Association for Behavior Analysis International

Matching theory, delay discounting, and behavioral momentum theory have been among the most conceptually generative advances in the quantitative analysis of behavior. Though these accounts often are divergent in terms of contemporary investigation and application, they are united in that they all focus on complex interactions between behavior, temporal dimensions of reinforcer delivery (e.g., reinforcer rate or delay), and the context in which reinforcers are encountered. Put another way, the timing of reinforcers is central to each of these accounts. With this underlying theme in mind, the present symposium will arrange four talks in which speakers present recent advances in these theoretical domains. In the first two talks, choice behavior will be discussed in relation to both the dynamics of response allocation across various levels of temporal resolution and the role of memorial processes in choosing between immediate and delayed rewards. The final speakers will discuss advances in the study of resistance to change. Specifically, these talks will focus on the unique contributions of reinforcer rate both to response strength and to contingency discrimination during disruption of ongoing behavior. Elizabeth Kyonka, an expert in quantitative analyses of behavior and decision-making processes, will serve as a discussant.

Behavioral momentum theory suggests that the association between discriminative stimuli and conditions of reinforcement during training is the main determiner of resistance to change under conditions of disruption. Studies have shown the limits of this interpretation by employing ratio schedules in which reinforcement rate is directly dependent on response rate. In contrast, the well-known autoshaping procedure arranges response-independent reinforcement and generates consistent operant responding. Two experiments compared resistance to change of autoshaped responding in pigeons. Experiment 1 evaluated resistance to change of autoshaped responding in a multiple schedule arranging two different probabilities of reinforcement with prefeeding and extinction. Experiment 2 compared resistance to change between subjects with pigeons responding on simple autoshaping procedures that arranged different probabilities of reinforcement. Results are generally consistent with predictions of behavioral momentum theory, and provide new challenges and directions in addition to an extension of resistance to change to procedures arranging only response-independent reinforcement.

Generalized matching is a quantitative description of the relation between the allocation of responses and relative rates of reinforcement. The functional relation of response allocation to rate of reinforcement can differ at different temporal resolutions. To illustrate matching analyses over multiple temporal resolutions, we used data from pigeons that pecked in concurrent-chain schedules. Initial links were dependent concurrent variable-interval (VI)-VI schedules. Terminal-link schedules were fixed intervals. The location of the initial link leading to the shorter terminal link varied pseudorandomly. Global analysis of functional relations across conditions reveals differences in sensitivity and systematic deviations from linearity that may be difficult to detect at more local levels. Within-session cumulative response plots illustrate the abrupt shifts in behavior that occur in this procedure. Local analysis of functional relations, including visit durations within initial links, can capture the patterns of response allocation that, in the aggregate, are responsible for differences in matching observed at global resolutions.

Behavioral momentum theory posits that the rate of responding during extinction is determined by both the discrimination of extinction contingencies, and an underlying behavioral mass or response strength established by baseline rates of reinforcement. However, evidence such as the partial reinforcement extinction effect (PREE) suggests that discrimination effects influence behavior to a greater extent than does reinforcement history. We present data from four identical conditions where two responses were maintained in separate stimulus contexts by differential rates of reinforcement. These stimulus contexts were then combined during extinction. Across successive replications, decreases were observed in both the absolute rates of responding and, to a lesser degree, the relative rate of responding between alternatives. However, differential responding towards the richer alternative was maintained throughout all 3 replications. The present results suggest that, while repeated extinction conditions enhanced discrimination effects and lead to faster decreases in overall responding, baseline rates of reinforcement still contribute to differential responding during extinction.