Acquisition, generalization, and maintenance of question-answering skills in autistic children.

We conducted an investigation to evaluate the effects of a training strategy for teaching autistic students generalized responses to three forms of wh--questions (what, how, and why). Students were taught, using modeling and reinforcement procedures, to answer questions with magazine pictures as the referents. Each question form was divided into two or more subcomponents reflective of common social usage and was taught within the context of a modified multiple probe design across subcomponents. Following acquisition of each subcomponent, generalization to natural context and storybook questions was assessed; additional probes were conducted to assess responding over time and whether acquisition of responses to questions promoted question-asking skills. Results showed that the picture training procedure was effective in teaching a generalized response to questions for which the relevant cue was visible, whereas specific generalization programming was required for situations in which the relevant cue was not visible. All acquired responses were durable over time.

Effects of fenfluramine on social behavior in autistic children.

Deficit in social interaction is a primary component of infantile autism. However, in the majority of drug studies, social interaction has not been measured consistently over time. Therefore, we examined, in a crossover design, the effect of fenfluramine on the social interactions of seven autistic children. Social interaction was measured one to three times per week, while the children were in open placebo, placebo, or drug phases of the study. The results demonstrated that the effect of fenfluramine on social interaction was inconsistent across children, with two children possibly demonstrating a tolerance to the behavioral effects of the drug. The results are discussed with respect to genetic and pharmacologic factors.

Training respite care providers for families with handicapped children: Experimental analysis and validation of an instructional package.

We conducted four experiments to evaluate a respite care training package. In Experiment 1, we assessed the effectiveness of an instructional manual on the acquisition of respite care skills and compared the relative effects of three different manual on the acquisition of respite care skills and compared the relative effects of three different manual presentation formats. Results showed that performance during simulated (role-played) respite care situations improved substantially for all six trainees following presentation of the instructional manual (with no significant differences between presentation formats) although some remedial training was necessary to achieve mastery criterion. In Experiment 2, we evaluated the effects of the manual presented as a whole with a larger group of trainees and compared it to a workshop training approach. Results showed that both the instructional manual and workshop training approaches were effective, but that the former appeared to be more cost-effective. In both experiments, correct responding generalized to a respite care situation with a multiply handicapped child. In addition, several measures of social validity were obtained. In Experiment 3, we evaluated a simplified training and assessment package that could be conducted using the resources typically available to respite care agencies, and in Experiment 4, we demonstrated that respite care agency personnel could successfully implement this program.

A direct prompting strategy for increasing reciprocal interactions between handicapped and nonhandicapped siblings.

We investigated the effects of a sibling training procedure, consisting of direct prompting and modeling, on the occurrence of reciprocal interactions between nonhandicapped and handicapped siblings. Data were obtained for training, generalization, and follow-up. Results of a multiple-baseline design across three pairs of siblings showed that: direct prompting of interactions was an effective strategy for increasing reciprocal interactions between handicapped and nonhandicapped siblings; the training procedure resulted in increased levels of initiations and responsiveness to initiations; reciprocal interactions between siblings generalized to larger play groups or across settings; reciprocal interactions between handicapped subjects and untrained, nonhandicapped peers increased without direct training; the siblings' levels of interactions were maintained at 6 mo follow-up; and these findings were judged socially valid by the siblings' parents.

Experimental analysis of response covariation among compliant and inappropriate behaviors.

Reliable changes in a variety of behaviors, or classes of behaviors, when only one is manipulated experimentally, have demonstrated that even topographically dissimilar responses can be functionally related. We investigated such a relationship between topographically different child behaviors (compliance and inappropriate activities) by using a methodology that tests for response covariation. Five conditions were provided to sequentially increase and decrease first one and then the other of these behaviors, with the degree of covariation between the two behaviors (i.e., the relationship between changes in the targeted and nontargeted behaviors) being the finding of interest. Results showed that, regardless of the intervention used, the behavior targeted, or the direction manipulated, the nontargeted behavior reliably covaried inversely with the targeted one. The findings have immediate relevance to the clinical treatment of multiple behavior problems exhibited by children. Furthermore, the study of relationships between responses and the processes underlying these relationships can have important implications for understanding the complexity characteristic of human behavior not yet analyzed by behavioral research.

Establishing generative yes/no responses in developmentally disabled children.

We evaluated the effects of two procedures for teaching four developmentally disabled children to respond yes/no appropriately. During baseline, tutoring was conducted in which five known items were individually presented with the question, "Is this a ----?", followed either by access to requested items or by remedial prompting contingent on responding. When tutoring did not improve performance, instruction was embedded in the regular classroom activities. In this condition, items requested by students were either presented or withheld on the basis of their response to the question, "Do you want ----?". Increases in correct responding were confirmed by a multiple-baseline design across all four students and were maintained with the introduction of new items. However, generalization to "Is this a ----?" questions did not occur in the tutoring setting until specifically programmed. Subsequently, students also demonstrated appropriate yes/no responding to questions involving actions, possession, and spatial relations.

Training mildly handicapped peers to facilitate changes in the social interaction skills of autistic children.

We evaluated the effects of a peer-training strategy, consisting of direct prompting and modeling, on the occurrence and duration of interactions between autistic students and nonautistic peer-trainers. Data were obtained in both training and generalization settings. The results of a multiple-baseline design across students demonstrated that:the direct prompting procedure produced immediate and substantial increases in the occurrences and durations of positive social interactions between the peer-trainers and autistic students; these increases were maintained across time at levels above baseline during subsequent free-play probes; these findings were judged by teachers to be socially valid; untrained peers increased their interactions with the autistic students in three of the four groups; generalization of behavior change across settings occurred only after specific programming; and interactions between untrained peers and peer-trainers decreased following training. Variables that may account for the results and the implications of these findings for peer-mediated interventions are discussed.

The class specific effects of compliance training with "do" and "don't" requests: analogue analysis and classroom application.

Two experiments are reported in which the relationship between compliance with "do" and "don't" requests was examined with developmentally disabled children. In Experiment 1, a multiple baseline design across subjects with counterbalanced treatment conditions was used to evaluate a compliance training program composed of four phases: (a) baseline, during which no consequences were delivered for compliance, (b) reinforcement for compliance with one targeted "do" request, (c) reinforcement for compliance with one targeted "don't" request, and (d) follow-up with reinforcement on a variable ratio schedule for compliance with any "do" or "don't" request. Results of probes conducted before and after training within each condition indicated that generalized compliance occurred only with requests of the same type as the target exemplar ("do" or "don't"). In Experiment 2, these results were replicated in a classroom setting. Following collection of baseline probe data on student compliance, a teacher training program was successfully implemented to increase reinforcement of compliance first with one "do" and subsequently with one "don't" request of a target student. Results of multiple baseline probes across "do" and "don't" requests indicated that the teacher generalized and maintained reinforcement of compliance with other requests of the same type and to other students, with a resulting increase in student compliance with the type of requests reinforced. The impact of treatment on both teacher and student behavior was socially validated via consumer ratings. Implications of these findings with respect to response class formation and compliance training programs are discussed.

Response-reinforcer relationships and improved learning in autistic children.

In order to affect more rapid response acquisition for autistic children, researchers have recently begun to investigate the functional relationships of reinforcers to other components of the operant conditioning paradigm. Previous research suggested that functional relationships between target behaviors and reinforcers might be especially effective. For example, locating a reward inside a container might be a more efficient way to teach a child to open the container than by handling the child a reward for opening an empty container. The present experiment assessed, within a multiple baseline design, the possibility of improving autistic children's learning by changing arbitrary response-reinforcer relationships (while holding target behaviors and reinforcers constant) so that the target behaviors became functional (i.e., a direct part of the response chain required for the child to procure the reinforcer). The results showed that: (1) arranging functional response-reinforcer relationships produced immediate improvement in the children's learning, and resulted in rapid acquisition of criterion level responding; and (2) high levels of correct responding initially produced by functional response-reinforcer relationships were continued even when previously ineffective arbitrary response-reinforcer conditions were reinstated. The results are discussed in terms of understanding and improving autistic children's learning.

Normal peer models and autistic children's learning.

Present research and legislation regarding mainstreaming autistic children into normal classrooms have raised the importance of studying whether autistic children can benefit from observing normal peer models. The present investigation systematically assessed whether autistic children's learning of discrimination tasks could be improved if they observed normal children perform the tasks correctly. In the context of a multiple baseline design, four autistic children worked on five discrimination tasks that their teachers reported were posing difficulty. Throughout the baseline condition the children evidenced very low levels of correct responding on all five tasks. In the subsequent treatment condition, when normal peers modeled correct responses, the autistic children's correct responding increased dramatically. In each case, the peer modeling procedure produced rapid achievement of the acquisition which was maintained after the peer models were removed. These results are discussed in relation to issues concerning observational learning and in relation to the implications for mainstreaming autistic children into normal classrooms.

Reinforcer variation: implications for motivating developmentally disabled children.

Motivating developmentally disabled children to participate in educational activities can be very difficult. This is especially true for children diagnosed autistic. Because there is some evidence to suggest that stimulus variation may influence motivation, the present study investigated the effects of constant vs. varied reinforcer presentation on correct responding and on-task behavior. Results from a reversal design showed declining trends in both correct responding and on-task behavior when the same reinforcer was consistently presented, whereas, varying the reinforcers produced significantly improved and stable responding. the results are discussed in relation to the literature on stimulus variation and its effects on responsiveness.

Measuring musical abilities of autistic children.

Three normal children with reported musical ability and three autistic children were tested for the ability to imitate individual tones and series of tones delivered by voice, piano, and synthesizer. Accuracy of imitation was judged by two independent observers on the basis of pitch, rhythm, and duration. The autistic children overall performed as well as or better than the age-matched normal children. These results are discussed and their implications for future neurological and clinical research are considered.