One of the most often used approaches in qualitative social sciences is Grounded Theory. Created by Glaser and Strauss, Grounded Theory revolutionised the qualitative approach to research. By creating a structured framework and clear steps for progression of research, qualitative inquiry could be seen as reliable, valid and “scientific”. I used quotation mark there because I’m not sure that there is one specific definition or a distinct and clear approach that makes something “scientific”. But Grounded Theory allowed qualitative methods to be taken more seriously within the social sciences.
The main reason for this approach being so quickly adopted on such a wide scale was that it has high levels of transparency. Often qualitative research can seem like a mysterious thing. There are no hard rules as to how much data is needed or how it should be collected or analysed. This makes it difficult to learn about and understand unless you have experience of doing qualitative research. Grounded Theory changed this by making it clear how research progresses, from the researcher’s expectations and prior knowledge of the field to sampling and data collection and theory creation. However, Grounded Theory is not a perfect system. There are flaws, particularly with the expectation that researchers should know as little as possible about the field so their analysis is not coloured by personal judgements or pre-existing knowledge. But one of the main problems students face when learning Grounded Theory is understanding how the analysis works. Whilst there is a large amount of transparency within grounded Theory literature, the emergence of themes remains a grey area. When I was first learning Grounded Theory we were told that transcripts must be gone over with a fine tooth comb, word for word, coded appropriately again and again until themes emerge. Always using that phrasing: a theme will emerge.
This is not particularly clear for many students, what are these themes, how do they emerge, how will we know when they emerge? For me and many of the people on my course, it didn’t make much sense. It didn’t sound “scientific”, it sounded like waving your hands over interview transcripts as if they were a crystal ball and magically a theme would appear from the ether. However, when I came to do a piece of Grounded Theory research that did not happen. I coded my transcripts word for word, went over them multiple times, grouped my codes, removed some that were unnecessary, cried, yelled at my transcripts. Then my Mum (in her infinite wisdom) told me to stop and gave me a pot of ice cream. If you live with someone doing research and you walk in to find them yelling “Be a theme! Why aren’t you a theme! Emerge Damnit!” at stacks of paper, this is the appropriate response.
Eventually themes did emerge. But it was only after having done the analysis that I realised the themes did not magically emerge or appear all at once, it had been a process. The hundreds of initial codes you begin with become streamlined as you go over the data again and again. Some will be removed as they only appear once or twice, others will be merged into a larger group of codes that have similar meaning. As this process continues, you will likely end up with five or six large codes that run through the whole of the data set. As these codes are a combination of smaller codes, they are quite complex and take some explaining, but it is clear why they are relevant to the data, and together these five or six large codes say something about the data as a whole, something that is relevant to the research question. These are your themes.
To my mind, themes do not emerge. They’re the result of a process of filtration. You begin with masses of data, hours of interviews and discussions and reading. Then you begin to condense this down by picking out key words or codes. Then you begin to group similar codes together, again narrowing the field. This process continues until you are left with the important bits. The things that are present in all of the data, the overarching things that say something about the data, the key themes. These themes are the essence of your data, they pure bits. The interesting bits that you can use to make a theory that will satisfy your research question. So a theme does not emerge, you have to discover it.