Variability is a dimension of behavior that may come under the control of consequent and antecedent stimuli (e.g., Page & Neuringer, 1985, Exp. 6). It has been already demonstrated that other operant behavior dimensions such as response rates can come under the control of stimuli trained in previous experimental conditions, when the contingencies change (e.g., Freeman & Lattal, 1992). The goal of this work is to extend our understanding of behavior variability as a dimension of operant behavior, answering the following questions: a) as a behavior dimension, does behavior variability come under the control of stimuli previously trained, when the contingencies change? b)does such stimulus control depend on the degree of variability required by the contingencies? and c)is the sensitivity to changes in the variability contingencies modulated by the similarity between the discriminative stimuli in the history and test phases? Experiments 1 and 3 of the present study are delineated to investigate questions a) and b), using an single-subject design in which a multiple schedule (mult) will arrange the reinforcement of Low Variability (LB) under a stimulus condition, and High Variability (HV - Experiment 1) or Repetition (REP - Experiment 3) under a different stimulus condition, during the History Phase. During the subsequent Test Phase, both components will be replaced by a yoked schedule that will deliver reinforcers in the same order in which they were delivered during the previous phase, without changes in the stimulus conditions. Experiments 2 and 4 are delineated to answer questions b) and c), using a single-subject design in which the stimulus conditions of the mult HV LV (Experiment 2) or mult HV REP (Experiment 4) will change from the History Phase I to the Test Phase I for only half of the subjects. The History Phase II and Test Phase II will be identical to the previous respective phases, but the stimulus conditions will change only for the subjects that did not experience such change during the previous Phase I transitions. The results of each experiment will be analyzed both quantitatively and qualitatively in terms of variability (U values and percent of possible sequences that were emitted, respectively)and in terms of the percent of reinforced sequences in each component.
News published in Agência FAPESP Newsletter about the scholarship: