Table 2.
Associations Between Caregiver Cognitive Stimulation and Developmental Outcome Domains
| Domain | N | r p | 95% CI | p | Prediction Interval | τ 2 | Q m | |
|---|---|---|---|---|---|---|---|---|
| Effect Sizes | Studies | |||||||
| Cognitive | 250 | 107 | .12 | [.10, .14] | <.001 | [−0.09, 0.32] | 0.004 | 159.10 |
| Psychosocial | 57 | 35 | .10 | [.07, .13] | <.001 | [−0.11, 0.30] | 0.001 | 43.27 |
| Physical | 23 | 14 | .17 | [.11, .25] | <.001 | [−0.04, 0.38] | 0.021 | 57.29 |
| Composite | 24 | 20 | .13 | [.08, .18] | <.001 | [−0.08, 0.33] | 0.008 | 31.06 |
Note. CI = confidence interval. rp = partial correlation coefficient (effect size). The sign of the partial correlation coefficients (rp) indicates whether caregiver cognitive stimulation is positively or negatively associated with child development. The prediction interval reflects the dispersion of partial correlation coefficients, such that 95% of studies comparable to those in the analysis will have the true effect size fall within the respective interval (Bornstein, 2024). Across outcome domains, heterogeneity was significant, τ12 = 0.01, τ22 = 0.01, Qm(4, 124) = 49.36, p < .001. While each domain is statistically different from zero, none of the estimates are significantly different from each other.