Association for Behavior Analysis International

The Evolutionary Theory of Behavior Dynamics (ETBD) is a complex systems theory designed predict adaptive behavior. It uses an algorithm to instantiate rules based on Darwinian natural selection on an artificial organism (AO) that is represented by a population of behaviors. The repeated application of these rules produces a record of behavior that can be compared to live-organism behavior. The ETBD has been shown to produce live-organism-like delay discounting, which is the tendency for reinforcers to lose value as they become more delayed. However, it is currently unknown how different parameters of the ETBD’s algorithm influence discounting and how these parameters might relate to live organism discounting. Understanding individual differences in delay discounting, particularly differences in the rate that reinforcers lose value, is important because these differences have been linked to a wide variety of unhealthy behaviors. We assessed AOs’ delay discounting using an adjusting amount procedure and explored the effects of various parameters of the ETBD’s algorithm. The data show that these parameters influence individual differences in delay discounting. These results suggest ETBD may be useful for modeling and understanding individual differences in live organism’s delay discounting.

Reinstatement refers to the reemergence of a target response that was previously diminished following the delivery of the reinforcer originally maintaining that response. Woods and Bouton (2007) found comparatively lower levels of response-dependent reinstatement for organisms whose target response was eliminated with partial reinforcement (PRF) relative to extinction. This could be because occasionally reinforced responses during PRF are associated with both reinforcement and extinction, whereas in extinction, the modulating effects of reinforcement erroneously predict a complete reintroduction of the reinforcement contingency. The present study replicated the procedures described in Woods and Bouton (2007) with artificial organisms (AO) animated by the Evolutionary Theory of Behavior Dynamics (ETBD). Consistent with Woods and Bouton, AOs whose behavior was previously diminished with PRF evidenced lower levels of reinstatement relative to AOs whose behavior was diminished with extinction. Results from the present study suggest that PRF may be a viable tactic for inoculating against response-dependent reinstatement. By replicating Woods and Bouton, the present study is the first to demonstrate reinstatement with the ETBD, further showcasing the generality of the algorithm for modeling dynamic behavioral phenomena.

The Evolutionary Theory of Behavior Dynamics (ETBD) has made two specific predictions about choice under concurrent random ratio schedules: preference becomes more extreme with (a) larger differences between the concurrent ratio requirements and (b) smaller absolute values of the ratio requirement for the denser alternative. In this study, we tested the ETBD’s predictions by evaluating human participant’s choice under various concurrent random ratio schedules. Sixty-three undergraduate students participated and were presented with two concurrently available response options on a touch screen monitor. The difference between the concurrently available ratio requirements were manipulated across conditions and the absolute value of the ratio requirement for the denser alternative was manipulated across groups. As predicted by the ETBD, participant’s preference for the denser alternative increased as the difference between the concurrent ratio requirements increased and groups with smaller absolute ratio requirements tended to display more extreme preference. However, a high level of heterogeneity was observed across human participants within each group that was not evident in the behavior of artificial organisms animated by the ETBD.