The impact of feedback on perceptual decision making and metacognition: Reduction in bias but no change in sensitivity

It is widely believed that feedback improves behavior but the mechanisms behind this improvement remain unclear. Different theories postulate that feedback has either a direct effect on performance through automatic reinforcement mechanisms or only an indirect effect mediated by a deliberate change in strategy. To adjudicate between these two competing accounts, we performed a large-scale study including 443 subjects; approximately half of them received trial-by-trial feedback on a perceptual task, while the other half did not receive such feedback. We found that feedback had no effect on either perceptual or metacognitive sensitivity. On the other hand, feedback significantly affected subjects’ response strategies by reducing bias and improving calibration in both the perceptual and metacognitive judgments. These results strongly support the view that feedback does not improve behavior through direct reinforcement mechanisms but that its beneficial effects stem from allowing people to adjust their strategies for performing the task.

Reducing Public Masturbation in Individuals With ASD: An Assessment of Response Interruption Procedures

Individuals with autism may engage in sexual behavior at inappropriate times and/or in inappropriate places. The current study investigated the effects of response interruption and redirection (RIRD) and response interruption (RI) on public masturbation (PM) of children and adolescents with autism. Initial assessments showed that PM was maintained by automatic reinforcement. During the treatment evaluation phase, we compared RIRD and RI to determine whether either procedure was successful in decreasing the duration of PM. In the RIRD condition, contingent on the occurrence of any PM the participant completed physical activities involving both hands (e.g., moving chairs, touching toes). In the RI condition, the therapist interrupted all instances of PM using physical and verbal prompts (e.g., saying in a neutral tone, “Stop that” and moving hands away from genitals). Both procedures were effective in decreasing the duration of PM but RI required fewer resources and less time. Clinical implications and suggestions for future research are reviewed.

Trial-Based Functional Analysis Informs Treatment for Vocal Scripting

Research on trial-based functional analysis has primarily focused on socially maintained challenging behaviors. However, procedural modifications may be necessary to clarify ambiguous assessment results. The purposes of this study were to evaluate the utility of iterative modifications to trial-based functional analysis on the identification of putative reinforcement and subsequent treatment for vocal scripting. For all participants, modifications to the trial-based functional analysis identified a primary function of automatic reinforcement. The structure of the trial-based format led to identification of social attention as an abolishing operation for vocal scripting. A noncontingent attention treatment was evaluated using withdrawal designs for each participant. This noncontingent attention treatment resulted in near zero levels of vocal scripting for all participants. Implications for research and practice are presented.

Duration of wheel-running reinforcement: Effects on reinforcement value and motivation in free-feeding and food-deprived rats.

Ten (pair housed) female Long-Evans rats were exposed to 5 s, 30 s, and 90 s wheel-running reinforcement durations on a response-initiated variable interval 20 s schedule as food deprivation was manipulated. On free feeding, never-deprived rats showed low wheel running and lever-pressing rates with long postreinforcement pauses (PRPs) for the 5-s reinforcement duration. Subsequently, when food deprived (Deprived 1), rats showed no effect of reinforcement duration on all measures. Under a second deprived condition (Deprived 2) with the rats maintained in single cages, there was no effect of housing (single vs. paired). When data from both deprivation assessments (Deprived 1 and Deprived 2) were combined, rats showed lower wheel running and overall lever-pressing rates with longer pauses on the 90-s duration compared to 30 s and 5 s bouts of wheel activity. The pattern of results challenges a reinforcement value interpretation, but is consistent with shifts in the motivational basis of wheel running. On free feeding, never-deprived rats were intrinsically motivated to run on wheels and operant lever-pressing was maintained at moderate rates by the automatic reinforcement of wheel running, except at the short reinforcement duration (5 s). When food deprived, motivation became food-related and rats showed high rates of lever pressing even at the shortest duration. The weak effects under initial deprivation (Deprived 1) raise questions about equivalence between wheel-running reinforcement duration and reinforcement magnitude using food reinforcement.

Reinforcement of a reinforcing behaviour: Effect of sucrose concentration on wheel-running rate

Wheel running, unlike typical operant behavior, generates its own automatic reinforcement that alters the control exerted by extrinsic reinforcement on wheel running. The current study investigated the implications of the automatic reinforcement of wheel running by arranging different sucrose concentrations as extrinsic reinforcement for operant wheel running in ad-lib fed and food-deprived rats. Eleven female Long Evans rats ran on fixed revolution 30 schedules that delivered a drop of sucrose solution as reinforcement. Sucrose concentration varied across values of 0%, 2.5%, 5%, 10%, and 15% sucrose (w/v). Results showed that under ad-lib feeding, only the highest concentrations increased operant wheel-running rate. By contrast, under deprivation, all concentrations of sucrose increased the rate of wheel running. Despite the differences in sucrose-reinforced operant wheel-running rates by deprivation level (ad lib vs. deprived), wheel-running rates did not differ at the highest concentrations. Prior research on operant lever pressing, a response generating low (or no) automatic reinforcement, has shown considerably higher lever-pressing rates as a function of increasing amounts of sucrose reinforcement when rats are food deprived. Together, these previous observations and the current study suggest that automatic reinforcement generated by an operant decreases the control exerted by extrinsic reinforcement. Additionally, the regulation by extrinsic reinforcement on automatically reinforcing behavior depends on the organism’s motivation or deprivation level (ad lib vs. deprived).