To interview or not

David Silverman is a professor emeritus at Goldsmiths, University of London, and author of a number of warks on Qualitative Research, including the snappily, and accurately, titled A Very Short, Fairly Interesting, Reasonably Cheap Book about Qualitative Research (Sage, 2007).  Because his work is accessible yet scholarly, and because many of my dissertation students choose topics with a qualitative emphasis, I often recommend that they read some of this work.

In its introduction, the short, interesting, and cheap book does warn that it’s also opinionated, and the most striking opinions expressed there are about the shortcomings of interviews as a source of original data.  To precis a rather complex subject in a few words, the argument is that interview results tend to reflect the choice of topics and questions, so can say more about the interviewer than the interviewee, and also that interviewees tell us what they’d like to tell the outside world, which may not reflect what they are really doing or thinking.  If you are studying the behaviour of a group of people (such as a team working together in a business) you will get a more accurate picture of the dynamics among the group by observing conversations, or even by intercepting emails between the people, than by carrying out interviews.

You can read the relevant section on a blog at and  watch David Silverman give a succinct account of his concerns about interviews, and some thoughts about what constitutes naturally occuring data, on YouTube at .  As an aside there is a lot of material on research methods on YouTube.

In the context of student dissertations, I think that these objections to interviews should be acknowledged.  But nevertheless interviews form a valuable tool and an effective way of gathering primary data, and my recommendation to students would be to mention the shortcomings of interviews in passing but, if they are appropriate, to use them nevertheless.

So why defend the value of interviews for dissertations?  To stress a point, I am talking about dissertations typically written by a final year undergraduate in business or management, or perhaps on an MSc in management: different circumstances apply to different courses.

At the undergraduate level it is difficult enough to gather any sort of primary data.  On occasions the availability of somebody to interview can be the key to a student’s ability to produce an analytical and reflective piece of work.

Particularly in business and management, a lot of research is ‘exploratory’ – as with case studies, the results should be treated with some caution.  For exploratory research it can even be counter-productive to be over-concerned about whether interview or survey results are representative of a population.  This is particularly true of areas of business or management that are changing rapidly: there is little value in surveying people about products which haven’t been invented yet (for instance 15 years ago few people would have expressed enthusiasm for a combined mobile phone and music player)

Sometimes interviews can provide privileged access to experts, and a student’s interpretation of an expert’s view can be very interesting and original

The ‘dead social scientist test’ mentioned in the methodspace blog should also be treated with some caution in the context of undergraduate dissertations.  One of the commonest criticisms that I hear levelled against dissertations, especially those that are competent but mediocre, is that the student’s voice didn’t come through, and that there’s no particular distinctive perspective discernible in the work.  Which could just mean that the student was over-zealous in ensuring neutrality of their data.

Of course in practice professional social scientists bring a whole range of views, experiences, and perspectives to their work, all of which contribute to their own distinctive approach, and they need to grapple with the tension between using these and avoiding bias.  But that for an undergraduate, and even a masters, dissertation there is little harm in erring somewhat on the side of demonstrating an individual perspective, provided that the work is based on real data and not just on hunches and conjecture.  And interviews form a clear, solid, and traceable way of acquiring authentic data.


2 Responses to “To interview or not”

  1. erlhov Says:

    Great post on research, Martin, thanks. What do you think of surveys as opposed to interviews? Needless to say, surveys may be easier to get people to respond to, it is narrower in its form, but perhaps broader in terms how how many opinions will be collected. Are there any other forms of research that might be particularly suitable for undergraduate dissertations? May any web 2.0 tools be of use?

    • martinrich Says:

      Thanks Erlend, in general I think a lot of the comments about interviews also apply to surveys. For an undergraduate in particular, it’s incredibly difficult to sample accurately, and there’s a danger that the survey results tell us more about the questions than the answers, but if you note the limitations they can still be very valuable. Incidentally one of the criticisms of both interviews and surveys is that you find out what the subject wants to tell the researcher, not what they tell other people. But the corrollary of this is that the things that people tell researchers can be rather interesting, especially in a survey which offers a measure of autonomy.
      I’m very interested in the use of web 2 tools by students: I did a conference presentation about this over the summer, and my paper on the subject will be on the web as part of the electronic journal of business research methods soon

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: