There are three main criticisms for interviews – specifically semi-structured interviews as a data gathering research method – that Assistant Professor of Communication Matthew Powers highlighted for a record audience at the second QUAL Speaker Series talk this fall. These three criticisms boil down to:
-
Assistant Professor of Communication Matthew Powers described his concept of interviews as social relations and the implications that follow for interviews as a research method.
Interviews conflate attitudes and actions – interviewees give their account of what they think happened or what they would have liked to happen, but not what really did happen;
- Interviews are too focused on individuals – as opposed to systems as an explanation for a phenomenon; and
- Interviews are often challenging in terms of finding coherence in narratives.
While there are good points brought up by these criticisms, Powers says, they treat interviews as a method that’s akin to a recipe in a cookbook. In reality, interviews are social relations and the social positions of the interviewee and interviewer shape the discourses of semi-structured research interviews, Powers argues.
Professor Kirsten Foot asked one of the questions that a record audience at the Matthew Powers talk had the chance to query.
To help illustrate the implications of looking at interviews as social relations, he shared examples from work he conducted with a co-researcher on journalism in France and the United States. As part of that research, Powers and his collaborator interviewed journalists in the U.S. who founded startup news outlets. Looking at the first answer given by an interviewee who successfully launched such a news site, one might conclude that startup news sites are founded by brave individuals in an opportunity/turmoil environment. However, this was the interviewee’s probably rather standard answer they were used to giving family, friends, and perhaps even investors and it came from their social position as a successful news entrepreneur.
Powers and his co-researcher knew from the literature that often those who successfully launch news sites have technical skills in particular, so they pushed, following up to get more details. It was only then that the interviewee told them he had social networks to lean on. While he did not confirm prior research (tech skills), this source did reveal that other factors (such as a social network to provide financial support) were key. It is this deeper probing and not taking at face value the initial answer, that Powers advocates for in order to get beyond the initial “cover” answers interviewees give out of habit and from their social position. The goal is to get to the common answers that illuminate a phenomenon more accurately.
Graduate students had many questions about the idea of viewing semi-structured interviews as social relations and the practical implications for that in their interview protocols.
There, of course, remain limits to looking at interviews as social relations, Powers admits. Most interviewees are people who feel like they have something to say and want to stay it. To remedy this self-selection bias, Powers recommends creative ways to get a strategic selective sample when that is appropriate to the research at hand. In the research he and his collaborator conducted on journalism startups in the U.S. and France, they found a lot of young journalists eager to share their stories, but fewer establish journalists interested in sharing theirs. Snowball sampling perpetuated this pattern.
However, Powers and his co-researcher decide to ask a young journalist in France specifically to introduce them to her mother, who happened to be an established journalist. After a successful interview with the latter, she introduced them to other established journalists in her network, by introducing Powers and his collaborator as friends of her daughter. Everyone was aware that Powers and his colleague were researchers, but their role as such was not enough to break into the establish journalists’ community. It was the trust and validation by a personal referral that opened up their sample to this otherwise elusive demographic.
When viewing interviews as social relations, to adjust accordingly to discover more common answers and to reach more varied voices among interviewees, Powers says researchers must do a lot of preparation work before and during the interviews, and to also do follow up analysis of the data gathered including checking with key informants if the interviewee’s answers ring true.
To see Matthew Powers’ presentation from this QUAL Speaker Series talk, including sample interview questions and answers illustrating his concept of interviews as social relations, click here.
