Thematic analysis pattern seeking

There are at least two types of thematic sources: a priori, where deducted themes are already identified – perhaps from a model or previous research work; and emergent, where inductive themes arise from the data (there is a third, a posteriori, which I will tackle in a later post). 

When seeking themes, there needs to be some systematic method for scrutiny, and some method for marking up the material to show the layers of thinking and seeing: even the act of “underlining key phrases ‘because they make some as yet inchoate sense’” (Ryan & Bernard, 2003, p. 88, citing Sandelowski, 1995, p. 373). Some approaches we can use to seek patterns in our data are as follows:

• Repetitions. Themes may be notable in the research data if they are repeated, or if, in the words of Guba, they form “recurring regularities” (1978, p. 53, as cited by Ryan & Bernard, 2003, p. 89).
• In vivo. This is where “the data is rooted in the participant’s own language” (Saldana, 2009, p. 6). Language is localised and may be used in specialised ways. For example, the term ‘skiving’ means something quite different to saddlers and teenage boys: to one it is the act of curriering a leather edge; to the other it is the act of bunking-off school, or of pretending to work. This used to be called “indigenous terms” which feels a bit colonialist these days.
• Metaphors. As humans we tend to present our thoughts as metaphors; as a story; as a common narrative. Metaphors tell us more than our words alone – they “communicate[…] unshared experiences typically involving a mapping process from the more concrete to the more abstract” (Creed & McIlveen, 2017, p. 18). The search for metaphors can become a method of analysis, an approach to make sense of the data.
• Transitions. It is not just what we say, but how we say it. For example, we may spring away from things, avoid things, or simply provide a change in direction, pace, tone or emphasis that also tells a story (Ryan & Bernard, 2003). Transitions may indicate many things, and researchers must be careful not to make unsubstantiated or unwarranted attributions.
• Similarities & differences. This is where we compare our textual data line by line, asking “What is this sentence about?” and “How is it similar to or different from the preceding or following statements?” (Ryan & Bernard, 2003, p. 91). We tell the story of the story, including the degree of shift as it progresses. We will compare interview with interview at the highest level.
• Causal language. While similar to transitions, this is also known as “linguistic connectors”, where the participant makes a pattern of showing the connection between an idea and a result (Ryan & Bernard, 2003): becauses; resultings; insteads; whens; and thens. Because this happened, this was the result, maybe demonstrating locus of control (Lefcourt, 1981). There may be a lot of time-factoring (afterwards, before) or use of negative terms (not, none, non- etc). Relationships may also be causal: attribution, functional, compared, high status or low, hot or cold; lucky or not, etc (Ryan & Bernard, 2003).
• Missing data. Also known as “lacunae” (missing pieces; lacuna is the singular). Missing data is the “everyone knows that” stuff; such as the need to turn a video recorder off after setting it up to record (which not everyone knows). Researchers will need good emic (Peters, 2015) knowledge in order to be able to spot what data is missing.
• Echoing theory. Interrogating our data for theories or theoretical components relating to “social conflict, cultural contradictions, informal methods of social control, [how] impersonal social relationships [are managed], status [acquisition and maintenance], and […] how people solve problems” (Ryan & Bernard, 2003, p. 93, citing Spradley, 1979) may also aid our data analysis, providing we don’t let it get too complicated.

There suddenly becomes much more detail we can supply in a methods section when we consider that in one single method - thematic analysis - there are so many ways of organising the data. 

Sam

References:

Creed, A., & McIlveen, P. (2018). Metaphor identification as a research method for the study of career. International Journal for Educational and Vocational Guidance, 18(1), 27-44. https://doi.org/10.1007/s10775-017-9345-2

Lefcourt, H. M. (Ed.). (1981). Research with the locus of control construct: Assessment methods. New York, USA: Academic Press.

Peters, B. (2015). Qualitative Methods in Monitoring and Evaluation: The Emic and the Etic: Their Importance to Qualitative Evaluators. American University. https://programs.online.american.edu/msme/masters-in-measurement-and-evaluation/resources/emic-and-etic

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

Saldaña, J. (2009). The Coding Manual for Qualitative Researchers. SAGE Publications Ltd.

Spradley, J. P. (1979). The Ethnographic Interview. Harcourt Brace Jovanovich College Publishers