Qualitative data processing methods

The steps of thematic analysis were laid out wonderfully clearly as "Analyzing text involves several tasks: (1) discovering themes and subthemes, (2) winnowing themes to a manageable few (i.e., deciding which themes are important in any project), (3) building hierarchies of themes or code books, and (4) linking themes into theoretical models" (Ryan & Bernard, 2003, p. 85). While I have written about thematic analysis before (here), there is always more to talk about! For example, there are at least four ways of processing the thematic analysis of our base data:

• Cutting & sorting. This is the “wholes and holes” (Suter, 2011, pp. 348-349, as illustrated). We could consider working with Lego: we first look through all our data and note where the similar pieces or elements fit together. We look for patterns. We deconstruct some pieces and reconfigure them, trying to understand how and why they were built in that particular way. We put them back together. We reorganise them. We log the data we have gathered. If we don’t have access to computer-aided qualitative data analysis software (or CAQDAS), such as NVivo, we could use a sorting hat such as Obsidian to note each data chunk and to theme it. Old school is to paste each phrase up on a library card, noting themes on the back (Ryan & Bernard, 2003); though why on earth anyone would want to have it in an unsearchable form is beyond me. This method can get complex on large projects, so more in-depth reading is recommended.
• Word lists. With CAQDAS (like NVivo), creating word lists is fairly easy. We can cut out certain arrays of common words (such as articles; or common verbs; or the 125 most commonly used words; Ryan & Bernard, 2003). This allows us to gather word count and phrase descriptive statistics, though phrase repetition may require a human eye rather than CAQDAS.
• Word collocation. This is where words appear in the same sentence, or close by (Ryan & Bernard, 2003). Again, CAQDAS can allow us to seek these patterns and provide descriptive statistics, leading to new themes, however this tool is also used with quantitative analysis tools using “co-occurrence matrices and applied factor analysis and dimensional plotting” (p. 97). Note that we should be awake to unintentional media influence: e.g., ‘law and order’; ‘poverty and depravation’ etc with this process.
• Metacoding. This is where we assign codes which connect the themes we have identified; we create a web of the interconnectedness of our ideas. This method too may be used with quantitative analysis methods via matrices. Interestingly, we can test the quality of our metacoding, using negative case analysis (Lincoln & Guba, 1985), where we check whether our emergent theory fits with those odd comments, people, incidents or cases which don’t really seem to fit with rest of the data. If it does fit, then we have a strong theory (Chatham-Carpenter et al., 2013).

It is quite useful to consider these different approaches; and to consider how careful - with a larger group of researchers - that all members must be in documenting how the methods should work... and how they actually work in practice.

Sam

References:

Chatham-Carpenter, A., DeFrancisco, V. P., Hall, T., Martin, C. R., & Palczewski, C. H. (2013). Foundations of Communication Research Methods: Qualitative, quantitative and rhetorical approaches to the study of communication. University of Iowa.

Lincoln, Y, S., & Guba, E. G. (1985). Naturalistic inquiry. SAGE Publications, Inc.

Ryan, G. W., & Bernard, H. R. (2003). Techniques to identify themes. Field Methods, 15(1), 85-109. http://doi.org/10.1177/1525822X02239569

Suter, W. N. (2011). Introduction to Educational Research: A Critical Thinking Approach. SAGE Publications, Inc.