Table 6.
Cluster robust meta‐regression for grade level for academic effect sizes.
| Effect (n of effect sizes) | Hedges’ g | 95% CI of g | df | p |
|---|---|---|---|---|
| Elementary schoola (36) | 0.36 | (0.09, 0.62) | 6.41 | 0.017 |
| Preschool (44) | −0.16 | (−0.43, 0.11) | 13.44 | 0.233 |
| Middle school (19) | −0.09 | (−1.14, 0.96) | 1.35 | 0.625 |
| High school (14) | −0.15 | (−0.41, 0.12) | 6.41 | 0.234 |
Note: N of studies = 24, N of effect sizes = 113, assumed ρ = 0.8, I 2 = 78.83. Cluster robust estimates with df < 4.00 should be interpreted with caution (Tanner‐Smith, Tipton, & Polanin, 2016). Wald test for null difference between preschool, middle and high school grade levels, F(3, 2.82) = 0.41, p = 0.761. Positive effect sizes favor Montessori over traditional education.
Abbreviation: CI, confidence interval.
a
Reference category/intercept.