Children with autism have a tendency to engage in repetitive behaviors across multiple domains, including play. One method for decreasing repetitive behavior that has been studied in humans and other animals in both laboratory and applied settings is the explicit reinforcement of variability. Basic research suggests that varied responding can be brought under the control of an antecedent stimulus that reliably predicts the availability of reinforcement for variability. The purpose of this study was to examine the effects of a lag 3 schedule on the diversity of forms constructed during block play by children with autism. In addition, the study evaluated the effects of a procedure designed to establish stimulus control over repetition and variation, and examined the extent to which variability would generalize to a novel task in which the same discriminative stimuli were presented.
Three boys diagnosed with autism, ages 7 to 9, served as participants. All three lacked appropriate play skills. During daily sessions, the experimenter presented the students with materials to either build block structures or create patterns on a pegboard. Variability was defined as the degree to which the forms created varied from one another. In an initial continuous reinforcement (CRF) phase, any form that the students created resulted in reinforcement. In a subsequent variability (VAR) phase, the experimenter asked the participant to make something “different,” and then reinforced only the creation of forms that were different from the previous three that were built (a lag 3 schedule). After this, students experienced a repeat (REP) phase in which they were instructed to make “the same,” and only forms that repeated one of the three most recently produced resulted in reinforcement (a rep 3 schedule). After experiencing multiple reversals between these two phases, an alternating schedule (ALT) phase was introduced in which each trial was determined by a random sequence to operate on either the VAR or REP contingencies, and the discriminative cues (“same” and “different”) were presented on each trial. An Sd absent (SDA) phase in which the discriminative cues were removed followed this. Finally, students were exposed to a novel task (painting a series of squares) in which the same cues were used. A reversal design was employed to compare the effects of these schedules on response variability.
All three participants increased the diversity of block forms produced under the lag 3 schedule. All three participants also demonstrated evidence of stimulus control in that responding varied more under the VAR component of the multiple schedule in the ALT phase, but the two conditions did not have the same effect when the discriminative stimuli were absent during the SDA phase. During the generalization task, students did not demonstrate stimulus control, in that the sequences of colors produced were equally likely to vary whether students were asked to make “same” or “different.” The implications of these findings for the education of children with autism are discussed, along with suggestions for future research directions.