A Functional Analysis of Stimulus Control Strengths of Antecedents and Consequences in Learning

We analyzed the contribution to the stimulus control in learning from the antecedents and the contingent consequences in this research. In the first experiment, we analyzed the stimulus control strengths of the preferred visual antecedent targets and prosthetic reinforcers in skill acquisition. The preferred- (PA) and non-preferred-visual-antecedent-stimuli (NA) were paired with two contingent consequences. In the conditions paired with the consequence of praise-for-correct-response (PC), researchers praised correct responses and implemented a correction procedure contingent on incorrect responses. In the conditions paired with the math-for-correct-response (MC), the procedure was the same except that a correct response was followed by presenting a non-preferred activity of doing a math problem. We measured the acquisition rates and maintenance of responses of PA and NA across the two consequence conditions. The results showed that all participants acquired the PA faster regardless of the consequence conditions. The findings suggest that the see-say correspondence of the PA may function as a stronger reinforcer in skill acquisition than the contingent consequence of prosthetic reinforcer delivered by the instructor. In the second experiment we controlled the reinforcement from the antecedent stimuli and conducted a component analysis of skill acquisition consequences. In the learn unit (LU) condition, researchers praised correct responses and implemented a correction procedure contingent on incorrect responses. In the praise-only-for-correct-responses (PC) condition, researchers delivered contingent praise for correct responses and ignored incorrect responses. In the correction-only-for-incorrect-responses (CI) condition, researchers ignored correct responses and implemented the correction procedure contingent on incorrect responses. We manipulated this independent variable across educational and abstract stimuli and measured acquisition rate, duration, and maintenance of responses. The results showed that the LU and CI conditions were both effective on teaching listener responses and were more effective than the PC procedure. The results suggested that the correction procedure was probably necessary and sufficient for skill acquisition and maintenance. Since the stimulus set sizes in the previous experiments were randomly decided, in the third experiment we conducted a systematic replication of Kodak et al.’s (2020) and Vladescu et al.’s (2021) research on the effects of stimulus set sizes on skill acquisition. We manipulated the stimulus set sizes by teaching three, six, and 12 sight words simultaneously during learn unit instruction. Researchers taught participants until the participant’s responding reached the acquisition criterion for 12 different sight words per set size condition. The acquisition criterion was set for an individual operant, whereby when accuracy met criterion for a single sight word, that sight word was replaced in the following session. The results showed that the set-size-three was more efficient than the set-size-six and -twelve in acquisition, which were more consistent with Vladescu et al.’s findings, but not consistent with Kodak et al.’s findings. However, the set-size-twelve reliably produced the highest maintenance levels for all participants. The opposing acquisition and maintenance results suggest further discussion on the definition of “effectiveness” in learning. To sum up, the results of these studies demonstrated that the strengths of stimulus control were a function of the synergistic reinforcement strengths across multiple correspondences of motivating operations, discriminative stimuli, and the contingent consequences