data analysis
Power of software to bring system to data-chaos
Saturday, April 4th, 2009 | learning, reflection, tools | No Comments
I am in a very interesting place in my research right now. A very frustrating one as well - but mostly wildly exciting. I’m starting to bring together my data and investigate what actually happened - or at least what actually happened according to me, since I subscribe to the view that analysis = story telling (to some extent at least).
So there I’ve been sitting for the last few weeks, attempting to find a loose end in the bundle of data, which I could grab hold of and start unraveling this mess. My problem was that I had too many ends which I was trying to hold in my hands simultaneously - half-formed understandings, hunches, ideas, frustrations, scraps of old feelings about things. Now that’s all good for story-telling, but how about analysis? In my view, the actual analysis part of analysis (the part that’s not story-telling) - is about bringing some system to the data, and making the story transparent enough to be credible.
After tip-toeing around the issue for weeks, I finally decided that maybe, after all, doing some coding wouldn’t be such a bad idea (yes, I know, the thought should have occurred to me much earlier, and it did, but I was frightened of disaggregating my lovely complex picture into little bits of meaningless junk - untill I realised that my complex picture, so far frighteningly resembled meaningless junk itself).
And yesterday I finally found a tool to do it with - on a Mac! I once did an interview with a Masters student for a project I worked on - and she talked about how using one particular tool - Excel instead of only Word - completely changed her outlook on research, data, and even herself. Suddenly she thought in tables instead of extended text and felt more ’scientific’ in her thinking. While we can probably all agree that Excel does not equal scientific approach to anything at all, I had the same feeling yesterday when I finally started coding. I got myself HyperResearch which does the job remarkably well and is easier to use that some of the large CAQDAS systems that run exclusively (!!!) on Windows. With very little fuss it puts at your fingertips the power of categorising text and retrieving the categories in various combinations. It may well lack some of the functionality of the larger packages - such as building large hierarchies of codes and modelling the data - but I am ready to question the usefulness of such ‘quantification’ tools for qualitative research any day. And althugh I’ve not explored the software fully yet and may well come ot miss a BIT more sophisticated functions (a bit of hierarchy is always good to put my head in order) - so far it’s managed to calm down a bit my search for a way to bring some system to the madness that is my understanding of my data.