Effectiveness of intervention using artificial intelligence for students with disabilities: systematic review and meta-analysis

Background Algorithmic systems do not sufficiently capture the contextual nuances of disability, thereby risking model bias and reducing ecological validity. Hence, this review aims to investigate the various pedagogical and ethical concerns that have resulted from the integration of artificial intelligence (AI) into special education. Methods We conducted a systematic review and meta-analysis to investigate the effects of intervention using AI on students with disabilities. For this purpose, 16 papers were selected from an electronic database, and relevant data were extracted. Results Our analysis revealed that the primary targets of AI intervention were students with learning disabilities (31.3%). Regarding methodological characteristics, only one study was a randomized clinical trial. The overall effect size (ES) was significantly large (Hedges' g = 1.263). The ESs of 14 studies were significant (87.5%), whereas those of two studies were insignificant (12.5%). Subgroup analysis revealed significant differences in ESs based on outcome, type of disability, grade, and implementer. Within each subgroup, the largest ESs were observed in the cognitive outcome, ADHD, and university student groups, as well as when the implementer was a researcher. Conclusions The results of this study suggest that AI interventions for students with disabilities are effective and that their effectiveness is increased when used as collaborative support tools, rather than replacing teachers in educational settings.