A computer task was developed for humans based on Zentall and Stagner's (2011) choice task for pigeons. In their procedure, pigeons were given a choice between one alternative that produced one of two stimuli that each always predicted three pellets of food (a mean of three reinforcers) and another alternative that on 20% of the trials produced a stimulus that predicted the delivery of 10 pellets of food and on 80% of the trials produced a stimulus that predicted the absence of food (a mean of two reinforcers).
Consistent with human gambling behavior but contrary to optimal foraging theory, the pigeons showed a strong preference for the alternative involving low probability and high payoff (men of 2 pellets) over the alternative with the greater net payoff (3 pellets).
To see the paper: http://rspb.royalsocietypublishing.org/content/early/2010/10/13/rspb.2010.1607
In our computer task, participants chose between two alternatives (two planetary systems: ZORB or ARTO). Choice of one alternative (ZORB) on 20% of the trials sent the participant to shoot at enemy spaceships on a red planet that was followed by the killing of 10 enemy generals or on 80% of the trials a green planet that was followed by the killing of no enemy generals. Choice of the other alternative (ARTO) sent the participant to shoot at enemy spaceships either on a blue planet on 20% of the trials or a yellow planet on 80% of the trials both of which were always followed by the killing of 3 enemy generals. Planetary systems and colors were counterbalanced. We recently used this task to investigate the similarities and differences in their tendency to choose "maladaptively."