Data Analysis

Now that I have collected my data, I need to analyse it – but how?

What is the best method for analysing the (mostly) qualitative data I have gathered.

Content Vs. Thematic Analysis

Looking at the descriptions in an Abubakar et al. (2018) journal, it seems that Content and Thematic Analysis are probably the most suited to the data I have collected.

These descriptions to me however, sound far too similar and I am struggling to understand how they differ in their approach. Both involve systematic coding to identify themes or patterns within the data. I need to delve a little deeper to understand which one is most suitable for my research.

A blog post I found on statisticsolutions.com states the following:

‘The analytical process for both thematic and content analysis is similar in that the researcher familiarizes herself with data and conducts coding on all data. In thematic analysis, the researcher systematically codes all data and then begins to organize the codes, based on some similarity, into larger and larger categories that may lead to a hierarchical structure of code -> subtheme -> theme. In content analysis, the researcher codes data and and generates categories and subcategories.’

(Statistic Solutions, 2020)

Reading through this I am struggling still to really see what the difference between the two are – in this case I feel like the word theme and category are interchangeable. The process of analysing is very very similar.

Perhaps a better approach is to understand what is being analysed and what you want to get out of the analysis. Content analysis can be used for quantitative data and qualitative data, thematic is only really used for qualitative and is more often used for analysing interview transcripts. While I do have some data that could be turned into quantitative data, the majority is qualitative and will remain as text.

Thematic analysis helps researchers understand those aspects of a phenomenon that participants talk about frequently or in depth, and the ways in which those aspects of a phenomenon may be connected.

(Statistic Solutions, 2020)

Whereas content analysis…

‘may help researchers with large amounts of textual data, as content analysis is useful for determining how words and word patterns are used in context.’

(Statistic Solutions, 2020)

Things like frequencies of words and pattern created in context are key to content analysis. For me that feels less relevant, I want to understand how the different aspects of what has been discussed in the focus groups relate to each other and see if there is an overarching ‘theme/s’ to help make sense of the data.

Its less about the individual and more about the collective opinion. Holistic I suppose is a good word for it.

Because of this I think Thematic Analysis maybe clinches it, and will be the approach I am going to take for analysing the data I have collected.

What is Coding?

Reading about different ways of analysis data, something frequently came up – coding. I will need to code my data, but I’m not really sure what that means? Coding to me is the stuff you do on computers to make websites – lots of </>.

It turns out that coding just means labeling part of the text/data to help organise it. And how you choose to code the data is entirely up to the individual.

I’m a very visual person so I am thinking – Sticky notes – highlighting – different colours? But I suppose I shall see what starts to make sense as I sift through the data.

Reflexive Thematic Analysis

I will be following Braun and Clarke’s six phases for Reflexive Thematic Analysis as a guide for my process.

The 6 steps are:

1. Familiarising yourself with the dataset.
2. Coding
3. Generating Initial Themes
4. Developing and Reviewing Themes
5. Refining, Defining and Naming Themes
6. Writing Up

They note that ‘with more experience (and smaller datasets), the analytic process can blur some of these phases together.’ My data set is quite small – two focus group transcripts so I anticipate that steps 3, 4 & 5 might become a bit blurred.

I will be approaching the reflexive TA in inductive, semantic and (critical) realist approaches, allowing the coding and themes to be directed by the content of the data.

1. Familiarising yourself with the dataset:

What have I done – I transcribed the audio recording into a transcript using AI. I started by using Otter AI and then realised they only gave you x3 30min transcripts and I had x2 1hr recordings, so instead I used Microsoft Word transcript which wasn’t as accurate. I started listening to the audio and correcting and amending the transcript. 30mins into the first one I realised I could make another Otter ai account and split my audio into sections. I transcribed the rest and carried on in the same way with the more accurate trancripts – listening to the audio and amending. I have spent a lot of time listening to this audio!

Because I was having to constantly rewind back and listen repeatedly due to people speaking over each other and the transcript getting confusing, this allowed me the familarise myself with the data quite a lot initially.

Step 2 – print out transcript and read through several times.

2. Coding:

I needed to decide on a method for coding my data. I decided going through and highlighting key points will be useful and using a post it to write what that points refers to succinctly.

At this point because I don’t know exactly what the themes are going to be (reading through it has generated some ideas but not anything concrete) I decided not to colour code anything.

3. Generating initial themes:

To figure out what the initial themes could be I decided to cut out the highlighted data and codes and begin grouping them together to see what felt addressed the same points. This also made it easier for me to shrink the dataset and make it more concise by removing all of the text that wasn’t needed for analysis.

You can progressively see how daylight disappeared during this process, and the warm yellow glow of my lightbulb takes over and makes my yellow highlighting and post-its all merge into one. Not the best for photographing!

4. Developing and reviewing themes:

The 12 initial themes I generated were:

1. Timing and Relevance
2. Thoughts on existing IT sessions
3. Lack of taught sessions and support
4. Need for frequent signposting to resources
5. Need for more check ins and prompts built into units for digital work.
6. Student involvement in development of resources
7. Issues and Barriers to creating these digital documents
8. Thoughts and suggestions for current CAD Club resources
9. New content suggestions
10. Students feelings and how individuals have differing opinions.
11. Interaction with CAD Club resources prior to focus group – quantitative.
12. Ideas to increase student engagement (Student suggestions)

How can I make sense of the themes – are they linked?

Some of the themes immediately feel like they fit together and I can see that there are distinct issues around ‘communication‘ and ‘teaching and support‘, some of the other themes like content suggestions and students feelings feel less relevant to these over-arching themes and don’t really feel like they help with answering my research question – which is about engagement. So although useful information – I don’t know how useful these themes are in contributing to answering my question.

I am going to digitise these themes and summarise the points for each to make it easier for me to process – as although the paper was useful it has started to get a bit messy and there is a lot of repetition and rogue themes that are throwing me off!

5. Refining, defining and naming themes:

I have already started to name my themes, and don’t know if I can refine them any further as there aren’t that many. It would be useful to finalise my over arching themes and what they mean in relation of my research question.

I have done this in a miro board:

6. Writing up:

How can I present this data within my presentation?

I need to work out how to summarise my findings in a concise and clear way.

To be continued….

References

• Abubakar, A. Douglas, S & Sani, Z (2018). Qualitative data collection, analysis and interpretation in research paradigms: The case of library and information science research. Applied Scientific Research. 6 (5): 211-215
• Braun, V. and Clarke, V. (2021). Thematic Analysis: A Practical Guide. London: Sage Publications.
• Statistics Solutions. (2020). What is the Difference between Content Analysis and Thematic Analysis? [online] Available at: https://www.statisticssolutions.com/what-is-the-difference-between-content-analysis-and-thematic-analysis/.