Semi-Structured Interviewing and Qualitative Analysis

Semi-Structured Interviewing and Qualitative Analysis

This chapter is an introduction to conducting and analysing semi-structured interviews. It will briefly put the use of this method within a theoretical context and will then outline the various stages of conducting a semi-structured interview project – producing an interview schedule, conducting the interview, analysing the material and writing up. This is a practically oriented chapter – intended mainly to help a reader with no previous experience of this type of psychological research method.

In general, researchers use semi-structured interviews in order to gain a detailed picture of a respondent's beliefs about, or perceptions or accounts of, a particular topic. The method gives the researcher and respondent much more flexibility than the more conventional structured interview, questionnaire or survey. The researcher is able to follow up particularly interesting avenues that emerge in the interview and the respondent is able to give a fuller picture. Then by employing qualitative analysis an attempt is made to capture the richness of the themes emerging from the respondent's talk rather than reduce the responses to quantitative categories. While there is no automatic link between semi-structured interviewing and qualitative analysis, and it would, for example, be possible to do a statistical analysis of the frequency of certain responses in an interview, this would be to waste the opportunity provided by the detail of the verbatim interview data. Therefore this chapter assumes a ‘natural’ fit between semi-structured interviewing and qualitative analysis. At the same time, after one has conducted a thematic qualitative analysis, it is also possible (if one wishes and it is appropriate) to include in the write-up some indication of the prevalence of the themes within the data set.

One can in fact adopt a range of theoretical positions when one is conducting an interview study. Broadly speaking, one may, at one [Page 10]extreme, believe that one is uncovering a factual record and a person's responses could be independently verified for their accuracy. At the other extreme one may assume that a person's responses form part of a locally organized interaction structure. The participant is answering in this way in order to perform certain interactive functions, for example appearing to be a good interviewee, or using expressions in order to convince the interviewer that he or she, the respondent, is an expert on this topic. It may, in the most extreme case, have no relationship to either a world outside (the factual record) or a world inside (beliefs, attitudes, etc.).

Between these two positions, one may consider that what respondents say does have some significance and ‘reality’ for them beyond the bounds of this particular occasion, that it is part of their ongoing self-story and represents a manifestation of their psychological world, and it is this psychological reality that one is interested in. The talk will probably also have some relationship to a world outside, though that is not the crucial point, but it will also be affected by the requirements of this particular interaction (Smith, 1995b).

This chapter is written from this middle position. It is assumed that what a respondent says in the interview has some ongoing significance for him or her and that there is some, though not a transparent, relationship between what the person says and beliefs or psychological constructs that he or she can be said to hold. This approach can be described as adopting a phenomenological perspective (see Giorgi, 1995). At the same time it is recognized that meanings are negotiated within a social context and that therefore this form of interviewing is also drawing on, or can be seen from, a symbolic interactionist position (see Denzin, 1995).

...