|
Field
Intervention (Hidden)
|
Before
We run an experiment in which subjects are divided into High and Low intelligence groups. This is determined by their relative rank in a Raven Test - those above the median are High, and those below, Low. They then complete a sequential 2 player public goods game, indefinitely repeated (discount rate = 0.75). They play as either the first mover or a second mover.
We have 4 treatments - HH, HL, LH, LL - where HH refers to a 'High First Mover, High Follower' and HL refers to a 'High First Mover, Low Follower'.
Rather than let subjects play the game actively, they specify a plan of action - i.e. we elicit their full strategies. We then match them to someone in the session, and the computer plays out the game.
Plans of action are limited so that subjects strategies can only be conditional on Period t and t-1 actions. As the strategy space is infinitely large, we must enforce some limitations. We do this for both first and second movers, ensuring that they both must make 30 decisions in the experiment.
In addition, we elicit subjects beliefs about the other players contributions in Period 1 ( a point estimate) that is incentivised.
|
After
We run an experiment in which subjects are divided into High and Low intelligence groups. This is determined by their relative rank in a Raven Test - those above the mean are High, and those below, Low. They then complete a sequential 2 player public goods game, indefinitely repeated (discount rate = 0.75). They play as either the first mover or a second mover.
We have 4 treatments - HH, HL, LH, LL - where HH refers to a 'High First Mover, High Follower' and HL refers to a 'High First Mover, Low Follower'.
Rather than let subjects play the game actively, they specify a plan of action - i.e. we elicit their full strategies. We then match them to someone in the session, and the computer plays out the game.
Plans of action are limited so that subjects strategies can only be conditional on Period t and t-1 actions. As the strategy space is infinitely large, we must enforce some limitations. We do this for both first and second movers, ensuring that they both must make 30 decisions in the experiment.
In addition, we elicit subjects beliefs about the other players contributions in Period 1 ( a point estimate) that is incentivised.
|