Abstract
Study objective: Traditional and more recent research on choice behavior has restricted the experimental procedures in three ways: restrictions on static properties of the alternatives and choice behavior; a disproportionate emphasis on reinforcers and reinforcement schedules; and disregard of possible regular variability or sequences in the participants' choices. After examining several recent studies that included procedures that might overcome these restrictions, we propose a predictive game task composed of two types of discriminative stimuli (predictive and reporting stimuli) and two types of responses (choice and performing responses). After a brief introduction describing game theory for analyzing choice behavior in predictive game tasks as normative theory, we discuss experimental procedures corresponding to five components of game theory:players, payoff matrices, alternatives, moves and strategies, initial conditions, and playing procedures. Based on this discussion, we conclude that the predictive game task is appropriate for a wide range of experimental procedures.
