at a loss

[RuffRhyder]

I'm doing a PhD in Economics and have been thinking about this quite a bit. Not really for my own thesis, but still. If you try to do something in game theory, you quickly have to chose between realism and tractability. If you want to run regressions like TWP suggests, you need a boatload of data to be able to conclude with any reasonable certainty. I guess it's feasible if you manage to get a lot of 2+2-ers interested in your project and you guarantee anonymity etc. All the best of luck, though.