To code or not to code?
"Any researcher who wishes to become proficient at doing qualitative analysis must learn to code well and easily.The excellence of the research rests in large part on the excellence of the coding." (Strauss, 1987, p. 27)
That's right, it's not so much about the coding (you can always justify that), but HOW you code (you also need to justify that!). As you probably know, Qualitative Research is about PROCESS. How you go about conducting the analysis should provide information on your own position towards the data, and also exemplify how you incorporate the conceptual framework into the analysis.
Essentially, it's all about perspective and acknowledging that perspective is where you can begin to see how much it drives what your findings will be. Let me give you an example.
A friend of mine in Research Psychology was recently describing to me a project that her professor worked on, in which they distinguished between a 'hit' and a 'toke' (yes, this was a study about marijuana consumption), and she asked me what my take was. I immediately launched into the different approaches one might take to smoking weed, either with a slow long inhale, or short breathy intakes. Each of these modes of smoking would depend on the smokers personality, what they believe about cannabis smoking etiquette, how they feel at the moment of smoking, etc. I was interested in the processes involved in smoking and the reasoning behind people’s choices/actions.
On the other hand, my friend's interpretation was to categorise a ‘hit’ and a ‘toke’ according to differences and try to define them through an observation of the behaviours (e.g., one long inhale is a ‘toke’ and three short inhales is a ‘hit’). Once the terms were defined, she began thinking about to quantify the terms in relation to each other (e.g, Is 3 tokes equal to 1 hit?), or within each smoking session (e.g., How many hits are needed in one session? How many tokes?)
We weren't talking about qualitative or quantitative research, but it struck me how quickly our ontological views on data and analysis surfaced, even within a casual discussion. It's easy when you put your own approach in contrast to others, but we don't often get to do this within our projects, so perhaps this is what makes positioning yourself so difficult. And yet, when this crucial step is missing in the research, it can prevent the readership from understanding where the researcher is coming from and how they came to their conclusions.
Lately, I've been talking to colleagues about coding in qualitative research and process can be very complex and intricate. This is one of the areas I really appreciate in quality qualitative research, where researchers can report their coding explicitly and build it into their analytical framework.
To be sure, the qualitative coding process in language education research has been largely influenced by Grounded Theory. This is where we get coding processes such as open coding and in vivo coding which enable the researcher to look through the data having consciously put their personal biases aside.
To move beyond descriptive coding into analytical coding, researchers can use different processes, such as axial coding, selective coding and comparative coding.
They can also incorporate the timing of the coding into the research methodology. For instance, will your initial coding be open coding? When do you start comparative coding? Why?
The key question to ask yourself after each step is WHY. I recommend writing down the questions and the answers you come up with in a .doc or a table to keep track of the whole coding process and how it evolves over the course of the study. Some questions I like to ask to keep myself in check are:
What is your overall plan for handling the data? Why did you chose this way of doing it? Does this tie back to your conceptual framework in any way?
How are you going to operationalise the coding, that is, what is the step by step process?
Will you use open coding first, then move on to axial coding? Why?
Will you do comparative coding throughout the entirety of the coding process? Why?
How are you planning to move from descriptive coding to analytical coding?
My belief is that JUSTIFICATION is one of the most important features in building up validity in qualitative research. If you can explain WHY this seemed like the best process/approach at the time, then you're at least on some kind of path. There's nothing to say your process can't be critique, improved, reflected on and tweaked, but at that moment with that data, that seemed like the best fit.
So bring out your inner 3-year and become a top-notch researcher: never stop asking WHY? WHY? WHY?