| Budget Amount *help |
¥3,800,000 (Direct Cost: ¥3,800,000)
Fiscal Year 1999: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 1998: ¥3,200,000 (Direct Cost: ¥3,200,000)
|
|---|
| Research Abstract |
We decided on the following two topics in our behavioral research of game situations. Both topics focus on the dynamic processes of choice behavior in game situations: 1) Methodological and theoretical considerations of games in animal learning research. 2) Identification of variables that control choice behavior in game situations.
We devised (or critically analysed) the Prediction Game Task (PGT) and showed that it involves parts of various experimental procedures, such as the concurrent chains schedule of reinforcement and the delayed matching-to-sample procedure with different outcomes. We also reviewed quantitative models in behavior analysis and related fields and examined several models that are appropriate for the simulation of choice behavior in game situations.
PGT is a procedure that consists of five successive events: (1) a predictive stimulus, (2) a choice response, (3) an informative stimulus, (4) required responses for obtaining reinforcers, and (5) reinforcers. Coincidenc
… More
e vs. non-coincidence between a predictive and an informative stimulus represents a cooperative vs. defect move of an experimenter (an environment) side, but coincidence vs. non-coincidence between a predictive stimulus and a choice response represents a cooperative vs. defect move of a subject side. According to a payoff matrix that characterizes a game situation, combinations of two moves determine the number of responses required to receive a reinforcer. Under a series of experiments that adopted "the chicken dilemma" as a payoff matrix, rats showed strong position preference and their behavior was not modified by time-out and correction method, prolonged inter-trial intervals, and shortened intervals between a choice response and the onset of an informative stimulus. Human subjects, on the other hand, followed the choice-distribution ratio of the experimenter side if there were no predictive stimuli, and followed the ratio of the predictive stimuli if they were introduced. Their behavior, however, depended on the information reliability correlated with these stimuli.
In our studies on resistance to change (RC), which is an another measure of preference and its dynamic process in a choice situation, pigeons showed higher RC during an alternative that produced higher reinforcement rate when their responding was disrupted by prefeeding, but not when it was disrupted by extinction that followed an exposure to a concurrent VI VI schedule. But in both the prefeeding and extinction procedures, no differences in choice proportion were obtained between baseline training and response disrupting operation. Less
|
