Comparing video modeling and least-to-most prompting with matrix training in teaching sociodramatic play to children with ASD

The purpose of this study was to compare the effectiveness and efficiency of video modeling (VM) and least-to-most prompting (LTM) with matrix training to teach sociodramatic play skills to children with autism spectrum disorder (ASD). Three children with ASD, aged between 8-9 participated in the study. The study employed single-case non-concurrent multiple baseline design across participants embedded with adapted alternating treatments design. The study was conducted at each child's home settings with the participation of their mothers as a play partner. Effectiveness findings showed that both VM and LTM with matrix training were equally effective to teach sociodramatic play in terms of acquisition, maintenance, and generalization. Efficiency findings indicated that there was no significant difference between the procedures in terms of number of sessions/trials to criterion and that LTM with matrix training was more efficient for two children in terms of number/percentage of errors to criterion. Additionally, the findings showed that LTM with matrix training was more efficient for all children in terms of number of minutes of instruction to criterion. Social validity findings positive evaluations of child participants, their mothers, and special education teachers working in the field.