Abstract
In this article the authors outline five types of debriefing and introduce a new type of debriefing, namely, that of debriefing the interpretive researcher. Next they present eight main areas accompanied by example questions to guide the interviewer when debriefing the researcher. They also present five authenticity criteria developed by Guba and Lincoln (1989) and include possible interview questions to document the degree to which the researcher has met these criteria. Finally, using Miles and Huberman's (1994) framework, they illustrate how displays such as matrices can be used to collect, analyze, and interpret debriefing interview data as well as leave an audit trail.
In quantitative research studies in the social and behavioral sciences, multiple instruments might be used in a single study, for example published surveys and/or known treatments (Campbell & Stanley, 1963). From the postpositivist point of view, in quantitative research the researcher is distant from the data, in fact so distant that the research is considered to be completely bias free (Yu, 2003). This leads one to believe that any researcher could conduct a specific quantitative study and the results would be similar; that is, to assume that an individual researcher's background, perceptions, and viewpoints do not influence the interpretations of the findings and implications of the research.
Unlike in quantitative research, when conducting a qualitative study, the researcher is considered as the main instrument for data collection, data analysis, and data interpretation (Paisley & Reeves, 2001). As noted by Lincoln and Guba (1985), “the instrument of choice in naturalistic inquiry [qualitative research] is the human” (p. 236). As such, in a qualitative research study researchers bring in their inherent (researcher) biases, which must be acknowledged and identified (Miles & Huberman, 1994). Bias itself represents a rich source of data. Indeed, Strauss (1987) coined the phrase experiential data when referring to researcher bias. According to Strauss,
These experiential data should not be ignored because of the usual canons governing research (which regard personal experience and data as likely to bias the research), for these canons lead to the squashing of valuable experiential data. We say, rather, “mine your experience, there is potential gold there.” (p. 11)
A major focus among many qualitative researchers is to capture authentically the lived experiences of people. According to Denzin and Lincoln (2005),
Such experience, it is argued, is created in the social text written by the researcher. This is the representational problem. It confronts the inescapable problem of representation, but does so within a framework that makes the direct link between experience and text problematic. (p. 19)
Denzin and Lincoln have referred to this problem as the crisis of representation. These authors also have argued for “a serious rethinking of such terms as validity, generalizability, and reliability, terms already retheorized in postpositivist … constructivist-naturalistic … feminist … interpretive … poststructural … and critical … discourses. This problem asks, ‘How are qualitative studies to be evaluated in the contemporary, poststructural moment?’” (pp. 19–20, italics in original). Denzin and Lincoln have called this problem the crisis of legitimation. These authors also discuss what they term the crisis of praxis in qualitative research. This crisis leads to the question, “Is it possible to effect change in the world if society is only and always a text?” (p. 20). As noted by Onwuegbuzie and Leech (2004),
The crises of representation, legitimation, and praxis threaten qualitative researchers' ability to extract meaning from their data. In particular, lack of representation means that the evaluator has not adequately captured the data. Lack of legitimation means that the extent to which the data have been captured has not been adequately assessed, or that any such assessment has not provided support for legitimation. Thus, the significance of findings in qualitative research is affected by these crises. (p. 778)
The crises of representation, legitimation, and praxis make it important for interpretive researchers to be aware of the role that their biases can play in exacerbating these crises during the course of a qualitative study.
Miles and Huberman (1994) identified two sources of researcher bias: (a) the effects of the researcher on the study participant(s)/key informant(s) (termed Bias A), and (b) the effects of the study participant(s)/key informant(s) on the researcher (termed Bias B). These biases can come to the fore at any stage of the qualitative research process. According to Miles and Huberman, Bias A prevails when the qualitative researcher disrupts or poses a threat to the existing social or institutional relationships. Bias A also can lead to informants' implicitly or explicitly boycotting the researcher, who is viewed as a spy, critic, nuisance, voyeur, or antagonist. In addition, Bias A can inhibit informants. In contrast, Bias B can lead the researcher to go native (i.e., become a complete participant, as opposed to a peripheral-member researcher who develops desirable emic perspective without participating in those activities central to person/group/society under study; Adler & Adler, 1987).
Unfortunately, these two sources of researcher bias (i.e., Bias A and Bias B) typically are not examined systematically, thereby preventing the qualitative researcher from reflecting critically (i.e., systematic reflexivity) to the greatest extent possible on the role that bias plays in the research process in general and in shaping the findings and interpretations in particular. Such systematic reflexivity also could facilitate qualitative researchers' process of “examining one's personal and theoretical commitments to see how they serve as resources for generating particular data, for behaving in particular ways vis-à-vis respondents and participants, and for developing particular interpretations” (Schwandt, 1997, p. 136). As such, collecting and critically reflecting on researcher bias in a systematic manner can greatly enhance the legitimation of accounts of social and behavioral phenomena (i.e., help to address the crises of legitimation and praxis). Obtaining this information can benefit the qualitative study by clarifying the findings (i.e., help to address the crisis of representation), elucidating possible problems (i.e., help to address the crises of legitimation and praxis), assisting the researcher in keeping bias from unduly influencing the results (i.e., help to address the crisis of legitimation), and, above all, helping the researcher understand the role that her or his bias is playing in the study (i.e., help to address the crises of representation and legitimation). Indeed, as noted by Fontana and Frey (2005), “many studies that use unstructured interviews are not reflexive enough about the interpreting process” (p. 713).
One way to obtain and use reflexive data from the researcher is by debriefing the researcher. Specifically, to debrief a researcher, someone who is not involved in the study interviews the researcher and collects debriefing data. The purpose of collecting these data is to help interpretive researchers to identify and to reflect on the degree to which their biases potentially might have influenced the various facets of the research study (e.g., formulating the research question, implementing data collection, and conducting analytical procedures), might have changed over the course of the investigation in general and interview process in particular, and might have affected interpretations of findings (i.e., interview data) and implications stemming from the findings (e.g., formulating analytical generalizations). In addition, debriefing interviews provide an opportunity for the researcher to evaluate initial hunches. The process of the researcher explaining and/or verifying initial hunches to the debriefing interviewer might illuminate to the researcher the plausibility and trustworthiness of these hunches in the conduct of the research. Despite the multiple benefits of debriefing or interviewing the researcher, however, to date this technique has not been advocated by qualitative researchers.
Thus, the purpose of this article is to introduce this new type of debriefing technique in the context of qualitative research. Often the qualitative researcher who conducts the study alone might not be optimally aware of information (e.g., thoughts, perceptions, feelings, and experiences that emerge during the conduct of the research). Therefore, being formally interviewed by another person can help researchers to promote reflexivity by reflecting “on their historical, socio-cultural, and geographical situatedness, the biases they bring to the study, their personal investment in and commitment to the study, and so forth” (Leech & Onwuegbuzie, 2008, p. 201).
In this article we outline a process of conducting a debriefing interview of an interpretive researcher. First, we present five common uses of the debriefing technique in qualitative research. Second, we present the new type of debriefing, that of debriefing the researcher, and discuss possible questions that the interviewer might ask the interpretive researcher. These questions are designed to tap potential sources of bias in terms of the researcher's interview background or experience; perceptions of the participant(s); perceptions of nonverbal communication; interpretations of interview findings and interpretations; perceptions of how the study might have affected the researcher; perceptions of how the researcher may have influenced the participant(s); awareness of ethical or political issues that might have arisen before, during, or after the interview(s); and identification of unexpected issues or dilemmas that emerged during the interview(s). Third, we present possible interview questions that are designed to document the degree that the researcher has met the authenticity criteria prescribed by Guba and Lincoln (1989). The five authenticity criteria are fairness, ontological authenticity, educative authenticity, catalytic authenticity, and tactical authenticity. Finally, using Miles and Huberman's (1994) framework, we illustrate how displays such as matrices can be used to collect, analyze, and interpret debriefing interview data (e.g., partially ordered, time-ordered, role-ordered, conceptually ordered, and effects displays) as well as to leave an audit trail. We contend that this process of interviewing the interpretive researcher can provide richer data that can enhance understanding (i.e., increasing Verstehen; Dilthey, 1961; Martin, 2000; Outhwaite, 1975) of the underlying phenomenon, thereby addressing to a greater extent the crises of representation, legitimation, and praxis.
The five commonly used types of debriefing
Until now there have been five major types of debriefing in qualitative research: peer debriefing, debriefing the participants on completion of the study, debriefing the gatekeeper, debriefing occurring among multiple researchers involved in the same study, and debriefing focus group moderators (Leech & Onwuegbuzie, 2008). To lay a foundation for understanding the importance of debriefing, we will briefly discuss each of these types below.
Peer debriefing helps the interpretive researcher to undertake an external evaluation of the research process (Leech & Onwuegbuzie, 2008; Lincoln & Guba, 1985; Maxwell, 1992, 2005; Merriam, 1988). This type of debriefing is essentially another form of intercoder reliability, the major difference being that it is not empirically based but logically based. Lincoln and Guba have depicted the role of the peer debriefer as the devil's advocate, an individual who keeps the researcher “honest,” who poses difficult questions about various aspects of the research study (e.g., procedures, interpretations), and who provides the researcher with the opportunity for catharsis by identifying with the researcher's feelings. The peer reviewer can be a colleague or any other person who is not directly involved with the research, whose role is to serve as a “disinterested peer” (p. 308). As noted by Lincoln and Guba, “[peer] debriefing is a useful—if sobering—experience to which to subject oneself; its utility, when properly engaged, is unquestionable” (p. 309).
Another type of debriefing, commonly known as debriefing the participants, involves the researcher interviewing the participants at the end of the inquiry in an attempt to explain the goals, objectives, purposes, and outcomes of the study, as well as to answer any questions or concerns that the participant might have (Leech & Onwuegbuzie, 2008). Most commonly, debriefing of participants occurs when there has been any undisclosed information about the study; however, it also could be conducted to notify participants of some or all of the findings, which exemplifies respect and appreciation for the important contribution that the participants made to the study. Furthermore, debriefing at the end of the study can help to resolve any potential tensions that might have developed between the participant(s) and researcher during the conduct of the research. Debriefing sessions also could be used to negotiate elements of the participant's story that will appear in the final report. Recently Collins, Onwuegbuzie, and Sutton (2006) coined the term prebriefing to denote a form of debriefing that occurs before the study begins. Prior to conducting the study, a qualitative researcher could use prebriefing sessions to identify potential participants, to explain the study protocol, to reiterate the importance of full participation in the investigation, to reassure the participant that confidentiality will be maintained, or the like.
A third type of debriefing occurs when the qualitative researcher debriefs the gatekeeper, who is the person from whom the researcher must obtain approval to gain access to the study's participants or to a research site. Debriefing in this situation often involves the qualitative researcher summarizing the study's findings, particularly those that are of most interest to the gatekeeper and/or that have the greatest implications for the participants or the study site. The goal of such debriefing is to motivate the gatekeeper to provide the researcher future access to participants and/or research sites.
A fourth type of debriefing is when more than one researcher is involved in an investigation. Ideally, the researchers should debrief each other on a regular basis, preferably in a comfortable, private location where they can discuss issues relating to the research process, including data collected, serendipitous findings, ethical dilemmas that emerged, and/or whether modifications to the research plan are needed (Leech & Onwuegbuzie, 2008). Being able to discuss thoughts, perceptions, feelings, and experiences with each other can help researchers probe deeper into the investigation, thereby capturing the participants' voices to a greater extent.
The fifth type of debriefing involves moderators (e.g., lead moderators or assistant moderators) of focus groups (Krueger & Casey, 2000). When there is more than one moderator involved, they should debrief one another immediately after the focus group interview in a private location (Krueger, 1994, 2000; Morgan, 1997). This debriefing session involves discussions of initial perceptions of the focus group participants (e.g., if one focus group participant did not contribute much to the session), noteworthy statements made by the participants, unexpected findings, and whether data saturation occurred (i.e., no new or relevant information emerged relative to previous focus groups). Most important, debriefing among moderators provides researchers with an opportunity to refocus so that they can start afresh with the next focus group (Onwuegbuzie, Dickinson, Leech, & Zoran, 2007).
Debriefing the interpretive researcher: The sixth type of debriefing
These five types of debriefing in qualitative research all represent useful techniques for collecting more data, complementing data analyses, enhancing data interpretation, and assessing the legitimation of findings. However, there is a sixth type of debriefing that has not received attention in the qualitative research literature, namely debriefing, or interviewing, the interpretive researcher.
To debrief the researcher, certain guidelines should be in place to guide the process. First, a trusted and knowledgeable person who is not involved in the study should conduct the debriefing interviews. Preferably, this debriefing interview would be audiotaped or even videotaped. Second, the debriefing interviewer, preferably, should not be a stakeholder; that is, one who depends on or has a vested interest to some degree in the research's findings, interpretations, and conclusions. Third, the interviewer should be someone who has interviewing skills, has experience conducting qualitative research studies, is somewhat knowledgeable about the research topic, and is available for multiple interviews with the researcher. Fourth, the debriefing interviewer should have read the transcripts representing the researcher's interview(s) of the study participant(s). Knowledge of the interview text could help the debriefing interviewer generate additional and or follow-up questions to ask the researcher. Fifth, the interview should take place in a private location and be recorded, and both the researcher and the interviewer should have sufficient time to devote to this important practice.
To assist the interviewer, we present possible questions to consider when interviewing the researcher as well as illustrate how displays (e.g., partially order, time-ordered, role-ordered, conceptually ordered, and effects displays) can facilitate in collecting, analyzing, and interpreting debriefing interview data.
Questions to use when debriefing the interviewer: Bias
Possible questions that the interviewer might ask the interpretive researcher that pertain directly to bias include those that tap the researcher's interview background and experience; perceptions of the participant(s); perceptions of nonverbal communication; interpretations of interview findings; perceptions of how the study might have affected the researcher; perceptions of how the researcher may have affected the participant(s); awareness of ethical or political issues that might have arisen before, during, or after the interview(s); and identification of unexpected issues or dilemmas that emerged during the interview(s).
In a debriefing of the researcher, it can be beneficial to consider a few general guidelines. First, it is important to ask questions in an open-ended manner (i.e., questions that cannot be answered with a yes or no). Doing so will increase the possibility of collecting rich, thick data. Second, the debriefing interviewer should consider asking questions that elicit the researcher's feelings regarding the study. Rich data can be extracted via the researcher's feelings as many decisions can be made based on how the researcher feels about the participants, the context of the study, the topic of the study, and so forth. Finally, questions should elicit information regarding the researcher, including the researcher's perceptions, past experiences, and so forth.
In Table 1 we have included the following eight main areas accompanied by example questions that were developed to guide the interviewer when debriefing the researcher: (a) the researcher's interview background/experience; (b) the researcher's perceptions of the participant(s); (c) the researcher's perceptions of nonverbal communication, (d) the researcher's interpretations of interview findings/interpretations; (e) the researcher's perceptions of how the study might have affected the researcher; (f) the researcher's perceptions of how the researcher might have affected the participant(s); (g) the researcher's awareness of ethical or political issues that might have arisen before, during, or after the interview(s); and (h) the researcher's identification of unexpected issues or dilemmas that emerged during the interview(s). These areas and example questions are only starting points for the debriefing interviewer; it is best if the interviewer knows the research topic guiding the study's design, researcher's background, and the study's goal, purpose, and rationale well enough to create questions that are specific to the study itself. Interview questions should mainly be open ended, thereby allowing the debriefing interviewer to phrase follow-up questions stemming from the researcher's previous responses.
Possible debriefing topics and questions relating directly to bias
Note: Debriefing interviewers are by no means expected to ask all of these questions in this table; rather, our goal in this table is to provide ideas of questions that debriefing interviewers might consider asking.
Questions to use when debriefing the interviewer: Authenticity criteria
Guba and Lincoln (1989) developed a set of criteria that stem directly from naturalistic/constructivist assumptions. These visionary authors called this set of principles authenticity criteria. Guba and Lincoln conceptualized the following five authenticity criteria: fairness, ontological authenticity, educative authenticity, catalytic authenticity, and tactical authenticity. Each of these criteria is discussed below.
Fairness. According to Guba and Lincoln (1989), fairness embodies a quality of balance. Specifically, it represents “the extent to which different constructions and their underlying value structures are solicited and honored within the evaluation process” (pp. 246–247). These constructions must be presented, elucidated, and member checked in as balanced a way as possible. Moreover, fairness relates to the thoughts, perceptions, feelings, concerns, assertions, and experiences of all stakeholders being represented in the text. That is, it involves adequately capturing the voice of all stakeholders. This can be accomplished by
deliberate attempts to prevent marginalization, to act affirmatively with respect to inclusion, and to act with energy to ensure that all voices in the inquiry effort had the chance to be represented in any texts and to have their stories treated fairly and with balance. (Guba & Lincoln, 2005, p. 207)
In essence, in striving for fairness, the interpretive researcher must be able to identify and negotiate the tensions invoked by contesting the contradictory constructions of the stakeholders. These tensions reflect different ways of knowing, experiencing, and valuing. There are two broad strategies for achieving fairness. The first strategy involves identifying all stakeholders and obtaining their constructions by identifying conflicts with respect to assertions, issues, concerns, and problems. As Lincoln and Guba (1986) have noted, determining differences among the underlying stakeholder constructions can be a challenging task, “but exploration of values when clear conflict is evident should be a part of the data-gathering and data-analysis processes” (p. 79), a task that we believe can be facilitated via the use of debriefing interviews. The second strategy for achieving fairness is via the “open negotiation of recommendations and of the agenda for subsequent action” (Guba & Lincoln, p. 246). This negotiation revolves around addressing unresolved assertions, issues, concerns, and problems.
Ontological authenticity. According to Guba and Lincoln (1989), ontological authenticity represents the criteria for assessing a raised level of awareness among the research participants. In other words, it refers to the extent to which the constructions of the research participants have evolved in a meaningful way as a result of participation in the study. Such authenticity yields “improvement in the individual's (or group's) conscious experiencing of the world” (Lincoln & Guba, 1986, p. 81). Ontological authenticity can be facilitated by the researcher's providing the participants with vicarious experiences that might help them to increase awareness of their own context. Evidence of ontological authenticity can be obtained via statements provided by the individual research participants and via audit trails that document the growth both in the research participants' understanding of their own lives and in the researcher's own “progressive subjectivity” (Guba & Lincoln, 1989, p. 248). Debriefing interviews can play a very important role in helping the researcher leave and articulate an audit trail of both the participants' and researcher's growth. These interviews also can help to challenge the researcher to seek testimonies of participants that provide evidence of their increased awareness.
Educative authenticity. This form of authenticity refers to the extent to which the individual research participants' “understanding of and appreciation for [but not necessarily agreement of] the constructions of others outside their stakeholding group are enhanced” (Guba & Lincoln, 1989, p. 248, italics in original). Moreover, the research participants should realize that the constructions of others stem from the value systems of these other individuals. As is the case for ontological authenticity, evidence of educative authenticity can be gleaned via testimony provided by the individual research participants and audit trails that document the growth both in the research participants' and researcher's understanding. Again, debriefing interviews can be useful here.
Catalytic authenticity. This form of authenticity pertains to the extent to which the new constructions and appreciations of the position of others that have evolved during the course of the study lead to some action(s) taken or decision(s) made by the participants. According to Guba and Lincoln (1989), three strategies can be used to assess whether catalytic authenticity has been met: (a) the researcher obtains testimony from all the participants and stakeholders regarding their interest in and willingness to act on the increased understanding, (b) the researcher obtains testimony regarding the joint actions of participants who have come to resolutions stemming from negotiations of tensions invoked by contesting and contradictory constructions of the stakeholders, and (c) the researcher systematically follows up within a given time frame to assess the extent to which participants' action(s) and decisions(s) stem from the increased understandings that emerged during the course of the study. Debriefing interviews can especially help to optimize the first two strategies.
Tactical authenticity. This form of authenticity pertains to the extent to which participants and stakeholders are empowered to act on the increased understanding that emerged as a result of the study. Guba and Lincoln (1989) have identified three strategies that can be used to assess whether tactical authenticity has been met: (a) the researcher obtains testimony from all the participants and stakeholders regarding the pathways and agency to action, (b) the researcher systematically follows up within a given time frame to assess which participants and/or groups do act and how they act, and (c) the researcher and participants (jointly) assess the degree of empowerment that took place during the study. This transforms the participants to being not only co-constructors of knowledge but also change agents. Once again, debriefing interviews can play a vital role.
In Table 2 we have included example questions pertaining to each of the five authenticity criteria. As is the case for Table 1, these example questions are only starting points for the debriefing interviewer. It can be seen that these questions are mainly open ended, thereby allowing thick, rich data to be collected. These questions can be combined with a selection of questions from Table 1.
Possible debriefing questions relating to authenticity bias
Elements of debriefing interviews
Optimally, in the debriefing interview, discussion would center not only on the study participant's comments but also on his or her nonverbal communication. Such nonverbal communication includes proxemic (i.e., use of interpersonal space to communicate attitudes), kinesic (i.e., body movements or postures), chronemic (i.e., use of pacing of speech and length of silence in conversation), and paralinguistic (i.e., all variations in volume, pitch, and quality of voice) (cf. Gorden, 1980). For example, the debriefing interviewer could ask the researcher to explore her or his reactions to observed nonverbal communication during the participant interview.
Although debriefing interviews are undertaken after all of the participant interviews have been conducted by the researcher, it also is plausible to embed within the study a series of debriefing interviews that could be undertaken throughout the interview process. In particular, in studies involving more than one participant being interviewed, debriefing interviews can be conducted after each interview. One benefit of conducting a series of debriefing interviews is that these interviews can serve as an audit trial documenting the evolution of the researcher's thoughts, perceptions, feelings, experiences, and the like over the course of the study. A second benefit is that the researcher can use debriefing data to identify problems that emerged during previous interviews and thus make appropriate adjustments to subsequent interviews.
In its most postmodern form a debriefing interview would involve a conversation between the researcher and the debriefer, resulting in the co-construction of knowledge. That is, the researcher, in collaboration with the debriefer, “activates diverse communicative resources as an integral part of exchanging questions and answers” (Holstein & Gubrium, 2004, p. 154). In such cases it might be appropriate for both the researcher and the debriefing interviewer to collaborate in analyzing the debriefing data. This collaboration can serve as a form of member-checking to confirm data's trustworthiness and plausibility. In turn, the researcher should consider asking the debriefing interviewer to coauthor the final report, which can represent some form of negotiated text.
It should be noted that if there was no other debriefing interviewer available, the questions posed in Tables 1 and 2 might be used with the interviewer him- or herself to frame more systematically a personal critical (autoethnographic) self-reflection. Notwithstanding, such a self-interview likely would not generate the same quality of data compared to a postmodern interview involving a second person because it is easier for another person (i.e., interviewer) to probe a response from the researcher (i.e., interviewee) than it is for a person to probe himself or herself. As such, whenever possible, we recommend that one or more persons be used to interview the interpretive researcher.
Use of displays in debriefing
We recommend that once the debriefing data are collected, the researcher consider using Miles and Huberman's (1994) framework, which includes matrices and displays, to collect, analyze, and interpret debriefing interview data as well as to leave an audit trail. Excel spreadsheets or a computer-assisted software package (e.g., NVIVO, QDA Miner) could be used to construct any of these matrices. The cell entries in these matrices might contain statements (i.e., transcriptions of data in raw or reduced form) made by both the participant(s) during the interview(s) and the researcher during the debriefing interview. These cell entries then would be coded and subjected to critical reflections that involve the search for patterns and processes. As such, the matrices would allow the interpretive researcher to leave an audit trail, which is recommended by many qualitative researchers as a method of evaluating legitimation or increasing legitimation, or both (Halpern, 1983). Moreover, use of these matrices would facilitate the research findings and interpretations being embedded within its context to a great extent. In the following section, we will discuss how five of the matrices/displays,—a partially ordered matrix, a role-ordered matrix, a conceptually ordered matrix, a time-ordered matrix, and an effects matrix—can be used with data from a debriefing interview.
Partially ordered displays. If the interviewer feels that it would be helpful to “uncover and describe what is happening in a local setting [i.e., with the researcher in regard to the study], no matter how messy or surprising it may be” (Miles & Huberman, 1994, p. 102), a partially ordered display might be helpful. One example of a partially ordered display that could be used by the interviewer is a checklist matrix. Checklist matrices can help the interviewer focus on one or several important concepts. For example, if the concepts of bias were emerging from the data (i.e., Bias A and/or Bias B), the interviewer might create a checklist matrix. There would be two columns, one for the area where bias might be present and one for examples of the bias. The rows would be areas where bias may be present (e.g., toward participants, toward the topic). This checklist could be expanded to include a column for how the bias impacted the results and/or interpretations.
A context chart also could be used to display the debriefing data. According to Miles and Huberman (1994), context charts are networks that display in graphic form the interrelationships among the roles and groups/organizations that “go to make up the context of individual behavior” (p. 102). For example, a context chart could be modified to graph the interrelationships among participants who have been grouped by the researcher based on their responses to one or more interview questions and the researcher's past (relevant) roles and the experiences that emerged from these roles. Such a chart might help the researcher place participant responses in a clearer context.
Role-ordered matrices. Another example of a display that might be helpful is role-ordered matrices. Miles and Huberman (1994) have defined a role as “a complex of expectations and behaviors that make up what you do, and should do, as a certain type of actor in a setting” (p.122). In this situation the role for the researcher represents being a researcher in the specific context of the study.
Creating a role-by-time matrix can illustrate how the researcher perceived her or his role over time (e.g., from a peripheral-member researcher to an active-member researcher in which she or he is involved with the central activities of the study participants, and/or to a complete-member researcher in which she or he becomes fully affiliated with the participants; cf. Adler & Adler, 1987) as different participants in the study were included. One way to set up the matrix is to have the row represent the researcher and the columns represent different participants (in the order they were included in the study or interviewed).
Conceptually ordered matrices. Conceptually ordered matrices can be used when the interviewer wishes to investigate simultaneously multiple concepts that have emerged from the data. For example, a conceptually clustered matrix, a type of conceptually ordered matrix, could be constructed in which the rows and columns are arranged to cluster items that are related theoretically, thematically, or empirically. For debriefing data, the rows could be the researcher's perceptions of the participants at the beginning of the interview, the middle of the interview, and the conclusion of the interview. The columns could be the different interview questions or clusters of interview questions. In constructing and reflecting on this matrix, the researcher would be seeking “conceptual coherence” (Miles & Huberman, 1994, p. 127).
Alternatively, a folk taxonomy could be used. A folk taxonomy typically represents a hierarchical tree diagram that displays how a person classifies important phenomena. Such a display could be used with debriefing data in which a hierarchical tree diagram is constructed to identify and exhibit how the researcher classifies emergent themes (Miles & Huberman, 1994).
A cognitive map, which displays the researcher's representation of concepts pertaining to a particular domain, also could be used, for example to illustrate the complexity of a particular researcher bias (i.e., Bias A or Bias B). Through the use of nodes and links, the debriefing data could be used to link (nonhierarchically) emergent themes extracted from the participant data to a specific bias revealed during the debriefing interview (Miles & Huberman, 1994).
A case dynamics matrix, which displays a set of elements for change and links the consequential processes and outcomes for the purpose of initial explanation, also could be employed. For example, the rows could comprise emergent themes extracted from the participant data, whereas the columns could contain the researcher's underlying assumptions and biases over the course of the participant interview. As such, it can help the researcher to identify the extent to which the emergent themes were caused by the researcher's assumptions and biases (Miles & Huberman, 1994).
Time-ordered matrix. Another method for analyzing how concepts and themes change over time is to use a time-ordered matrix (Miles & Huberman, 1994). The columns are time periods (e.g., beginning, middle, and end of the interview; first, second, third, etc., interview of the same participant), and the rows are concepts that the researcher is interested in investigating further. For example, each summary of the researcher's responses to the debriefing questions can be placed in a column. The rows then could be time-ordered if the analysis suggested that the researcher's reflections had a time component (i.e., time-ordered matrix).
Effects matrices. Effects matrices could be used when there are “ultimate” outcomes (Miles & Huberman, 1994, p. 137). For example, the ultimate outcomes could be the themes/codes that emerge from the original interview that were extracted by the researcher. The effects might be researcher reflections that emerge from the debriefing interview. Such a matrix would allow the researcher to directly compare these effects (e.g., biases that were extracted from the debriefing interview) to the ultimate outcomes (e.g., researcher's interpretations of each participant's response).
Concluding thoughts
In this article we outlined how the interpretive researcher could be interviewed or debriefed. We presented five types of debriefing in qualitative research, designed to assist the qualitative researcher in understanding how helpful debriefing can be in the context of research. We introduced a new type of debriefing, namely that of debriefing the researcher. Next we presented eight main areas accompanied by example questions to guide the interviewer when debriefing the researcher. Then we presented five authenticity criteria developed by Guba and Lincoln (1989) and included possible interview questions that are designed to document the degree that the researcher has met these criteria. Finally, using Miles and Huberman's (1994) framework, we illustrated how displays such as matrices can be used to collect, analyze, and interpret debriefing interview data as well as leave an audit trail.
The debriefing interview has logical appeal because it promotes a reflexive approach to interviews. Moreover, to some degree it mirrors what occurs in helping professions such as counseling and social work, wherein professional helpers (e.g., counselors) who experience vicarious trauma, burnout, compassion fatigue, and the like are encouraged to debrief informally and process traumatic material with supervisors and peers (Horwitz, 1998; Regehr & Cadell, 1999; Urquiza, Wyatt, & Goodlin-Jones, 1997). Accordingly, debriefers also can serve as support members for researchers who experience countertransference, a phenomenon that occurs when the researcher unknowingly transfers his or her own feelings, beliefs, and so forth onto the participant, or other negative reactions after interviewing a participant who discloses a traumatic experience.
Our rationale for developing the interpretive interview framework described in this article is based on our belief that interviewing (i.e., debriefing) the interviewer (i.e., researcher) has significant potential for transforming the interview process in qualitative research studies, going far beyond the use of field notes. Indeed, our framework here is consistent with Holstein and Gubrium's (1995) concept of active interviews, wherein interviews represent active meaning-making endeavors.
Our concept of debriefing the interviewer promotes the idea that the interview process represents a contextually based story that is not only co-constructed by the participant and the interviewer (cf. Gubrium & Holstein, 2002) but also reflects a collaboration between the researcher and debriefing interviewer. More specifically, debriefing interviews allow the researcher to go beyond telling the story (i.e., the “what”) by making transparent the processes, negotiations, and other interactive facets that occur both between the participant and the interviewer and between the researcher and debriefing interviewer (i.e., the “how”) and incorporating this information into the final report. Indeed, if the researcher deems this form of storytelling appropriate and meaningful, the debriefing interview process could facilitate the researcher's voice being interspersed with the participant's voice in an autoethnographical manner (Ellis & Berger, 2002), which can yield a deeper contextual understanding (Banister, 1999). Moreover, the debriefing interview process can help the interpretive researcher reach the hermeneutic circle, which represents the circle of understanding (Warren, 2002). Simply put, using debriefing interviews as a component of the research transforms the interview process into what we call a methodology of story-sharing. In addition, we contend that the debriefing interview process can be used to extract more meaning from data collected via other means, such as observations and focus groups. These extensions to the debriefing interview framework will be developed in subsequent articles, as will other techniques for analyzing debriefing data.
The debriefing process is particularly useful in transformative-emancipatory research, wherein the researcher's goal is to conduct research that is emancipatory, participatory, antidiscriminatory, and the like, which focuses directly on the lives and experiences of marginalized persons or groups such as women; ethnic, racial, or cultural minorities; individuals with disabilities or exceptionalities; religious minorities; and members of gay, lesbian, bisexual, and transsexual communities (Mertens, 2003). In such inquiries the researcher serves as an advocate of the study participant(s), hoping to be able to use the findings to advocate social policies to reflect the interests of the participant(s) and, subsequently, elevate the visibility of marginalized groups in society. Here, debriefing interviews can help ensure that the advocacy stems directly from the study participant's voice and does not merely represent preconceived biases that the researcher introduces overtly or covertly into the study.
In advancing our debriefing interview concept, we are providing an alternative epistemological model of the postmodern interview in an attempt to “expand our ways of understanding how we come to know about our inner lives and social worlds” (Holstein & Gubrium, 2004, p. 157). We believe that by using the process of interviewing the researcher, the data obtained can provide richer information that can enhance understanding (i.e., increasing Verstehen; Dilthey, 1961; Martin, 2000; Outhwaite, 1975) of the underlying phenomenon, thereby addressing, to a greater extent, the crises of representation, legitimation, and praxis. However, we recognize that our alternative model of co-constructing knowledge adds a layer of complication to the interpretive interview process, yet we believe that this is offset by the fact that this model of interviewing encourages the researcher to go beyond attempting to document what was stated by the participant in an interview. Rather, the debriefing interview challenges the postmodern researcher to reflect on what was said by the participant in relation to “how, where, when and by whom experiential information is conveyed, and to what end” (Holstein & Gubrium, 2004, p. 158). Furthermore, the debriefing interview helps to transform the interview process by enabling the researcher to recognize the processes underlying the construction and interpretation of interview data. The participant's comments are not viewed as data stemming from a fixed single source that “speak for themselves”; rather, the interview process is viewed as a meaning-making process that involves the deconstruction of the participant's voice by showing how what is being said relates to the thoughts, perceptions, feelings, and lived experiences of both the participant and the interpretive researcher. Thus, via the debriefing interview process, a new form of deconstruction-ism occurs (cf. Derrida, 1976), in which, as noted by Fontana and Frey (2005),
the influence of the author is brought under scrutiny. Thus, the text created by the rendition of events by the researcher is “deconstructed”; the author's biases and taken-for-granted notions are exposed, and sometimes alternative ways of looking at data are introduced (Clough, 1998). (p. 714)
We recognize that debriefing interview brings to the fore its own set of methodological and analytical guidelines, principles, and stances. However, we hope that this article represents a first step towards capturing the voice in an even more meaningful and trustworthy manner. At the very least, we hope that the ideas we have presented heretofore motivates a conversation among interpretive researchers.