Software and Qualitative Analysis
7. Drawing and Verifying Conclusions
This is a “next step” only conceptually. In practice, it is rare that you first do all your coding, then do all your conclusion drawing. Far more often, there is an iterative process of coding, exploring conclusions and explanations, coding further, testing conclusions, etc. More on this iterative process will be presented later.
The process of drawing conclusions begins early in the coding process. Even as you first begin reviewing and coding your data, you are beginning to form ideas about the important phenomena they indicate, and beginning to generate propositions about these phenomena, and the relationships among them. Once the data are coded the researcher will begin to pursue those early propositions, and to examine the data for other conclusions.
The first step here is often to examine the individual themes which have been identified. The researcher gathers together all of the text passages coded for a theme. Reading all of these passages together (while also referring back to their original contexts for accurate interpretation) can enable the researcher to better understand the theme. Often it becomes clear that there is more than one theme captured by the code, and it must be partitioned. At other times, after reading several themes it becomes clear that several should be combined, or one subsumed within another. Eventually, the researcher is able to write a faithful representation of all the data on that theme. Such a summary captures the common features of the reports of the different respondents, but avoids glossing over differences among them. It is important to illustrate each of these theme summaries with quotes. This allows the reader to assess the interpretation the analyst has made. Further, it is important to have a scheme for tracking respondents in these quotes—say with ID numbers, pseudonyms, or fictitious initials—so that the reader can see that you are not overly relying on a subset of respondents. This kind of use of quotes thus constitutes a fundamental tactic for insuring reliability in qualitative research.