Using NVivo for literature reviews


Lyn Lavery Those of you who have attended any of my recent NVivo training will be aware that I’m a huge fan of using NVivo for literature reviews. I also love the fact that as soon as I mention it to other researchers, I can see their eyes light up with the possibilities that the software has to offer. It’s easy to understand why researchers are making the connection between NVivo and literature reviews. The processes involved are very similar to those involved in qualitative data analysis. In both, we read and reflect on text, make comments, identify key themes, look for great quotes, identify contradictions, and so on.

Using NVivo for literature reviews has several potential advantages. One key benefit is the ability to store everything related to your review in one place. I’m the type of researcher that if I let myself get away with it, I’d have information everywhere - my desk would be a mess of journal articles printouts, handwritten notes and post-its, while my computer would have electronic files in various random locations. With NVivo, I can have the PDFs of my journal articles, notes I’ve taken from books, and reflections on my readings, all stored in one place. Even better, these can all be coded at nodes in NVivo that represent the various topics of interest for my review. I can also store various attributes about my references such as year of publication, methodology used, and country of origin.

These attributes bring me to the second advantage of using NVivo for a literature review. NVivo has a powerful query tool which allows you to examine your coding across attributes. Let’s say you have a review on the literature relating to transcription, and you’re interested in the way authors discuss the importance of software and technology across different disciplines. If you code at a node for ‘Software and technology’ and include an attribute for author discipline, you can run a Matrix Coding Query in NVivo to examine this. Likewise, you may be interested in how authors discuss a particular topic over time – Matrix Coding Queries are an easy way to assess this.

The query tool also allows you to query the text itself. Text Search Queries allow you to look for particular words or phrases in the literature, and the results of these can be easily converted to nodes. Hence, it’s a very fast way to locate key topics and have the associated text organised in one place for later reference. Likewise, the Word Frequency Query allows you to examine the most frequently occurring words across the literature, which is an excellent way to identify key topics in the beginning stages of a review. Both Text Search and Word Frequency Queries can also be displayed visually, which is an excellent technique for exploring patterns in the literature.

When you get to the write up stage, that’s when it starts to get really exciting. Using our transcription example above, let’s say you are interested in the impact of difficult recordings on the person transcribing. If you have coded your literature at a node for this, all you need to do it open the node and you have all the related text in one place to assist with your write-up. The Matrix Coding Queries mentioned above can also be very useful at this point.

Some of my other favourite NVivo literature review tips include:

  • Have a node where you code any ‘Great Quotes’ that you want to include in your write-up.
  • Use Annotations for brief reflections and See Also Links for making connections between authors.
  • Information from reference management software (EndNote, Zotero, Mendeley, and RefWorks) or online note taking software (Evernote and OneNote) can be imported into your literature review project.
  • Framework Matrices are an excellent way to compare across authors or summarise findings from key articles.

If you’re interested in learning more, we offer a two-hour online course on using NVivo for Literature Reviews. For further information about NVivo, visit QSR International.