Tablet-based interventions for individuals with ASD: Evidence of generalization and maintenance effects

Despite positive effects of and established empirical evidence for tablet-based interventions for individuals with autism spectrum disorder (ASD), it is not known whether such findings can be applicable to maintenance and generalization effects of those interventions.

This systematic review evaluated peer-reviewed single-case experimental design (SCED) studies concerning evidence of generalization and maintenance effects of tablet-based interventions for individuals with ASD.

To evaluate the methodological rigor of the literature in terms of generalization and maintenance effects, the authors adopted four of the basic design standards developed by What Works Clearinghouse (WWC; Kratochwill et al., 2010/2014). In addition, the authors used Tau-U effect size measure and attempted to calculate effect sizes differentiated by the type of generalization and maintenance teaching strategies.

A total of 21 studies assessed generalization and/or maintenance effects. In the first evaluation of evidence of generalization and maintenance effects, it was found that more than half of the studies included in this review collected interobserver agreement (IOA) on at least 20% of sessions across all generalization and maintenance conditions and met the minimum quality thresholds of IOA. Meanwhile, less than one third of the studies included more than three data points in each generalization and maintenance condition. With regard to maintenance of effects, about half of the reviewed studies did not report the latency to the maintenance measure, which may hamper the assessment of the clinical and practical significance of the effect of the tablet-based intervention. In the second evaluation, the omnibus Tau-U effect size for baseline to generalization comparisons resulted in a moderate effect. For the contrasts between intervention and generalization comparisons resulted in a small effect. The omnibus Tau-U effect size for baseline and maintenance comparisons resulted in a strong effect. For the contrasts between intervention to maintenance comparisons, the omnibus Tau-U effect size resulted in a small effect.

Findings in this review suggest that efforts should be made to establish a system for appraising generalization and maintenance procedures in SCED studies. In addition, future studies should investigate if tablet-based interventions are truly effective in creating sustainable behavioral change in individuals with ASD.