Exclusive Interview with Ann Lewins and Christina Silver
In our last issue's article about the CAQDAS 07 conference, we mentioned an inspiring presentation given by Ann Lewins and Christina Silver on their new book, Using Software in Qualitative Research: A Step-by-Step Guide.
Already labeled "The New QDA Bible" because of its exhaustive, yet user-friendly, treatment of the subject, Using Software in Qualitative Research: A Step-by-Step Guide weighs each major QDA software package evenly and offers crucial instruction and support, no matter which product you use. Each chapter offers rigorous comparisons of the major programs. In short: a great book for everyone doing qualitative research.
But who are the people behind Using Software in Qualitative Research? What inspired them to take on the formidable task of writing it? What changes do they foresee in the world of QDA?
We sat down with Lewins and Silver for a chat on these topics and more.
Tell us a little about your research background and your fascination with qualitative data analysis.
Ann Lewins: I've had a patchy research career. I only got my first degree at 43 – after my husband died – and then went straight into a PhD, which I eventually interrupted because my paid work was too demanding and interesting. My part-time PhD research was all about the change in direction of the British Labor Party towards Europe during the 1980s and I found it very difficult to remain at one point in history because politically, things were changing so fast.
At the same time I was fortunate to be in at the beginning of the CAQDAS Networking Project – though Nigel Fielding, Ray Lee and Jane Fielding had been holding informal events before then to support CAQDAS. So, unlike the PhD, my paid work began to be all about new developments, new methods, and new debates! I also took part in the analysis of several qualitative datasets collected by East Surrey Health Authority and later with other local authorities, each project concerning the needs of and provision for older people. For the former, it was the Research Manager who requested that I use ATLAS.ti. So the report was delivered along with a live ATLAS.ti presentation.
Christina Silver: I became interested in social research at school and subsequently during my first degree in Sociology and Politics. I first undertook qualitative data analysis when I was doing my undergraduate dissertation, a process which I greatly enjoyed. I constructed a paper-based matrix of respondents by codes on my bedroom wall and stuck photocopied quotes in the relevant cells. At the time I didn't know CAQDAS packages were available.
I subsequently used software for both my Masters and PhD dissertations, which compared the experience of school-based sex education in the Netherlands and England and Wales from a socio-political and historical perspective. I have also undertaken many different research projects, mostly qualitative, for both academic and applied purposes and usually using software, and I have taught methods courses at several universities in the UK and Europe.
What made you decide that the researchers out there need a new guide that helps them with computer-aided qualitative research?
CS: The book is our attempt to pull together our own experiences in teaching and using qualitative software, and the knowledge we have gained from talking to and working with so many different researchers and kinds of projects over the years.
What was your overall goal of writing this book?
AL & CS: We needed to provide practical ideas about what was useful in any software, what was exceptional in each of them and how real users adapted software to suit their own needs.
Can you tell our readers what kinds of research are best supported by the ATLAS.ti software?
AL: I would keep it very general and link it to types of data rather than types of research. One general category would be any research that generates observation notes, in depth interviews or focus group data.
Focus group data for instance is notoriously difficult to code. How much text to code concerning, say, a topic within a topic: one speaker's contribution? Part of a contribution? The whole discussion? For me, one of the central attractions of ATLAS.ti is the way when you are flicking through coded passages (having double clicked on a code) you see the passages highlighted in their source context as you single click on each quote in the list.
Other softwares do this in different ways and quite well, but there is just something about the layout of ATLAS.ti which puts you in close touch with more of your data at moments like that. Context is so important. So what I mean is that if you have by chance only coded a small nugget within a sentence, you can always get back to the context surrounding the nugget very easily. Also, flexible ways of working and the level of interactivity between functions in ATLAS.ti are second to none.
For me, interactivity is a very important part of the instantaneous, reflective nature of the way you need to work when unscrambling complex thoughts and connections.
CS: ATLAS.ti will support most types of qualitative research very well. In comparison to other options available, a particular benefit is the range of data types which can be handled together and in the same way with ATLAS.ti. There are an increasing number of projects which require multimedia and textual data to be integrated thoroughly. Hyperlinking between points in data is dealt with best in ATLAS.ti. I don't think you can truly integrate different media without hyperlinking tools such as those in ATLAS.ti.
At ATLAS.ti we are curious about your opinion: what do you two think are the strongest features of our software?
AL: I think I've already touched on some basic – but major – aspects. I recognize that the things we users think of as simple or basic are probably the most difficult to write for a software designer!
The interactivity between functions is very efficient and effortless. I like the layout, and the movement between different functions and the way the data is always given utmost importance in the amount of screen it occupies. That's a major lack with at least one of the other software programs – actually it is quite stressful and annoying continually having to adjust the panes to get more data on view. So although there are some amazing other features, it's the layout of ATLAS.ti that I like best.
CS: I agree with what Ann says about the interactivity between tools and the interface. In addition I particularly like the way the memo system is given the same prominence as documents, quotes and codes. People often tend to forget to write notes etc. when they are in the flow of coding or analysis and the centrality of the memo system in ATLAS.ti both stresses the importance of writing in QDA and also serves as a reminder to the user to do so!
Do you think QDA software changed the way qualitative research is done and perceived? If so, can you elaborate a little?
CS: Software enables many qualitative tasks which are difficult or not possible when working manually, and therefore the potential for conducting analysis in different ways is great when using software.
For example, word searches, frequency counts, reliable and fast retrieval and the query tools, not to mention the ease with which one can change one's mind and move beyond the basic level of coding etc. etc.
However, many people use software in quite similar ways to how they would work manually, often not utilizing the more sophisticated tools available. There are also still many skeptics, and others who are unaware of software or misunderstand its purpose and possibilities. So I would say that there is a relatively small minority of methodologists and researchers who are stretching the boundaries and who see great potential in developing the ways we can work with qualitative data due to the potential that software affords.
You can see this in particular with mixed methods and multi-media work. But it's also important to be aware that the majority of researchers are probably using software in quite basic ways, and to acknowledge that this is okay as well. So it can change the way QDA is done and perceived, but it needn't.
In your introduction you mention the debate over whether different QDA packages "steer" the way we perform analysis. In your opinions, can one's choice of software affect research results? If so, how?
CS: I think as researchers we need to be reflexive about the impact of software use on QDA and acknowledge explicitly which tools were used to what effect in our methodological writings.
For too long QDA has been seen as a mystical process and the use of software both allows and necessitates transparency in analytic processes. That said, it is the researcher who is conducting the analysis and the one deciding which software to use, and which tools within that software to use in which order etc. It can be argued that the structure of software programs encourages certain types of working and thinking – for example, that visibly hierarchical coding schema structures encourage hierarchical thinking etc.
But I think this is a somewhat simplistic argument which negates the independence and reflexivity of the individual researcher. Just because a tool is available does not mean that the researcher will use it; just because a software developer had something in mind when designing a tool does not mean the user will interpret it in that way. We know very well from our work with the CNP that researchers use software very differently, and they often manipulate tools very creatively to suit their own particular needs.
On the other hand, sometimes people contact us and say, "I don't think I'm using ATLAS.ti (or another QDA software) properly, or to its full potential…" and ask for guidance on how they "should" be using a certain program. They seem to need reassurance that they are doing their analysis "right."
But the software is irrelevant in answering that question. Just because you use software does not make the quality of your work of a higher standard, and if you do not use software, it doesn't mean your analysis is necessarily less thorough or useful. The quality of the analysis essentially rests with the person using the tool, not the tool itself.
What's the most interesting research project you've heard of through your help-line?
CS: Difficult to say as there are so many! I always find it interesting to work with people from the less traditional disciplines. One particular project I remember was using ATLAS.ti for analyzing hieroglyphics in Egyptology.
AL: The other day I held an advanced follow-up session with several ATLAS.ti users. It was such a fascinating mix of researchers. It was just so interesting hearing how each researcher had evolved different ways of working and how they were bringing different aspects of qualitative methodology to the software. I think I managed to get over some new ATLAS.ti techniques and tricks…but I think I was the main beneficiary!
What major changes do you foresee in the QDA world?
CS: What we'd like to see is more collaboration between developers and users, as these dialogues can be very interesting. It will be interesting to see what happens about data exchange, as this is very much in the interests of large, multi-disciplinary and multi-national research teams. I think we may be heading towards web-based interfaces and more add-on type functionality.
What kind of feedback do you get from the readers of your book?
AL: It's a bit early to say, but people seem to like it.
CS: People find the comparisons at the end of each chapter useful in thinking about which package will best suit their needs. We've also heard that methods teachers like the step-by-step sections, as they are quite detailed and useful to direct students to.
AL: It is very important to us that we remain independent of any commercial links or a perceived dependence on any one software, so we look for the upsides, but were also quite frank in our book about the drawbacks of each software.
Thank you so much for taking the time to talk to us, Ann and Christina!
Now all that remains to be said is: check out the book!
