Some colleagues of mine and I are considering putting a research proposal together under the auspices of experimental economics. We keep seeing these really cool results in our game theory class-such as the centipede game-that theoretically don't match up to experiments performed in previous experiments. Beyond the self-selection issues inherent in laboratory economic experiments, we believe the amount of information (and payoff structures in a game of perfect information) can be controlled to produce expected results.
For example, if you have a 5 to 10 minute chat with the person you are soon to play a game against, we believe an "altruistic effect" may be larger, i.e. a person would be willing to achieve higher payoffs regardless if the effect conforms to a Nash Equilibrium.
I don't want to give away too many details (it would be nice to ensure this particular experiment hasn't already been performed!). We'll see where it goes.
No comments:
Post a Comment