| Sumario: | AbstractDeficits in facial expression recognition often hinder social interaction and communication for children with autism spectrum disorder (ASD). While numerous interventions have been developed to address this, there is still no clear consensus on which methods are most effective across different learning phases. To address this gap, this meta-analysis evaluated the effectiveness of facial expression recognition interventions by analyzing data from 22 single-subject design studies using the Tau-U statistical method within a random effects model. The results demonstrated moderate to large effect sizes during the intervention and maintenance phases, while the generalization phase yielded a medium effect. Moderator analyses revealed significant differences based on intervention duration and methods. Specifically, a duration of 6–8 wk and conventional methods significantly outperformed interventions lasting over 8 wk and technology-based approaches, respectively. Although other targeted variables were not statistically significant, the data highlighted important practical trends. Interventions yielded higher point estimates for children with ASD aged 7–10 and those with mild to moderate ASD. For program design, the findings suggest utilizing a single interventionist, targeting 3–6 expressions, practicing for 20–45 min daily, and conducting sessions in low-structure environments. These results provide clear, evidence-based guidelines for designing individualized facial expression recognition programs for this population.
|
|---|