Frank Robertson, Diego Flores, Marcia Ventura, Jordan Sgro and Veronica Taite, Brigham Young University
Social and Behavioral Sciences
Behavioral research has often focused on the impact of reinforcers on behavior. Punishers are occasionally used alongside reinforcers, and the relative strength of the two is a question of major importance in behavioral psychology. The matching law is a mathematical formula that attempts to evaluate the relative influence of two sets of reinforcers and punishers. The generalized matching law contains two parameters, sensitivity, which indexes how quickly participants respond to differences between reinforcement frequencies, and bias, which indicates how much one reinforcer is preferred over another. We have developed a new method for data analysis in situations containing both reinforcers and punishers, which expands the generalized matching law by a single parameter. The new parameter is punishment bias, where a value between 0 and 1 indicates the relatively greater effect of reinforcers (versus punishers) on behavior. A value greater than 1 indicates that behavior is impacted more by punishers. If this variation on the generalized matching law is valid, it would provide a simple metric for use in situations where negative and positive events both occur. Ten participants were asked to play a video game in which they moved a submarine around obstacles and collected coins. They could choose to do so on either the left or right side of the screen. Variable- interval reinforcement schedules were operative on both sides of the screen and changed every six minutes. Punishers were added to the left-side reinforcers in half of these schedules. We recorded each participant’s mouse clicks on each side, as well as the reinforcers and punishers they received. We analyzed this data using both the new method and the generalized matching law. We found that the new version fit the data better than the standard version and present it as a possible tool for future studies of the joint use of reinforcers and punishers.