28 Feb 2017

Analysis continued

I’ve been trying to get my head around the data analysis side of things. The Braun & Clarke (2006) diagram of thematic analysis helps and I will definitely use this approach. I have been taken of on a tagent a bit by use of the AntConc software. It’s really interesting to see commonality in data items, but it’s difficult to know how or if to interpret patterns that are more than a word count.

I think i’ll still use the tool, but I’m re-reading the transcriptions now to look for emergent themes.

What seems to stick with me though, is those themes from the lit. review of Digital Capabilities, Definitions of Learning Technology, Institutions constraints and I think this helps categorise the data, though I want to allow themes to emerge from the data where appropriate.

I’ve also been in touch with my fellow MA students and colleagues, as having moved away from the area, i’m feeling a bit lost and unmotivated to write. I got a link from one of them to Generating best evidence from qualitative research: the role of data analysis, which looks like a handy article on thematic analysis and i’ll try to summarise here.

Green (2007) maintain that transparency in the process of data analysis provides an insight into any conclusions drawn and this adds rigor to the research. Critical to this is the “process of examining the information collected and transforming it into a coherent account of what was found” and that “It is the task of the researcher to make the link between the accounts that are described and the claim to the knowledge produced”.

Thematic analysis is described here as a four-step process, underpinned by theory. It is a fluid process where the researcher is constantly moving between immersion in the data, coding, categorising and creation of themes.

The first step of immersion is to become familiar with the transcripts so the researcher is able to drop back in and out as necessary. It helps join up the data and provide clarity.

The second step of coding is not just about generating descriptive labels for the text, but about contextualising the data and identifying what the researcher is asking from it. This is an iterative process and data may be re-coded or assigned to multiple codes

The third step of categorisation is about finding a good fit between codes. Data may be categorised, but it will still need explaining and this may be through defining a theme (see Green’s example of ‘time’ category). Many studies stop at this stage and present a description of categories along with illustrative quotes, but it is important to acknowledge the full extent of the data that doesn’t fit within the category.

Which leads to the fourth step of identifying themes. This is moving beyond description to explanation and interpretation. This involves testing the explanation against the data and the theory. A theme must explore the significance of a category.

Green, J., Willis, K., Hughes, E., Small, R., Welch, N., Gibbs, L. and Daly, J. (2007), Generating best evidence from qualitative research: the role of data analysis. Australian and New Zealand Journal of Public Health, 31: 545–550. doi:10.1111/j.1753-6405.2007.00141.x