The following piece is guest authored by Sociology Faculty Roberta Garner.
The UVA Gang-Rape/Rolling Stone story grabbed my attention because my daughter taught at UVA for four years and lived near Rugby Avenue and the fraternity houses. I found it easy to call to mind the stunning Rotunda and residence halls designed by Thomas Jefferson, the lovely magnolia trees, the crowds of student party-goers, and the complex cultural dynamics of the town. The landscape seemed familiar, but I knew it harbored many mysteries.
I circulated my remarks about interviewing to colleagues and the conversation has been great! Bruce Evensen underlined Margaret Talbot’s point that Sabrina Rubin Erdely (the Rolling Stone reporter) should have interviewed more people and made an effort to talk to the accused. Greg Scott sent me a link to a Columbia Journalism Review article by Judith Shulevitz that invoked Karl Popper and the need for openness to the possibility of “falsification” by data of one’s initial hypothesis. (Someday, as a Marxist, I am going to take on the Popperites’ self-contradictory attacks on Marxist analysis—but that is another story). Here is the reflection I circulated, and I hope we can keep the conversation going.
Research methods: The challenges of interviewing:
The sensational UVA/Rolling Stone gang-rape case demands attention from anyone interested in research methods. I am always surprised when I hear students respond “Let’s interview” to the question: how can we learn more about this situation, action, culture, or institution? Of all research methods, interviewing is the most problematic and the one least suited to neophytes. I want to make clear that I am not out to slam the RS reporter (though Evensen and Talbot’s criticisms are well-taken), attack the victim’s story, bash Rolling Stone for not fact-checking, or debate UVA policies. My focus here is on what happens in interviews and why interview data are extremely complicated and often deeply flawed—exactly why multiple sources and fact-checking are so important.
A lot has been written about how there are huge gaps between what people do and what they say they do (people lie or are mistaken in their answers, so we cannot reach conclusions about their behaviors from their verbal assertions).
But I want to go beyond this obvious defect in interview methods to a more troubling insight, often associated with postmodern ideas about knowledge. The interview is an interactive discourse that elicits and creates stories with only a loose and untraceable association with the verifiable, empirical facts that we would like to establish—facts about behaviors, cultural norms, practices, beliefs, and so on. Without unreservedly embracing the postmodern view of interviews, Greg Scott and Roberta Garner give it careful attention, saying “We have to be aware that we are constructing social reality by the way we ask questions and that the data we have produced (not collected) in the form of an interview transcript are the discourses in which we participated (2013: 281).”
Consider two alternatives to the interview: One is the first-person audio-taped or written narrative that expresses the experiences and feelings of an individual –a subject, not only in the sense of research subject, but in the sense of an “I” who takes action and can be held responsible for it. The narrative might be a lie or a self-deception, but we know who produced it. The other is the research report of a trained social scientist or journalist that expresses the observations and data collected by an individual: the material may contain other people’s lies and self-deception, as well as incorrect interpretations proffered by the researcher; but an individual researcher or journalist is responsible for the account and must defend its objectivity, show awareness of misrepresentations that may have been recorded from research subjects, and reflect on contending interpretations.
Carry out two corresponding thought experiments for the UVA incident: Imagine Jackie writing an account of what happened to her (as trauma victims are often encouraged to do to contribute to their recovery); and imagine the reporter spending several years writing an ethnography about UVA and its Greek life. Would these two narratives resemble the Rolling Stone interview-based article?
In contrast, the interactively and collaboratively created interview narrative—the result of a dialog—exists in a bright, shiny “interview space” where no one is exactly responsible for the product. It is brought into being in an interaction, and it is difficult to ascertain the limits of responsibility for what is said and how it is framed, even if we have a complete transcript of it. I believe that this ambiguity and haziness would have arisen even if Erdely had interviewed more individuals.
The “bubble space” of the dialogic interview (especially the one-shot, parachuted-in type of interview) becomes even more problematic when we are not “merely” trying to establish facts (Who? What? Where? The name of the fraternity?), but venturing conclusions about “culture.” How often we read about “culture”—corporate culture, ghetto culture, the culture of rape, and so on….as if culture were a sort of tarpaulin that we can pull down and tie over a group of people—as if there were no variation, ambivalence, transgression, and contradictory performances among these people. The interview—decontextualized from a complete ethnography — is a particularly difficult method for reconstructing the way people live cultures. This problem is related to the “attitudinal fallacy” discussed in a recent ASA paper by Jerolmack and Khan, but it goes even further: not only is there a gap between what people do and what they say, but they consistently do and say many contradictory things, in a tangle of words and actions that (as Jerolmack and Khan point out) cannot be learned from interviews.
Jerolmack, Colin and Khan, Shamus. “Talk is Cheap: Ethnography and the Attitudinal Fallacy.” (ASA paper)
Scott, Greg, and Roberta Garner. Doing Qualitative Research: Designs, Methods, and Techniques. Pearson 2013.
Shulevitz, Judith. 10:15 AM – December 6, 2014 What happened at Rolling Stone was not Jackie’s fault Overcome confirmation bias with responsible reporting.
Talbot, Margaret. New Yorker article on December 7, 2014
Note: Roberta Garner is a full professor and Interim Department Chair of Sociology at DePaul University. Her research interests are political sociology, sociological theory, urban sociology, and research methods.