Richard Holton (MIT) writes:
There's now quite a widespread perception amongst philosophers that interviews don't do any good, and may do some harm, in hiring the best faculty. Much of this is based on the belief that there is a body of psychological research that has established that this is the case. I had subscribed to this myself, though with something of a bad conscience. Whilst I was familiar with plenty of general research that might make one sceptical of interviews (work on salience effects and the like), and I knew the much cited discussion in Robyn Dawes' book House of Cards of how interviews hadn't helped find better students for Houston Medical School, I didn't know of any studies that had specifically investigated the effectiveness of interviews of different types in selecting the best people for jobs.
So over the summer I did a little research, and was surprised at what I found. There is virtually nothing on academic interviewing (there isn't enough of it, and it's hard to get independent criteria for good outcomes), but there is a lot of research on interviewing for other jobs. The data turn out to be fairly complex. It does seem that free-wheeling interviews, with no agenda and no prior expectations of what the candidate should say, are not very helpful. But there is a fair bit of evidence that a well structured interview, especially one that probes the skills that are needed for the job, yields good information. Such interviews typically involve either questions about how the candidate has handled certain things in the past (so-called 'behavioural description interviews'); or about how they would handle certain hypothetical happenings (so-called 'situational interviews'). To overcome salience biases, it is important to have fairly objective criteria of success, and to record the candidates' performance at the time (reliance on memory is not good). With that in place, the evidence is that both kinds of interview provide useful information.
I had a brief correspondence with Allen Huffcutt, one of the leading interview researchers, about the relevance of these findings to philosophy positions. Here are his comments:
Some may argue that academic positions are unique so the results of research do not apply to them. I would argue against that. The process of the interview is to identify key characteristics needed for a position, and then ask questions to assess those characteristics. The research out there has been applied to a vast range of job types, a number of which have very similar key characteristics to faculty positions.
What I might suggest is a carefully thought out approach to hiring philosophy faculty. In the job talk, looks for key elements related to presentation. E.g., are they well organized? Do they find ways to make the material interesting and relevant to the students? Do they deliberately bring people into the presentation (e.g., ask them questions) or just stand there and lecture the whole time? Do they handle questions well?
During the interview, look for key elements pertaining to social interactions. For instance, using the behavior description format, you could ask them to describe a time when they had to handle a difficult student, a time when they went out of their way to help a student, etc. Same for their interaction with other faculty.
None of this is conclusive, but it has changed my mind. I remain convinced that the kinds of unfocused interviews I often had to conduct in the UK, typically with a majority of non-philosophers on the panel, are, at best, unhelpful. In contrast, discussing someone's paper, which is what many APA interviews and on-campus talks consist in, is rather like a mix of a behavioral descriptive interview and a situational interview; you spend much of it finding out how the candidate responded to various philosophical problems (BDI), and then you find out how they would respond if someone said something like what you then say (SI). Of course there remain worries: you need to work hard to make sure that the knowledge from the interview doesn't crowd out the knowledge you have from other sources (and typically interviews don't provide much information about teaching abilities); there is a real question about whether interviews tend to increase the effects of sexual and racial bias (I don't know what the findings are on this); and there is a question of whether it is worth the cost of both interviewing at the APA and bringing candidates onto campus (I think that for us it's not). But I do now think that well conducted interviews or on-campus talks can have a useful role. At any rate, there's no knock-down evidence that they don't.
Comments are open; submit comments only once, they may take awhile to appear. As usual, signed comments will be strongly preferred, and all submitted comments must include a valid e-mail address.