Midday meals as an early childhood nutrition intervention: evidence from plantation communities in Sri Lanka

BMC Public Health

Table 7 IV regression model results- by Gender

	IV– Random Effects Regression
	Female			Male
	WAZ	HAZ	WHZ	WAZ	HAZ	WHZ
Treatment var
Trt	0.033** (0.012)	0.052** (0.016)	0.033 (0.025)	0.033** (0.008)	0.039* (0.018)	0.001 (0.026)
Other controls?	Yes	Yes	Yes	Yes	Yes	Yes
Observations	8567	5285	3943	8647	5024	3840
σ_u	0.039	0.090	0	0.060	0.109	0
σ_e	0.299	0.335	1.225	0.350	0.243	1.136
ρ	0.017	0.068	0	0.029	0.167	0
Kleibergen-Paap LM Stat	9.28**	68.84**	38.25**	15.85**	65.35**	34.55**
Cragg-Donald Wald F Stat	31.95	689.29	118.05	37.61	785.13	230.37
Sargan-Hansen Test	0.457	0.666	8.162*	0.132	0.019	5.473+

Robust standard errors in parenthesis, **p < 0.01, *p < 0.05, + p < 0.1
K-P LM test [Null: matrix of reduced form coefficients has rank = K1–1 (underidentification)], C-D Wald F Test [Null: Equation is weakly identified], S-H Test [Null: Overidentification restriction is satisfied]. Other controls include child age, age squared, gender, zscore of first weight, zscore of birthweight, zscore of first height, pcare, ethnicity, CDC status and birth-year fixed effects either as direct controls or instruments

ISSN: 1471-2458