NVivo coding tips and tricks
We’ve used NVivo extensively for coding over the years and have discovered a number of tips and tricks that we’d like to share. We’re not suggesting that these are the “right” way to code (we don’t believe there is such a thing), but we’ve certainly found that we’ve saved ourselves some time, not to mention headaches, by following the suggestions below! Note that this blog post refers to 'nodes' – if you're using a recent version of NVivo, these have been renamed 'codes'.
Create descriptions for all your nodes
Take the time to enter descriptions when you create nodes. This can be beneficial both to the coding process, as well as your overall data analysis. Nodes often represent abstract or complex concepts, so it’s easy to forget what you intended, and meanings can evolve over time as your analysis progresses. For those of you working in teams, descriptions are particularly important to ensure that everyone has a shared understanding of the nodes. Descriptions can be entered as you create a node within the ‘New Node’ dialog box, or you can insert and amend them at a later stage.
Keep an eye on the number of nodes in your project
One of the problems with NVivo (admittedly a positive one) is that it is very easy to create nodes and perform coding. The downside of this is that it is very easy to get carried away! Having a large number of very specific nodes can slow you down and negatively impact on the overall quality of your data analysis.
One of the most common questions we’re asked is “How many nodes should I have in my NVivo project?” Our usual response is that there is no magic number – this will be driven by your methodology and research questions, the type of analysis you are looking to perform, as well as the nature of your data. Keep in mind though that ‘less is more’!
Be selective about how much data you code
Another area where the ‘less is more’ principle applies is in relation to the amount of text to select when coding. Many researchers intuitively want to include surrounding contextual data. Unfortunately, this can create problems later on as you end up with lots of extra reading around the primary content. It also makes it difficult to discern patterns in the data. We recommend that you code only the data that specifically relates to the node that you are coding to. And, remember that NVivo allows you to easily view contextual information if needed. Simply right-click over coded text and select either Open Referenced File, Coding Context, or Spread Coding.
Keep track of your coding
I know that I like to keep track of my coding as I go, just to make sure the information has been coded to the right node. There are a few ways to do this - my preference is to use the ‘Coding Density Bar’ (you can select this from 'Coding Stripes'). This appears as a stripe down the right-hand side of your document, and if you ‘hover’ your mouse over the stripe it will list all the nodes that relate to that section of data. As this is updated as you code, you’re able to keep an eye on which node(s) the data has been coded to.
Allow sufficient time for coding
Lastly, don’t under-estimate how long the coding process will take. Even with the benefit of a tool such as NVivo, qualitative data analysis can be a complex process and you need sufficient time to do it justice. When planning your research, we recommend that you leave yourself more time than you think you will need for this stage of the process. It’s also important that you avoid coding for too long in one sitting, and we always like to take a break between coding each file. These simple steps will help maintain a high standard of coding for your research project.
There’s only room to share some of our coding advice in this blog, so if you’d like to extend your learning in this area, consider attending our upcoming virtual event Research Accelerator 2021. We’ve got a lot more tips to share!