Content area
Full Text
Abstract
Issues associated with the analysis of qualitative data receive much less attention in the market research literature than data collection issues. In the absence of defined analysis procedures and standards, qualitative data analysis is perceived as an idiosyncratic process, relying heavily on the personal biography and the philosophical stance of the analyst. One of the more recent developments in qualitative research is the increasing use of computers and software programs designed specifically to assist with data management and analysis. These programs operate on the assumption that there are commonalties in the various approaches to data analysis. There are few published reports on programs in the market research literature and the aim of this article is to examine the assumptions about data analysis on which these programs are based, review how programs can help with analysis and discuss some of the methodological issues associated with their use. Additionally, we suggest how programs can be used by qualitative market researchers.
Introduction
Matthew Miles (1979) described qualitative data as an attractive nuisance; rich in context and meaning but presenting formidable problems for the analyst. Qualitative data are voluminous and even a small number of groups or depth interviews will generate many hours of video or audio-recorded material, pages of interview transcripts and interviewer notes and observations. Usually these raw data lack an obvious and identifiable structure, particularly where the researcher has adopted an iterative approach to data collection and sampling whereby early data and respondents shape subsequent decisions as to what further data are gathered and from whom (Spiggle 1994).
Qualitative data analysis can be a solitary, labour-intensive and largely manual task where the analyst immerses him or herself in the data, working with them until such time as conclusions emerge. Unlike quantitative research, there are few universally accepted procedures and standards for qualitative data analysis which is portrayed as a creative, largely subjective and idiosyncratic process. This, along with the tendency for qualitative researchers not to report on how they reached their conclusions or arrived at their interpretation from the raw data, lead many to doubt the trustworthiness of the findings (Hunt 1994). Indeed the penchant for qualitative researchers to talk about the single illuminating quotation in a pile of transcripts (Hayward 1989), eureka...