Themes

In qualitative research, data collection typically occurs to the point of saturation. Essentially, this means that researchers continue interviews to the point where little new information is shared by participants. In other words, people continue reporting essentially the same ideas and the law of diminishing returns is at work in the information-gathering procedure. Collecting more data, at that point, does not produce novel results.

Thematic analysis is one of many methods used to assess whether or not saturation has occurred in the data collecting process. Typically, reoccurrence is a prime means of analyzing data for themes. That is, researchers assess the interview transcripts for repeated statements, phrases, and words.

However, themes are not generated simply by counting words. Rather, themes are assessed by examining constructs that occur in the data. For example, the terms anger, upset, frustrated, or mad may be used by various participants, but all refer essentially to the same overarching construct. Computer-assisted software often can be helpful to qualitative researchers when assessing data for related ideas, using different words or phrases.

Themes typically are derived from codes generated by the qualitative researcher. Reading the material presented by the participants multiple times, using constant comparison among the ideas presented throughout the interviews, can become overwhelming to the researcher. Consequently, experienced qualitative inquiries involve researchers making memos to themselves as the data collection process occurs. These memos may involve hunches, impressions, or ideas for further exploration. Tracing through one's memos often prompts the researcher for potential reoccurring ideas.

These potential reoccurring ideas are coded, simply meaning that the constructs are physically noted (handwritten or in a software database). Sometimes what originally appeared to be a theme does not turn out to be aptly supported by the data. Those codes eventually are ignored or discarded. Other codes, however, occur repeatedly—both in terms of breadth, and, when appropriate, depth of occurrences. These reoccurring coded phrases, terms, and expressions (and the like) formulate constructs that seem to be shared by most or many of the participants of the study. When sufficiently grounded in the data collected, they become the study's themes.

Quantitative researchers sometimes overlook the inductive method of qualitative themes. As a hallmark, qualitative researchers use an inductive method of analysis. Most quantitative researchers begin with hypotheses that they attempt to prove or disprove statistically. Essentially, it is a deductive method. The researcher begins with a conclusion (null hypothesis) and goes to data for its support. Qualitative researchers, in contrast, are inductively driven. They begin with the data, and from it, develop hypotheses or conclusions. Most often, the themes generated from the data are the study's conclusions.

Finally, themes are assessed by qualitative researchers in more than just transcribed interviews. Triangulation involves sifting through documents, websites, test results, public relations materials—anything that might substantiate or negate the stability of themes found from participant interview transcripts. In other words, information from related documents should support the overall themes generated from the coded transcript data.