Twenty-two studies were included (n=323). Mean quality score was 2.5 (range 1.2 to 3.6).
Intelligence quotient: ABA intervention produced a significant impact on general IQ (ES 1.19, 95% CI 0.91 to 1.47; 18 studies). ABA intervention had a significant impact upon non-verbal IQ (ES 0.65, 95% CI 0.17 to 1.13; 10 studies). There was evidence of heterogeneity for these comparisons.
Language skills: ABA intervention produced favourable effects on receptive language performance (ES 1.48, 95% CI 0.96 to 1.97; 11 studies). Favourable effects of ABA intervention were observed for expressive language skills measured by standardised assessments (ES 1.47, 95% CI 0.85 to 2.08; 10 studies). Favourable effects of ABA intervention were reported for general language skills (ES 1.07, 95% CI 0.34 to 1.79; five studies). There was evidence of heterogeneity for these comparisons.
Adaptive behaviour domains: ABA intervention produced significant effects for communication (ES 1.45, 95% CI 1.02 to 1.88; 11 studies). ABA intervention produced significant effects on socialisation (ES 0.95, 95% CI 0.53 to 1.37; 11 studies). ABA intervention produced significant effects on an adaptive behaviour composite measure that combined all four adaptive behaviour domains (ES 1.09, 95% CI 0.70 to 1.47; 15 studies). There was evidence of heterogeneity for these comparisons.
ABA intervention produced significant effects for daily living skills (ES 0.62, 95% CI 0.30 to 0.93; 11 studies). ABA intervention yielded significant effects on motor skills (ES 0.71, 95% CI 0.19 to 1.22, p=0.008); there was no evidence of heterogeneity for these comparisons.
Further results and sensitivity analyses were presented in the paper. Publication bias was present for all outcomes except for daily living skills, motor functioning and composite adaptation.