Friday, May 29, 2009

experiments + surveys

I'm working on an interesting combination of data right now. Primarily, it's experimental data: the dictator game played with alterations on framing and whether the pot is earned through effort or luck. But, through a different project, there is also a wealth of personal data for each participant.

Experimental economics mostly aims to see how markets work in real life with exogenous real world randomness minimized, and with little regard for personal details of participants. The hope is, after all, that there is some kind of social scientific law that is indifferent to such details. But when play varies between groups of people, even if it almost always averages out to a certain "law" of aggregate behavior, all that personal data can probably say a whole lot about why people do the things they do, hence why markets work the way they work on a micro level, and hence how they evolve dynamically and what is the mechanical operation of the invisible hand.

Sounds pretty thrilling to me. In fact, that's serendipitously eerily related to what I wrote my NSF proposal on (never thinking that would actually be what I got to work on.)