Wednesday, 10 March 2010

Drowning in data

I've been back in the UK for five weeks exactly, and it's already feeling like a lifetime!

I have to admit I've not really been concentrating on work for most of this past month and a bit. Apart from the expected post-fieldwork brain-scatter, I've also been grappling with a ridiculously complicated moving situation. I won't get into it now, let's just say that it won't actually be fully sorted out for another 3 weeks or so. This whole thing has resulted in a monumental decline in efficiency, productivity, health, and mental wellbeing. I now find myself in a period of relative calm and stability, so I've been able to actually make some progress (and post this blog entry).

So, as you may recall, I am now in a situation where my plans have been changed completely, and I am facing a long stint of data analysis. It is the view of my supervisors that the data I have collected so far will be enough for my thesis. I would actually love to have another field season, but it might not be the best course of action, given the amount of time I have left as a PhD student.

If I use the funding I have budgeted for another field season to employ someone to digitise and enter data for me, then I would save a couple of months of mindless tedium. If I actually go to the field, not only do I lose 3-4 months of valuable data analysis time, I actually increase the amount of data I'd have to enter. In effect, I'd lose about 6-7 months of analysis and writing-up time. So the first scenario makes much more sense.

All this is yet to be finalised and agreed on. In fact, I have my annual review meeting with my supervisors in a couple of days, so I am hoping to have a new gameplan for the PhD, so I can start fresh on it next week.

I leave you with a pretty picture I made in my GIS - see, I've been doing some work! It shows the relative proportions of the main different habitat/land use types at my survey sites across Cyprus. This sort of map should hopefully help me deal with any potential spatial autocorrelation issues. More on that some other time.