Table 3.

AWH-E Impacts on female menstrual health practices and knowledge.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Indic talks to mother about menstruation	Indic activ not affected by menst	Indic impr menst mgmt	Indic approp disposal	Knowledge - indic girls reach puberty first	Knowledge - indic menstruation frequency	Knowledge - indic menarche allows pregnancy
Panel A: East Hararghe sample
=1 if in AWH-E community	0.046*	0.102	0.262**	0.042	0.002	0.102**	0.090**
	(0.023)	(0.075)	(0.100)	(0.086)	(0.040)	(0.044)	(0.037)
Control mean	0.079	0.785	0.195	0.361	0.725	0.351	0.586
Observations	493	112	112	92	607	605	607
Panel B: South Gondar sample
=1 if in AWH-E community	0.222***	−0.060	0.180	−0.104	0.016	0.158***	−0.052*
	(0.042)	(0.154)	(0.166)	(0.184)	(0.035)	(0.038)	(0.027)
Control mean	0.145	0.952	0.466	0.463	0.287	0.584	0.810
Observations	602	50	49	47	645	644	645

This table presents results from regressions for adolescent female outcomes, including coefficient estimates and standard errors (in parentheses). All regressions include as covariates adolescent age at the time of recruitment, an indicator for households that had multiple adolescents eligible for the study, sampling block and survey month fixed effects. In additional, all regressions control for household size, a household asset index, an indicator for the household head being literate, an indicator for the household head being female, and an indicator for the household having ever received PSNP benefits. Regressions are weighted to maintain initial population proportions, and standard errors are clustered by kebele. ^* denotes statistical significance at p < 0.1, ^** at p < 0.05, ^*** at p < 0.01.