Theses and Dissertations

Determining the More Effective Behavior Analytic Intervention for Children With Autism Who Exhibit Pica Behaviors

Jennifer J. Lanham, Nova Southeastern University


This dissertation was designed to determine which behavior analytic intervention was more effective in the treatment and reduction of mouthing non-nutritive substances in children diagnosed with autism. This study included four participants in an A-B-A reversal design with a component analysis across four intervention phases. The study participants were enrolled in a center-based treatment environment for children with autism who displayed frequent pica behaviors that present a danger to health and create a barrier to learning. Parents have voiced significant concern for each of their children who were participants and sought out effective interventions that are generalizable across environments. The participating parents of the children and adolescents for this study received required training for appropriate replacement behaviors to maintain low to zero levels of mouthing behavior each day following the fading out of the intervention.

The writer established three treatments beginning with redirection paired with a mild punishment procedure; next was skill acquisition with discrimination training; and the third treatment applied discrimination training with a food exchange procedure, before the removal of pica treatment. Prior to treatment, each participant took part in a series of procedures that included a functional analysis and a preference assessment for foods, items, and activities that might be reinforcing. The first phase was the preintervention and baseline data collection, along with caregiver meetings and completing documentation and assessments. The second phase began with the first intervention, which used redirection of the pica behavior, response blocking with a verbal reprimand to stop the chain of behaviors leading to mouthing before contact with the non-nutritive substance and the child’s mouth. The second intervention, introduced in the third phase, consisted of response-blocking paired with discrimination training and noncontingent access to preferred foods. The fourth phase included the training and execution of a food exchange procedure to replace pica items with appropriate food consumption. Lastly, the fifth phase tracked the maintenance and generalization of pica behavior when pica interventions were removed. The level of intrusiveness between phases increased with each intervention. The component analysis was based upon passage of time, which lasted 2 weeks per treatment, with the same amount of intervention for each phase received by each participant.

An analysis of the data revealed that the participants’ data across all phases of intervention showed a decreasing trend, which resulted in low to zero levels of pica behavior, although two participants showed spontaneous recovery of the behavior yet also reduced pica instances during the next sessions. The intervention that reduced pica behavior to the lowest levels for all participants was Phase 4 in which the participants exchanged pica items for preferred foods. All parents expressed satisfaction with the reduction in pica, reduction of all maladaptive behavior, and improved language and social skills.