«LEARNING TO INTERVIEW Learning to Interview in the Social Sciences Kathryn Roulston Kathleen deMarrais Jamie B. Lewis University of Georgia A large ...»
Not all students showed that they were able to reflect deeply on the interview process, that is, the how of interaction rather than the what of interaction. Although some students’ reflections devoted a substantial proportion of discussion to preliminary analysis of the content of talk (produced by the respondent) and how this related to the interviewer’s research topic, few Roulston et al. / LEARNING TO INTERVIEW 659 looked deeply at the interview from the perspective of how the interaction was accomplished. This certainly reflects the literature extant in the field of interviewing because it is the content of interviews that is seen to be of primary interest, not how that content was achieved by speakers within the interview setting. Indeed, Briggs (1986) commented that by “leaving the interview situation itself out of the analysis, we have cleverly circumvented the need to examine our own role in the research process” (p. 4).
We take one vivid example to illustrate this point. One of our novice interviewers, while examining some of her interviewing practices and explaining how they might be improved did not identify significant features of her interview in her reflective statement. These included (a) a large number of assessments (Pomerantz, 1984) and formulations (Heritage & Watson, 1979) in her talk and (b) use of closed questions. We discuss each of these in turn.
Assessments. Assessments are statements used by speakers to claim knowledge about that which is being assessed (Pomerantz, 1984). In her analysis of interview data, Kelly (2001) showed how interview respondents may use assessments for the purposes of praise and criticism. In the following, we see examples of assessments provided by Heather in her interview that do the work of praising her respondent.
I bet she was so excited.
You must have been a superstar.
I bet he saw you and just was taken You would have been wonderful.
Kelly’s analyses show how individuals do considerable “identity work” to establish an entitlement to give these assessments or opinions. In the preceding examples, the interviewer’s lack of entitlement to provide these opinions is shown by her use of phrases such as “I bet,” “you must,” and “you would.” Post hoc analyses provide no clue as to what kind of data would have been produced had such assessments been withheld—however, it is useful for interviewers to reflect on what such assessments contribute to the talk.
Formulations. Formulations are statements in which speakers paraphrase prior utterances through preserving, deleting, and transforming information produced by other speakers (Heritage & Watson, 1979). In the following excerpt, we see how the researcher preserves information (her respondent comes from a family of 10 children), deletes information (her respondent’s description of her mother as not doing much “but have babies”), and transforms the information (her respondent’s mother as “busy” and washing clothes for the family).
Participant: I had older brother and sisters, older than I was; I am the youngest of the first five. And they all worked, my brother went into the service and he came out the year before my dad died and got a job and helped out, and so, when I got into 660 QUALITATIVE INQUIRY / August 2003 12th grade I worked on weekends, my mother had never worked public work in her life. She had never done much of anything, but have babies ((laughter)).
Researcher: She was doing a lot, she was busy.
Participant: And my Dad thought she was a queen. She didn’t even wash dishes or do anything, ‘cos she had so many kids.
Researcher: Yes she didn’t have time. Bless her heart. I can’t imagine 10 children, just washing the clothes for 10 kids.
Participant: She never did do that, she probably did it when the oldest ones were small, but she didn’t do it when we grew up, everybody had a job...
In this extract, we see the participant providing responses that show disagreement with the interviewer’s prior utterances. For researchers who propose to gain an understanding of their participants’ views, it is important to be aware of the researcher’s work in making formulations and the consequences for data generated.
Closed questions. When posed closed questions that could have been answered by yes/no responses, the interview respondent repeatedly gave short answers that provided little scope for elicitation of further talk via further probes from the researcher. Two examples of closed questions illustrate
Q: Do you still sing in church?
A: I sing in church.
Q: Do you still have a class reunion?
A: We hadn’t had a class reunion in about 20 years.
In failing to recognize features of their own interactional styles as in the preceding example, interviewers overlook the implications for the kind of interview data they are likely to co-construct with their participants. This is further complicated if the transcription process is investigated in relation to what is audible on the tape. As noted earlier, although most of the participants provided close and accurate transcripts of their interviews, some did not. For example, in several transcripts we read while listening to the audiotapes, some sections of data were missing, possibly because they were considered to be irrelevant by the transcribers.
Other students accurately pinpointed some of their question-posing problems within their reflections. For example, Irene commented on her use of a
I still have problems phrasing some of my questions in a jerky sort of manner.
The question that starts in line 343, “So, with the activities—how do you decide which ones you will use vs., or where do you get them from even?” is an example. Here, I also asked two questions instead of just one.
Roulston et al. / LEARNING TO INTERVIEW 661 Our point here is not to critique these students’ reflective abilities or their beginning work as interviewers but to consider ways of facilitating learning experiences—for example, through providing pertinent readings, timely discussions, and apt questions—that will aid students to reflect deeply on their interviewing practice and the implications for their work as researchers.
CONCLUSION AND RECOMMENDATIONSOur analysis of data revealed that the interview process challenged students. This occurred in multiple ways. Students’ concentration was thwarted by unanticipated behaviors of participants and distractions in interview settings, and the research focus was sometimes lost through the researchers’ own actions, such as poor phrasing and delivery of questions and not listening closely. Some novice interviewers had difficulty constructing and using interview questions that were open ended and focused on their research purpose. In addition, once these questions were refined, they found it difficult in the actual interview not to elaborate extensively, consequently forfeiting the clarity of the question. Using probes to extend an interview participant’s points was challenging for many of these novice interviewers. Some tended to accept a participant’s response and move directly to the next question rather than ask for elaborations or clarifications of meaning, whereas others provided formulations of prior talk that served to transform the meaning of their respondents’ utterances. The interview process was at times complicated by the students’ difficulty in being present or active listeners. Rather than listening, some students reported being engaged in analyzing the way the participants’ experiences fit with their own research interests or thinking about the next question. Students spoke of being overwhelmed by all the things they had to attend to in the interview setting and referred to their selftalk within the interviews in which they worried about their performance as researchers. In several of these interviews, students were confronted with the emotional aspects of the interview process—both in experiencing difficult emotions themselves and in understanding the emotional effect of the interviews on their participants. Furthermore, the transcription process that followed the interview proved to be yet another hurdle for some, often tedious—if not self-revealing.
We conclude the article by presenting some points of consideration for those who teach interview skills for social science research purposes with the hope that we who are teachers and researchers might assist novice researchers who plan to use interviews as a data generation tool to develop sound skills in the area (and not repeat our own errors). Specific strategies to improve the teaching of qualitative interview methods include the following.
Conduct interviews as part of authentic research projects. Engaging students in multiple interviews as part of real-life research studies provides them with 662 QUALITATIVE INQUIRY / August 2003 opportunities to gain and develop skills in negotiating entry with people who are not familiar to them. In authentic studies the purpose of the interview is clear, whereas this may not be so when students conduct interviews (often with known individuals) for the purposes of a class assignment. Students are also more likely to feel a responsibility for high-quality interviewing that may not be present in a practice interview conducted with a class member or peer.
Close, guided analysis of interview tapes and transcripts. We suggest multiple ways that close, guided analysis of interview tapes and transcripts might be achieved, including peer review of transcripts and audiotapes, use of videotaping with peer review, and small group analysis of interview data, including attention to content and process. As a starting point, we include possible topics and questions for analysis and reflection in Appendices A and B.
Class discussions concerning research design and researchers’ assumptions. This could include discussion of topics such as the interrelationship between research questions and focus and the interview questions and the researchers’ assumptions and conceptions of interviewing as a research method.
Our initial experiences in using these strategies with other students have proved highly productive and will be subject to further investigation. We believe that it is important to study the development of interview skills by novice researchers for two reasons. First, the research interview is widely used in the social sciences as a method of generating data. As Briggs (1986) noted, “the validity of a great deal of what we believe to be true about human beings and the way they relate to one another hinges on the viability of the interview as a methodological strategy” (p. 1). Therefore, it is imperative that adequate training of social science researchers employing interviews be provided. Second, although some researchers have provided specific guidelines for the construction of interview questions and the conduct of interviews (Foddy, 1993; Kvale, 1996; Spradley, 1979; Weiss, 1994), others have provided strong critiques of interviewing as a research method (Atkinson & Silverman, 1997; Scheurich, 1995). Although we are aware that researchers should be cognizant of such critiques, we argue that with appropriate guidance, novice interviewers can develop more effective interviewing skills. Although there are a plethora of texts about interviewing as a research method, we have found little empirical research regarding the teaching of interviewing to guide us as university educators in the development of curricula. Therefore, this research, which specifically investigates one such interview course, will be potentially useful to those teaching qualitative research methods courses in addition to novice researchers undertaking such courses. We believe that we can assist researchers to develop interview skills and that this development is assisted through guided practice. We conclude by quoting Leah, one of our novice interviewers: “There is nothing more convincing than the saying ‘the most effective way to learn how to interview is by doing it.’” Appendix A Sample Grading Rubric: Interview Project
Qualitative Interview Project The following are questions to consider as you reflect on your interview process and write those reflections for submission with your transcript and tape. Be sure that you provide examples from the tape to support your points.
1. How do you think you did with the explanation of the research purpose?
2. How do you think you did with the explanation of the consent process and form?
3. How did the interview context enable or constrain the interview process?
4. How did you do in building rapport with the participant?
5. What kind of questions did you ask in the interview?
6. What kinds of responses did you get?
7. How did your questions influence the participants’ responses?
8. Did you put possible responses into the questions?
9. Did you ask closed-ended questions? Open questions?
10. Did you use more than one question in your utterance?
11. How did you handle your wait time within the interview?
12. Was there overlapping talk in the interview? Interruptions?
13. Did you use continuers such as um, okay, mm-hmm?
14. Do you treat interviews as conversation? If so, what was your input into the conversation?
15. Did you evaluate the participants’ responses to your questions within the interview? If so, how did the participant respond to this evaluation?
16. Were your interview questions focused on the purpose of the research and your research questions?
17. What would you do differently if you were able to do the same interview again?
18. What suggestions for improvement do you have for your own interview techniques?
1. Pseudonyms are used throughout the article for all students and interview participants.
2. The requirements for the certificate include the completion of five qualitative courses (three of which are core) and a dissertation that uses a qualitative design.
666 QUALITATIVE INQUIRY / August 2003
REFERENCESAtkinson, P., & Silverman, D. (1997). Kundera’s Immortality: The interview society and the invention of the self. Qualitative Inquiry, 3, 304-325.
Baker, C. D. (1997). Membership categorisation and interview accounts. In D.