Tuesday, January 22, 2008
Discrepencies in data analysis
I feel like my brain is going to start exploding soon as I have been doing so much reading and trying to sort out what's going on with my data. I have good data, I know that as I did all the experiments myself! My problem is that my data doesn't fit nicely to any typical models, which is OK and there are many likely reasons for this, but just means lots of work to explain and justify why this is and that it's not that I have a rubbish or pointless experimental system.
One of my main problems is that there are lots of ways to plot out the data, and mostly that's fine and standard undergraduate stuff. What is really getting to me is the fact that one particular way of presenting the data is particularly bad yet half the text books I have been going through plot the majority of examples using this representation (after first pointing out its serious inadequacies in earlier chapters) and all the background papers I have been reading relating to my experiment also use this particular projection. Gah! I guess it's most annoying for me as my data definitely shows its deviations from typical using this particular representation, and it makes it really hard to evaluate other peoples research when shown this way.
Another problem I am finding as I try to really analyse my data is just how complicated my system really is. Individually each part is quite straightforward and well characterised and understood. But put all together it actually tends to brain melting complexity. One of the things I have a lot of trouble with is adequately explaining by research in detail to other people. I can explain stuff and justify it at quite a simple level, and it all works out and makes sense in my head at more detailed and complicated levels, but actually getting that across to people in my own words I find is a nightmare. Still practise practise practise and one day I might be perfect!
One of my main problems is that there are lots of ways to plot out the data, and mostly that's fine and standard undergraduate stuff. What is really getting to me is the fact that one particular way of presenting the data is particularly bad yet half the text books I have been going through plot the majority of examples using this representation (after first pointing out its serious inadequacies in earlier chapters) and all the background papers I have been reading relating to my experiment also use this particular projection. Gah! I guess it's most annoying for me as my data definitely shows its deviations from typical using this particular representation, and it makes it really hard to evaluate other peoples research when shown this way.
Another problem I am finding as I try to really analyse my data is just how complicated my system really is. Individually each part is quite straightforward and well characterised and understood. But put all together it actually tends to brain melting complexity. One of the things I have a lot of trouble with is adequately explaining by research in detail to other people. I can explain stuff and justify it at quite a simple level, and it all works out and makes sense in my head at more detailed and complicated levels, but actually getting that across to people in my own words I find is a nightmare. Still practise practise practise and one day I might be perfect!
Labels: data analysis
Subscribe to Posts [Atom]