research process. The following five analysis and the purpose of the research” major research areas demonstrate significant (p. 36). The positivist approach is most closely differences between the positivist and inter- associated with quantitative content analysis, pretive approaches: surveys, and experiments, techniques discussed in detail in subsequent chapters. The interpre- 1. Role of the researcher. The positivist tive approach is most closely connected with the researcher strives for objectivity and is sepa- specific research methods discussed in this rated from the data. The interpretive chapter. Research methods, however, are not researcher is an integral part of the data; in conscious of the philosophy that influenced fact, without the active participation of the their selection. It is not unusual to find a researcher, no data exist. positivist using focus groups or intensive Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 5 Qualitative Research Methods 119 interviewing, two methods commonly catego- but the questions discussed in the groups con- rized as qualitative, in connection with a centrate on how group members interpret a quantitative study. Nor is it rare to find an inter- political ad, what meanings they derive from a pretive researcher using numbers from a survey negative ad, the context of their viewing, and or content analysis. Thus, the guidelines for what makes them feel cynical toward politics. focus groups discussed in this chapter, or the The focus groups stand alone as the source of discussion of survey research in a subsequent data for the analysis. The interpretive researcher chapter, are relevant to both paradigms. uses induction to try to find commonalities or Although the methods may be the same, the general themes in participants’ remarks. Thus, research goal, the research question, and both researchers use focus groups, a method tra- the way the data are interpreted are quite ditionally defined as qualitative, but each uses different. the method somewhat differently. To use a concrete example, assume that a Despite the differences, many researchers positivist researcher is interested in testing now use a combination of the quantitative the hypothesis that viewing negative political and qualitative approaches to understand ads increases political cynicism. The fully the phenomenon they are studying. As researcher conducts focus groups to help Miles and Huberman (1994) state: develop a questionnaire that measures cyni- cism and exposure to what is defined as neg- It is getting harder to find any methodolo- ative advertising. A statistical analysis is then gists solidly encamped in one epistemology conducted to determine if these two items are or the other. More and more “quantita- related and if the hypothesis is supported. tive” methodologists … are using naturalis- An interpretive researcher interested in the tic and phenomenological approaches to same question might also conduct focus groups, complement tests, surveys, and structured A CLOSER LOOK Methodology and Methods The words methodology and methods are investigating and how the research should sometimes confused. Methodology is the study proceed. of methods and the underpinning philosophical Different methodologies are associated with assumptions of the research process itself. Dif- different paradigms. Quantitative methodology ferent research questions suggest different meth- generally adopts the positive paradigm, whereas odologies. A researcher interested in how the qualitative researchers promote the critical para- Internet affects copyright laws would probably digm. Those who accept the critical paradigm gen- choose the methodology of legal research. A erally follow the methodology of the humanities. researcher who wants to trace how radio pro- In contrast, a method is a specific technique gramming has evolved since the introduction of for gathering information following the assump- television would probably choose historical tions of the chosen methodology. Researchers methodology. A study about the effects of tele- who choose the positivist paradigm will use vision on children might use scientific methodol- such methods as surveys and experiments, ogy. In short, methodology deals with the whereas those who choose the interpretive par- question of “why” to do research in a certain adigm will rely on methods such as focus way. It is a guide to what problems are worth groups, ethnography, and observation. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  120 Part Two Research Approaches interviews. On the other side, an increasing approach, called “mixed methods” (see number of ethnographers and qualitative below) has become popular in many of the researchers are using predesigned conceptual social sciences. frameworks and prestructured instrumenta- Although qualitative research can be an tion…. Most people now see the world excellent way to collect and analyze data, with more ecumenical eyes. (p. 20) researchers must keep in mind that the results of such studies have interpretational In past years, an occasional “turf war” limits if sample sizes are small. Researchers between the two approaches has erupted interested in generalizing results should use (Kover, 2008). In the authors’ opinion, fram- large samples or consider other methods. ing the debate as the qualitative versus the However, in most cases qualitative research quantitative approach is not productive. It studies use small samples—respondents or is more useful to look at ways the two meth- informants that are not necessarily represen- odologies can be integrated. As Neill (2007) tative of the population from which they are puts it, “More good can come of social drawn. Like quantitative research, qualita- science researchers developing skills in both tive research is a useful mass media research realms than debating which method is tool only when its limitations are recognized. superior.” In fact, recent developments All too often, the results from small-sample suggest that many researchers have adopted qualitative projects are interpreted as though the ecumenical perspective of Miles and they had been collected with large-sample Huberman. Although it has yet to become quantitative techniques. This approach can widely used in mass media research, a new only cause problems. A CLOSER LOOK Qualitative Research Definition—A Final Note Although most qualitative research projects use Qualitative research uses a flexible question- small samples that eliminate a researcher’s abil- ing approach. Although a basic set of ques- ity to generalize the results to the population, the tions is designed to start the project, the truth is that it is easy to increase sample size to researcher can change questions or ask avoid this problem. This is often done in both follow-up questions at any time. private- and public-sector research and therefore Quantitative research uses a static or stan- eliminates the primary argument against using dardized set of questions. All respondents qualitative research. So what’s the problem? are asked the same questions. Although If large sample sizes are used, then the differ- follow-up questions (and skips) can be ence between qualitative research and quantita- designed into a questionnaire, they must tive research must relate to something else. It be included in the questionnaire or measure- does—it relates to how questions are asked. ment instrument before the research project When all the clouds of controversy are eliminated, begins. Interviewers conducting the interview the difference between qualitative research and are not allowed to stray from the quantitative research boils down to this: questionnaire. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 5 Qualitative Research Methods 121 MIXED METHODS RESEARCH a researcher may follow up a survey by con- ducting intensive interviews with some of the As defined by Creswell (2003), a mixed respondents in order to more fully under- methods approach is one in which the stand the results. researcher collects, analyzes, and integrates The mixed methods approach has several both quantitative and qualitative data in a sin- advantages. First, the technique can produce gle study or multiple studies in a sustained pro- stronger evidence for a conclusion through a gram of inquiry. The mixed methods approach convergence of findings (akin to the triangula- draws from the strengths of both qualitative tion notion mentioned in Chapter 2). Second, a and quantitative techniques. Researchers who researcher can answer a broader range of advocate this approach are less interested in research questions because the research is not debating whether quantitative philosophy is confined to a single method. Finally, the tech- compatible with qualitative philosophy and nique can provide information and insight that are more interested in using the approach, or might be missed if only a single method were combination of approaches, that works best in used. examining the research question. There are disadvantages, as well. As is There are several models of how mixed obvious, mixed methods research requires methods research may be designed. Figure 5.1, more time and effort because the researcher adapted from Creswell (2007), shows three is actually conducting two studies. In addi- basic approaches. A qualitative approach is tion, the technique requires the researcher abbreviated QUAL, while QUAN denotes a to be skilled in both qualitative and quanti- quantitative approach. tative methods. If these skills are lacking, it In the concurrent design, both qualitative might require a research team. Lastly, data and quantitative data are collected at the analysis might be more difficult, particularly same time and both are weighted equally in if the methods yield conflicting results. analysis and interpretation. An example Although still relatively rare in mass might be a survey questionnaire or interview media research, studies using the mixed meth- that contains both closed-ended quantitative ods approach are beginning to appear in items and open-ended qualitative items. mass communication journals. See, for exam- In the sequential designs, one method ple, Lieberman, Neuendorf, Denny, Skalski, precedes the other. For example, a researcher and Wang (2009); Greenwood (2010); and might conduct focus groups that generate Gunther, Kautz, and Roth (2011). Those items to be used in a subsequent survey, or Figure 5.1 Mixed Methods Research Designs I. Concurrent Model QUAL + QUAN II. Sequential Models QUAL → QUAN QUAN → QUAL Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  122 Part Two Research Approaches readers who want to examine more examples the data are arranged in chronological order should consult the Journal of Mixed Methods according to the sequence of events that Research and search the Internet for “mixed occurred during the investigation. Further- methods research.” more, each piece of information is coded to identify the source, and multiple photocopies DATA ANALYSIS IN QUALITATIVE and computer files of the notes, transcripts, and other documents are mandatory. RESEARCH The data are then organized into a prelim- Before examining some specific types of quali- inary category system. These categories might tative research, let’s discuss qualitative data arise from the data or they might be suggested and methods of analysis in general. Qualitative by prior research or theory. Many researchers data come in a variety of forms, such as notes prefer to do a preliminary run-through of the made while observing in the field, interview data and record possible category assign- transcripts, documents, diaries, and journals. ments in the margins. For example, a qualita- In addition, a researcher accumulates a great tive study of teenage radio listening might deal of data during the course of a study. produce many pages of interview transcripts. Organizing, analyzing, and making sense of The researcher would read the comments all this information pose special challenges and might write “peer group pressure” next for the researcher using qualitative methods. to one section and “escape” next to another. Unlike the quantitative approach, in When the process is finished, a preliminary which analysis does not begin until all the category system may emerge from the data. numbers are collected, data analysis in quali- Other researchers prefer to make multiple tative studies is done early in the collection copies of the data, cut them into coherent process and continues throughout the project. units of analysis, and physically sort them In addition, quantitative researchers generally into as many categories as might be relevant. follow a deductive model in data analysis: Finally, some of the toil in qualitative data Hypotheses are developed prior to the study, analysis can be made easier by commercial and relevant data are then collected and ana- software programs. Some that are widely lyzed to determine whether the hypotheses are used include NVivo, a program that allows confirmed. In contrast, qualitative researchers users to import, sort, and analyze video and use an inductive method: Data are collected audio files, photos, and text documents relevant to some topic and are grouped into and to display results in models and charts; appropriate and meaningful categories; expla- and HyperRESEARCH, a similar program nations emerge from the data. The remainder that enables a researcher to code, retrieve, of this section follows a modified version and build models using data from audio, of the phases of qualitative data analysis sug- video, and text sources. gested by Miles and Huberman (1994): Many qualitative researchers like to have (1) data reduction, (2) data display, (3) con- a particular room or other space that is spe- clusion drawing, and (4) verification. cially suited for the analysis of qualitative data. Typically, this room has bulletin Preparing the Data: Reduction boards or other arrangements for the visual and Display display of data. Photocopies of notes, obser- To facilitate working with the large amounts vations written on index cards, large flow- of data generated by a qualitative analysis, the charts, and marginal comments can then be researcher generally first organizes the infor- conveniently arrayed to simplify the analysis mation along a temporal dimension. That is, task. Because it is an efficient way to display Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 5 Qualitative Research Methods 123 A CLOSER LOOK Software for Qualitative Data Analysis Software can help ease the labor-intensive task Weft QDA http://www.pressure.to/ of analyzing qualitative data. Current pro- qda/ grams can provide simple word counts, isolate themes, show interconnections among the data, QDAP http://www.umass.edu/ and produce graphical displays. Many pro- qdap/ grams also provide the option for the ATLAS http://www.atlasti.com/ researcher to add comments and marginal notes to the data. Listed below are the names NVivo http://www.qsrinternational. and URLs of some of the more popular qualita- com/#tab_you tive data analysis programs. (This information HyperRESEARCH http://researchware.com/ was current as of late 2011.) Note that the first two on the list are free but have limited Ethnograph http://qualisresearch.com/ capability. The others offer a free trial period. the data to several people at once, this “ana- has subsequently been refined (Lincoln & lytical wallpaper” approach is particularly Guba, 1985). At a general level, the process helpful when there are several members of consists of four steps: the research team working on the project. 1. Comparatively assigning incidents to Finally, the researcher is the main instru- categories ment in qualitative data collection and anal- ysis and therefore must prepare before 2. Elaborating and refining categories beginning the task of investigation. Maykut 3. Searching for relationships and and Morehouse (1994) describe this prepara- themes among categories tion as epoche, the process by which the 4. Simplifying and integrating data into researcher tries to remove or at least become a coherent theoretical structure aware of prejudices, viewpoints, or assump- tions that might interfere with the analysis. Comparative assignment of incidents to cat- Epoche helps the researcher put aside per- egories. After the data have been prepared sonal viewpoints so that the phenomenon for analysis, the researcher places each unit under study may be seen for itself. of analysis into a set of provisional catego- ries. As each new unit is examined, it is com- Data Analysis: Conclusion pared to the other units previously assigned Drawing to that category to see whether its inclusion Qualitative data can be analyzed with many dif- is appropriate. It is possible that some initial ferent techniques. This section discusses two of categories may have only one or two inci- the best known: the constant comparative tech- dents assigned to them, whereas others may nique and the analytical induction technique. have a large number. If some units of analy- The constant comparative technique (fre- sis do not fit any preexisting category, new quently called Grounded Theory) was first classifications may have to be created. Units articulated by Glaser and Strauss (1967) and that fit into more than one category should Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  124 Part Two Research Approaches be copied and included where relevant. topic and help determine the research Throughout the process, the emphasis is on outcome. comparing units and finding similarities After scanning all the data cards in the among the units that fit into the category. “interpersonal connections” category, a For example, suppose a researcher is con- researcher might write a proposition such ducting a qualitative study about why indivi- as “People subscribe to social networking duals join social networking sites such as websites to expand their circle of casual Facebook or Twitter. Interviews are conducted friends.” Similar statements are written for with several people and transcribed. The the other categories. researcher then defines each individual asser- tion as the unit of analysis and writes each Searching for relationships and themes statement on an index card. The first two among categories. The third phase of the cards selected for analysis mention keeping in method involves searching for relationships touch with current friends. The researcher and common patterns across categories. places both of these into a category tentatively The researcher examines the propositional labeled “interpersonal connections.” The next statements and looks for meaningful connec- statement mentions social pressure to belong tions. Some propositions are probably strong to a group; it does not seem to belong to enough to stand alone; others might be the first category and is set aside. The next related in several important ways. Whatever card mentions finding out about what old the situation, the goal of this phase is to gen- high school acquaintances are doing. The erate assertions that can explain and further researcher decides this reason is similar to clarify the phenomenon under study. the first two and places it in the interpersonal In our online example, the researcher connections category. The next comment might note that several propositions refer to talks about not wanting to be left out. The the notion of expansion. People use social researcher believes that this comment, like the networking sites to expand their circle of earlier one, reflects social pressure and starts friends, or to expand their sources of new another category called “peer pressure.” The music, or to expand the number of groups process is repeated with every unit of analysis, to which they belong. The analyst then gen- which can be a long and formidable task. eralizes that the expansion of one’s social However, at some point during the process, and cultural space is an essential reason for the researcher begins to refine the categories. joining. Elaboration and refinement of categories. Simplifying and integrating data into a During the category refinement stage, the coherent theoretical structure. In the final researcher writes rules or propositions that phase of the process, the researcher writes a attempt to describe the underlying meaning report that summarizes the research. The that defines the category. Some rules for results of the foregoing analyses are inte- inclusion might be rewritten and revised grated into some coherent explanation of throughout the study. These rules help to the phenomenon. The researcher attempts focus the study and also allow the researcher to write a brief explanation, but in sufficient to start to explore the theoretical dimensions detail to convey an idea of the scope of the of the emerging category system. The ulti- project. The goal of this phase of the project mate value of these rules is that they reveal is to arrive at an understanding of the people what the researcher learns about a chosen and events being studied. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 5 Qualitative Research Methods 125 The analytic induction strategy blends essentially the same reasons, but also adds hypothesis construction and data analysis. that he uses the prices advertised on the It consists of the following steps (adapted channel to comparison shop. Once again, from Stainback & Stainback, 1988): the hypothesis is refined. The researcher posits that the home-shopping channels are 1. Define a topic of interest and develop viewed for practical, consumer-related rea- a hypothesis. sons: finding bargains, learning about pro- 2. Study a case to see whether the ducts, and comparing prices. hypothesis works. If it doesn’t work, At this point, the researcher tries to find reformulate it. cases that might not fit the new hypothesis. A 3. Study other cases until the hypothe- colleague points out that all the people inter- sis is in refined form. viewed so far have been affluent, with substan- 4. Look for “negative cases” that might tial disposable income, and that perhaps disprove the hypothesis. Reformulate people who are less well-off economically again. might watch the home-shopping channels for other reasons. The researcher interviews a 5. Continue until the hypothesis is ade- viewer from a different economic background quately tested. and discovers that this person watches because Note that in this method, an explanation he finds the salespeople entertaining to watch. for the phenomenon, in the form of a Once again, the initial hypothesis is modified hypothesis, is generated at the beginning of to take this finding into account. the study. This process contrasts with the The researcher then finds other respon- constant comparative technique, in which dents from different economic levels to check an explanation is derived as the result of the the validity of this new hypothesis and con- research. tinues to gather data until no more cases can A simplified example demonstrates how be located that do not fit the revised hypothesis. the analytic induction approach works. Sup- This process can be exhausting, and it pose that a researcher is interested in explaining can be difficult for the researcher to deter- why people watch home-shopping channels. mine an exact stopping point. One might Colleagues tell the researcher that the answer always argue that there are still cases in the is obvious: People watch because they want environment that would not support the to buy the merchandise. The researcher is not hypothesis, but the researcher simply did convinced of this but decides to use the expla- not find them. nation as an initial hypothesis and finds a person who is known to be a heavy viewer Verification: Reliability and of these channels. During the interview, the Validity in Qualitative Data person says that although she has ordered a Qualitative researchers must pay attention to couple of things off the air, her primary reason several different concerns that may call the for watching is to find out about new and credibility of their research into question. unusual products. First, there is the matter of the completeness Armed with this information, the of the data. If qualitative researchers do a researcher reformulates the hypothesis: Peo- sloppy job taking notes or otherwise record- ple watch the home-shopping channels to ing what was observed, there is the possibil- buy and find out about new products. ity that incorrect interpretations may be Another viewer is interviewed and reports drawn from the data. A second problem Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  126 Part Two Research Approaches concerns selective perception. Qualitative 1. Multiple methods of data collection. researchers cannot simply dismiss data that This factor is similar to the notion of trian- do not fit a favored interpretation of the gulation that was discussed in Chapter 2. data. They must analyze these cases and The use of interviews along with field obser- offer explanations as to why the data don’t vations and analysis of existing documents seem to fit. Finally, qualitative research often suggests that the topic was examined from raises the question of reactivity—when the several different perspectives, which helps act of observing some situation changes build confidence in the findings. the situation itself. Would the same things 2. Audit trail. This factor is essentially a have occurred if researchers were not there? permanent record of the original data used Reactivity is a difficult problem to overcome, for analysis and the researcher’s comments but researchers must try to minimize it when- and analysis methods. The audit trail ever possible. Taken together, these three fac- allows others to examine the thought pro- tors suggest that qualitative researchers, cesses involved in the researcher’s work much like quantitative researchers, must and allows them to assess the accuracy pay attention to the reliability and validity of the conclusions. of their data. 3. Member checks. In this technique, However, the concepts of reliability and research participants are asked to read a validity have different connotations for qual- researcher’s notes and conclusions and tell itative data. As we discuss later, quantitative whether the researcher has accurately methods use distinct and precise ways to cal- described what they were told. culate indexes of reliability and several artic- 4. Research team. This method assumes ulated techniques that help establish validity. that team members keep each other honest Yet these concepts do not translate well into and on target when describing and inter- the interpretive paradigm. As Lindlof (2002) preting data. When appropriate, an outside points out, interpretive research recognizes person is asked to observe the process the changing nature of behavior and percep- and raise questions of possible bias or tion over time. Nonetheless, though envi- misinterpretation. sioned differently, reliability and validity are no less important in qualitative research. Creswell (2007) suggests another method They help readers determine how much con- to aid verification—debriefing. This consists fidence can be placed in the results of the of having an individual outside the project study and whether they can believe the question the meanings, methods, and inter- researcher’s conclusions. Or, as Lindlof pretations of the researcher. If more detail (1995) says, “Basically, we want to inspire is needed, Onwuegbuzie and Leech (2006) confidence in readers (and ourselves) that describe 24 methods for assessing the validity we have achieved right interpretations” (p. of qualitative research. 238). Or, as Hammersly (1992) expressed Barusch, Gringeri, and George (2011) it, validity is achieved when the description provide evidence of the relative use of the of the observed phenomenon accurately above verification techniques. They studied depicts what was observed. 100 articles that used qualitative methods Rather than emphasizing reliability and and found that 60% used multiple methods validity, Maykut and Morehouse (1994) of data collection, 31% used member checks, address the trustworthiness of a qualitative the same percentage used the research team research project. They summarize four fac- method, while 9% used an audit trail. tors that help build credibility: Debriefing was the least common at 7%. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 5 Qualitative Research Methods 127 IN-PERSON VS. ONLINE to use depends primarily on the research question and the objectives of the research. While the Internet has opened new possibili- And, of course, it is always possible to com- ties for qualitative researchers, there are bine elements of both approaches. An online advantages and disadvantages involved. focus group might suggest a follow-up par- Let’s take a broad-spectrum look at the gen- ticipant observation study in real life. The eral strengths and weaknesses of online qual- balance of this chapter discusses five com- itative research and in-person qualitative mon qualitative techniques: field observa- research. Note that the strength of one tion, focus groups, intensive interviews, case approach is usually a weakness in the studies, and ethnography. For each method, other. With in-person qualitative research: we describe the traditional procedures and The data are “richer,” in that observers then examine their online counterparts. can see the physical responses and sur- roundings of their respondents. Body FIELD OBSERVATION language and facial expressions can add to understanding. Field observation is useful for collecting data and for generating hypotheses and theories. Respondents do not need special com- Like all qualitative techniques, it is concerned puter or keyboarding skills. more with description and explanation than Projective tests and product demon- with measurement and quantification. strations are possible. Figure 5.2 shows that field observations are Group dynamics can offer clues to classified along two major dimensions: (1) analysis and interpretation. the degree to which the researcher partici- The researcher is an integral part of the pates in the behavior under observation and data collection. (2) the degree to which the observation is concealed. With online qualitative research: Quadrant 1 in Figure 5.2 represents overt Coverage of wide geographic areas is observation. In this situation, the researcher possible. Neither the respondent nor the is identified when the study begins, and those researcher has to be in the same spot. under observation are aware that they are being studied. Furthermore, the researcher’s The online behavior of large groups role is only to observe, refraining from par- (for example Facebook users) can be ticipation in the process under observation. observed. Quadrant 2 represents overt participation. Responses may be more thoughtful and In this arrangement, those being observed contain more information. Recruiting also know the researcher, but unlike the sit- busy respondents is easier since the uation represented in Quadrant 1, the research can take place at the respon- researcher goes beyond the observer role dents’ convenience. and becomes a participant in the situation. There is no bias for or against vocal or Quadrant 3 represents the situation where outgoing respondents. the researcher’s role is limited to that of Expenses are often substantially lower observer but those under observation are than other approaches. not aware they are being studied. Quadrant 4 represents a study in which the researcher Both approaches have inherent strengths participates in the process under investiga- and weaknesses. The choice of which method tion but is not identified as a researcher. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  128 Part Two Research Approaches Figure 5.2 Dimensions of Field Observation Overt 1 2 Observer Participant 3 4 Covert To illustrate the differences among the Some examples of field observation stud- various approaches, assume that a researcher ies in mass media research include Gieber’s wants to observe and analyze the dynamics (1956) classic study of gatekeeping (informa- of writing comedy for television. The tion flow) in the newsroom and Epstein’s researcher could choose the covert observer (1974) description of network news opera- technique and pretend to be doing something tions. Bielby, Harrington, and Bielby (1999) else (such as fixing a computer) while actu- observed meetings of soap opera fan clubs in ally observing the TV writing team at work. their study of the way fans relate to soap Or the researcher could be introduced as opera characters. Smith and Krugman someone doing a study of comedy writing (2010) did in-home observations of how and watch the team in action. If the research people used their digital video recorders, question is best answered by active participa- and Christie (2009) observed how people tion, the researcher might be introduced as a used new communication technologies in researcher, but still participate in the writing bookstore-cafes. process. If the covert participant strategy is used, the researcher might be introduced as a Advantages of Field Observations new writer just joining the group (such an Although field observation is not an appro- arrangement might be made with the head priate technique for every research question, writer, who would be the only person to it does have several unique advantages. For know the identity of the researcher). The one thing, many mass media problems and choice of technique depends on the research questions cannot be studied using any other problem, the degree of cooperation available methodology. Field observation often helps from the group or individual observed, and the researcher define basic background infor- ethical considerations. While covert partici- mation necessary to frame a hypothesis and pation may affect subjects’ behavior and to isolate independent and dependent vari- raise the ethical question of deception, the ables. For example, a researcher interested information gathered may be more valid if in how creative decisions in advertising are subjects are unaware of being scrutinized. made could observe several decision-making Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 5 Qualitative Research Methods 129 sessions to see what happens. Field observa- Disadvantages of Field tions often make excellent pilot studies Observations because they identify important variables On the negative side, field observation is a and provide useful preliminary information. poor choice if the researcher is concerned In addition, because the data are gathered with external validity. Validation is difficult firsthand, observation is not dependent on partly because the representativeness of the the subjects’ ability or willingness to report observations made is potentially questionable their behavior. For example, young children and partly because of problems in sampling. may lack the reading or verbal skills neces- Observing the TV viewing behavior of a sary to respond to a questionnaire concern- group of children at a daycare center can pro- ing their TV viewing behavior, but such data vide valuable insights into the social setting of are easily gathered by the observational television viewing, but it probably has little technique. correlation with what preschoolers do in A field observation is not always used as other places and under different circumstances. a preliminary step to other approaches. Besides, since field observation relies heavily on Sometimes it alone is the only appropriate a researcher’s perceptions and judgments and approach, especially when quantification is on preconceived notions about the material difficult. Field observation is particularly under study, experimenter bias may favor spe- suitable for a study of the gatekeeping pro- cific preconceptions of results, while observa- cess in a network television news department tions to the contrary are ignored or distorted. because it is difficult to quantify gatekeeping. Potential bias is why it is rare to use only one Field observation may also provide access observer in a field observation study—observa- to groups that would otherwise be difficult to