Coding and analysis tips for qualitative researchers

In this post, I share some of the tips I provide in my workshops on how to approach coding and analysis of qualitative data. Before I start, I want to put in the big disclaimer that most of this will depend on your research design. Having a clear idea of your research design before you begin your analysis is very helpful. If you don’t have much of an idea, these two books are a good place to start: Handling Qualitative Data by the amazing Lyn Richards. This is a link to some online resources associated with the book. The other one I recommend (and often mention) is The Coding Manual for Qualitative Researchers by Johnny Saldana.

So, what are some things to think about before you start coding?

Identify how you think and process information and start exploring the data that way

Do you process best by listening? If o, then focus on listening to the audio files rather than looking at a transcript and take notes while you do this. As you process what people are saying, you will start to understand and make sense of the data. You can then selectively transcribe the key points. Some qualitative software, or Computer-Assisted Qualitative Data Analysis Software (CAQDAS for short) allows you to do this within the tool itself. Some like NVivo even let you code directly from the audio or video files.

If you think better while looking at a hard copy rather than the screen, print off the transcript (if you can), get the highlighters and pens out and start that way. Later on, transfer these notes and codes into the software of choice. Why would you transfer to CAQDAS? Well, it would give you the opportunity to delve into the data in a more nuanced way. You could investigate the silences, as well as explore any new themes or relationship between ideas that seem to be emerging.

Create a code called “Great Quotes”

Create a category/code called “great quotes” and code all the great quotes to it. Also code the same text or piece of data to the other themes or categories to which it relates. For example, if you have a great quote that relates to power dynamics, then code that quote to power dynamics, and the great quotes code. When it is time to write up on power dynamics, you simply do a search for all content in great quotes that relate to power dynamics and you have all the useful quotes related to that topic right there.

Ask yourself: “Is this code worthy?”

It is extremely easy to create a lot of codes when using CAQDAS, and even to overcode, which means code absolutely everything. These can leave you disconnected from the data, meaning you may find it harder to understand what people are telling you, to find patterns, or even to make sense of the key points. There are two ways to look at this question.

Is this worthy of creating a new code or category? Ask yourself if what is being said is important enough to be created a separate category. If you are unsure, you can create a code called “other” and then check in on that at different times to see if there are any patterns or themes emerging.

Does this piece of text need to be coded at all? One of the temptations when coding is to code absolutely everything. Do stay focused on your topic and research question. More importantly, keep trying to understand and unpack what is important to the participant and code that. There is also a school of thought that says data reduction is one of the purposes of coding in the first place, though this of course depends on your research design.

Explore and experiment with different CAQDAS

I definitely suggest everyone has a play with different CAQDAS if they can. Software programs connect you to the data differently, display connections between ideas in unique ways, and have a variety of features help you analyse the data. Finding the best one will be determined by your research question, what you are wanting to find out from the data, as well as the way you think and process information. Some CAQDAS tools are very visual (such as Quirkos), others very easy to use and text based (DelveTool), some sit in between qual and quant, almost blurring the boundaries between the two (Dedoose), and others have a multitude of features useful for lots of types of projects, (For example: NVivo, MAXQDA, Atlast.ti). This site has a list of the different types of software and also provides reviews. Most software companies offer a free trial, so you can have a play around with the software to see if it is the right fit for you and your research.

Remember why you are analysing data

The most important thing to remember is the purpose of the analysis, namely, to answer your research question! Of course the purpose will differ slightly depending on the research design, but broadly, it is to understand what the participants are trying to tell you, and what is important to them. Stay focused on connecting with the data, finding the meaning behind what people are telling you, hearing their story, and making sense of it all. Things will start to fall into place.

Photo by Hugo Rocha on Unsplash