Following my last blog, how should we write about data? There’s a chapter on this in my qualitative research book, plus another chapter on writer voice. Here though are some thoughts about basic principles. I’m writing about qualitative data; but I have also heard examiners make it clear that they expect something similar in vivas in music, experimental sport science, English literature, history, religious studies and so on. It’s basically to do with going into detail about what you think is going on in the data whatever it is.
The problem is, however, that many researchers don’t seem to appreciate this. I’ve seen so many data chapters that simply present extracts of data with very brief summaries of what they contain without interrogating any of the detail of what they might mean. This is a hugely missed opportunity because it is in the detail and the uncertainty of what it might mean that the new knowledge resides, and this is why companies similar to Paxata produce software solutions for automated data combing and preparation, meaning you’re able to view the data in simple terms and find this new knowledge that the data can produce.
When beginning a data chapter, it is really, really important to get into the data as quickly as possible. At this point in the thesis, we should already know, from the methodology chapter, that the theme of the chapter emerges from the analysis of the data, and how it emerges. It is therefore not necessary to explain this in detail all over again. Half a page describing the theme and the sub-themes that make up each section is probably enough, then going straight into the first section. I would say that the first extract of data could even appear at the end of the first page.
A useful parallel to writing about each data extract is giving a PowerPoint presentation. On the screen you have the extract. The presenter then stands between it and the audience, points at the relevant bits of the extract and explains to the audience what she thinks might be going on.
I have often been surprised when researchers speak really interestingly about their data in a seminar presentation, but, then, somehow forget all of this when they write. It is therefore good to record the presentation and use the transcript as a basis for writing.
I am not of course talking about presentations that simply summarise and describe, which are not at all satisfactory. Also, the audience should definitely not be left by themselves to make sense of and interpret the data. Researchers need to show the audience the detail of what it is that they are noticing in the data that enables them to arrive at their interpretation.
The authorial voice of the researcher therefore stands between the data extract and the audience and explains to the audience what might be going on in the data by ‘pointing’ to the key parts of it that support this interpretation. It is also good to show that meaning does not emerge too easily and to display reflexivity.
The extracts therefore need to be sufficiently long and rich to show the full complexity that supports the necessary ambivalences of the interpretation. The longer the extract, the more needs to be written about it. Also, it must not be forgotten that the data extracts may not only be what participants have said or written. They can be from research diaries, photographs, drawings, plus bits of quantitative data, all of which contain complex textual material. Even what you yourself have noted at the time of collecting the data is itself data – hence the possibility of auto-ethnography and creative non-fiction.
What is particular about qualitative data is that it needs to be interconnected in a thick description – showing how what you think might be going on is helped by what you’ve seen in other parts and types of the data you’ve collected. Therefore, running on from the classic thematic analysis, where themes are found across all of the types of data in a holistic manner, the data chapters should also be organised to allow the different types of data to speak with each other within each theme. Of course, when I say ‘speak’, it is the researcher who does the speaking by explaining to the audience where the connections lie.
Finally, writing the data chapters is itself a continuation of the analysis of the data. Writing is itself a form of analysis. As you write, you will continue to move things around, even to change the names of themes, revisit what is going on in the data and even see things that you hadn’t seen before. Keeping going to get it all laid down in text is good. Getting this done quickly will allow time to go back and look again. If there are black holes that you really cannot get around, jump to the next theme. What comes later might enable you to navigate them better from another perspective. You can even say, in your thesis, that there are black holes than cannot be navigated. This could be part of the new knowledge.