Association for Behavior Analysis International

The matching law has proven robust across species, methods, and reinforcers. Research in behavioral economics has demonstrated the asymmetrical effects of gains (reinforcers) and losses (punishers) on human decision making in hypothetical scenarios primarily involving monetary outcomes. We applied matching-based analysis to the measurement of the gain-loss differential in three parallel experimental series involving human participants, pigeons, and rats. The experimental procedure included conjoint interdependent concurrent variable-interval schedules of reinforcement and punishment. Human participants played a customized video game in which the concurrent schedules were associated with the two sides of a screen containing moving submarines and underwater targets. Participants gained or lost on-screen points or actual coins (dispensed by and inserted into a device next to the screen). Nonhuman subjects received access to food or lost that access while responding in standard experimental chambers (pecking for pigeons and lever pressing for rats). Molar data analysis utilized a set of models derived from the generalized matching law to produce model-specific parameter values for gain and loss, respectively. Molecular data analysis utilized event recording to identify behavioral patterns characteristic of the transition from gain-only to gain-plus-loss conditions and vice versa. We will describe the results from each experimental series, then compare and contrast the three sets of results.

Organisms can anticipate an environmental event when it occurs regularly in time or when it is preceded by a cue. Three experiments were conducted in order to explore the interaction between these two strategies. Each experiment involved to maintain constant or variable certain temporal relation between the conditional stimulus (CS, noise) and the unconditional stimulus (US, water) in a Pavlovian conditioning procedure. Rats were used as experimental subjects. In Experiment 1 subjects were exposed to a fixed or a variable intertrial interval; in Experiment 2 subjects were exposed to a variable interstimulus interval, but one of the stimuli occurred at fixed interval; and in Experiment 3 subjects were exposed to fixed or variable duration of the CS. Approaching to the water dispenser was established as the target response. When the US-US interval was fixed subjects gradually increased responding over the interval. In contrast, responding was relatively constant when the US-US interval was variable. Furthermore, responding increased when the CS was presented. This increase depended on the temporal proximity of the CS with the US. These results suggest that the organisms adjust their behavior to the temporal regularities of the stimuli occurrences.

We performed two experiments in which pairs of participants were asked to coordinately respond in order to achieve a common goal. Each participant would move a piece in a virtual chessboard and meetings in adjacent cells in specific areas were reinforced according to an ABAB design, experiment 1, and ABA design, Experiment 2. For experiment 1, meetings were not reinforced at conditions A. Meetings were reinforced by successive approximations at conditions B. The results from the first experiment showed that the variability index of IBCs and aggregate products were smaller in conditions B than in conditions A. However, they did not completely stereotyped. Experiment 2 was design to evaluate if longer exposure to intervention condition would enhance stereotypy in the IBCs and their aggregate product. The second experiment used ABA design where conditions A where intervention condition, and condition B baseline. Results of experiment 2 shows that a complete stereotypy on IBCs and aggregate product was found. Both experiments may be indicating that contingent reinforcing consequences may have a selective effect upon IBCs and their aggregate products.