How an NVivo expert puts strategies before tactics
Christina Silver manages the CAQDAS Networking Project at the University of Surrey and is a co-author of Using Software in Qualitative Research (Lewins and Silver). She is also an authorised trainer of NVivo software. Here Christina explains the importance of letting strategies (not tactics) guide the use of NVivo.
For the past few years Nick Woolf and I have been unpacking the way that expert users of CAQDAS packages like NVivo harness these programs powerfully, so we can better help new researchers learn these skills without years of trial and error.
The result is Five-Level QDA, a pedagogy that spans methodologies, software programs and teaching modes (Silver & Woolf, 2015) which we presented in the Digital Tools stream at the 12th International Congress of Qualitative Inquiry and which we will be discussing at the forthcoming Kwalon conference which will reflect on the future of QDA software, in Rotterdam, the Netherlands, 25-26 August.
Here I’m sharing some of the reasons why Five-Level QDA workshops delay teaching how to operate the software until part way through, something that is both uncommon and unexpected.
What is Five-Level QDA?
First some key points about Five-Level QDA:
- Five-Level QDA is not a different way of doing analysis, but a method of teaching CAQDAS packages, including NVivo, that describes what expert users already do unconsciously.
- It describes the process of harnessing the software powerfully, and so is relevant whatever sector or discipline you’re working in, whatever methodology you’re using to guide your project, and whether you’re working as an individual researcher or in a team.
- The method focuses on thinking differently about strategies - what you plan to do - and tactics - how you plan to do it. Analytic strategies must always drive software tactics, never the other way round. This doesn’t always come naturally, and the key skill to
learn in using NVivo powerfully is how to “translate” individual analytic tasks into the most appropriate software operations.
- Learners must have a complete and high-level understanding of the software’s architecture before being able to learn to implement “translation”. In contrast, learning to operate the software is the easiest and least important part of learning, and comes last.
First, be clear about analytic strategies
Many workshop participants expect to open NVivo straight away and be taught how to operate it, but this is not the way we teach.
In two day workshops we don’t open NVivo until after lunch on the first day. In three day workshops we look at NVivo on day one, but don’t do anything with it until day two. Some might think this goes against the idea that people learn best by doing.
We agree that experiential learning is important, as is the need to make learning contextually relevant. But harnessing NVivo powerfully requires being clear about analytic strategies and the architecture of the program first, as a necessary preparation for the all-important process of “translation”.
Make the best use of NVivo
Operating NVivo is easy. I can teach it in 10 minutes. But learning to operate NVivo doesn’t lead to powerful use.
A workshop that opens with learning to operate the software inevitably creates a mindset that the progress of the analysis is driven by the extensive capabilities of the software.
Researchers then look around their project for ways to use all of the software’s features – which means letting the tactics rather than the strategies drive the analysis, instead of the other way around.
NVivo can do pretty much whatever you need it to do, that’s what makes it so potentially powerful. But it’s tricky for new users to know what it is they need to learn first. Five-Level QDA gets over this by providing a process for identifying individual analytic tasks – the strategies – and teaching how to “translate” them into software operations.
We do this by focusing on NVivo’s lower-level “components” (the things within it that can be acted on, of which there are 16) rather than its higher-level features (the things NVivo can do – of which there are hundreds). NVivo users must obviously learn to operate the software, but to use the program powerfully that is not the first item of learning on the agenda.
Find out more
We’re currently working on writing up the Five-Level QDA method in a textbook, in the meantime, you can read more about it in our recent article.
During the past 20 years Christina has trained and supported almost 10,000 researchers in the use of qualitative software and has contributed to graduate level methods courses in several universities.