Documentation:Research Commons/NVivo Symposium 2015

Keynote Talk: Analysis Beyond Coding

Two decades ago, Steiner Kvale wrote of the “1000 pages of interview transcripts” problem in qualitative research. I have regularly encountered the follow on problem: “Now I’ve coded all my data, what do I do?” – one that has become evident since the advent of qualitative software. Qualitative software has been criticised for fostering an emphasis on code and retrieve as a primary tool for analysis of qualitative data, and indeed, the majority of users do not go beyond these functions. The consequence is a plethora of studies that rely on descriptions of key codes (a.k.a. themes) as a report of completed ‘analysis’. But is software really to blame?

Both of these ‘problems’ are at least partially consequences of researchers ignoring a fundamental tenet of almost all approaches to qualitative analysis: the need to begin analysing as soon as data are available (or even earlier, treating literature as data). Early analysis, involving an holistic review of and reflection on each source as it comes in, and for some traditions a detailed working through, is likely to prompt a modification of questioning and perhaps sampling strategy so as to gain more relevant data. Further, it will ensure the researcher is already developing ideas to check against further data, contributing to answering research questions.

Rather than being to blame for these problems of post-transcript and post-coding blockage, software, especially modern versions of specialist qualitative data analysis software, provides elegant (and useful!) solutions for deeper analysis. These include the use of memoing, links back to evidence, tools for querying relationships between codes, and modelling options. Be prepared, therefore, to experiment, to ‘play’, to take time to understand how the software ‘works’, and thus, to learn to use it flexibly and effectively for whatever methodological purposes you choose.

So, our researcher with the problem of all those codes, faced with a daunting array of choices, a learning curve and tight deadlines, uncertain about what steps to take, is tempted to resort to the tried and true method of ‘thematic analysis’ – except that what he or she describes are not themes (these typically involve a relational statement with two nouns and a verb), but a set of unrelated categories. In this context, I developed a non-threatening, easily remembered, three-step ‘formula’ to help move our beleaguered researcher forward: Take your set of codes, and Describe, Compare, Relate. Application of these three sequential but also iterative steps, each of which builds on codes and attributes and is fully supported by software, leads the qualitative analyst out of the code-and-retrieve mire. And it does so gently, by starting with description of a single code. Then, when comparisons involving that code are made across subgroups, questions to pursue arise – to be further explored through assessing relationships with other codes. And so to another cycle with another code, until a web of relationships is built, and theory is happening.

Software will not always be the answer to every need. There can be times when you just want to map out some ideas using the flexibility of paper, white board or sticky notes. And software is not sufficient in itself: ultimately good research requires a researcher who is prepared to take time with their data, to reflect deeply, to constantly challenge their data and what they are saying, to be persistent in pursuing leads, and who mixes intuition with discernment, good management, and logic.

Lightning Talks

Lightning Talks are rapid-fire, high interest presentations. As part of our all-day NVivo Symposium, we will be showcasing various uses of NVivo qualitative data analysis software through this Lightning Talks event. We welcome submissions from graduate students and other researchers about how they have been using NVivo to further their work -- you're welcome to discuss aspects of methodology, showcase your visualizations, or highlight interesting uses of the software. Please consider submitting a proposal. Accepted presenters will be guaranteed a seat in the post-symposium advanced workshop on Queries and Visualizations on April 24th.

Master Class: Theory and Practice of Qualitative Data Analysis in NVivo

In this interactive seminar, Pat Bazeley will work with actual research projects to discuss the theory and practice of qualitative data analysis using NVivo. Potential challenges and opportunities in these projects will be addressed. This is a great chance to learn from an expert NVivo user in a very practical way.

NVivo for Mac Workshop

In this half-day workshop, a trainer from QSR, the company that makes NVivo, will cover the fundamentals of NVivo for Mac. Topics addressed may include an overview of NVivo for Mac; the design framework; importing data; working with variable type data, such as demographics; starting the analysis process; setting up your own project in NVivo. This workshop is suitable for Mac users with little or not experience working with NVivo. Space is limited ; registration by lottery is required.

About the Keynote Speaker

Pat Bazeley is Director of Research Support P/L and part-time Associate Professor in the Centre for Primary Health Care and Equity at the University of New South Wales. Since graduating in psychology she has worked in community development, project consulting and in academic research development. For over 20 years Pat has been providing research training and project consulting across a wide range of disciplines to academics, graduate students and practitioners across Australia and internationally. Her particular expertise is in helping researchers to make sense of qualitative, survey, and mixed methods data and the use of computer programs for management and analysis of data. Pat’s research and publications, which include Qualitative data analysis with NVivo, focus on qualitative and mixed methods data analysis and on the development and performance of researchers. She serves on the Editorial Boards of the Journal of Mixed Methods Research and Qualitative Health Research, and is President Elect of the Mixed Methods International Research Association.