Table 6.

Two-stage least squares impacts of program participation on program components (“Treatment on the Treated”), instrumenting for participation using treatment status and controlling for baseline characteristics (percent, unless otherwise noted)

Program component	Treatment mean	Control mean	Impact	p-Value	Adjusted p-value	Sample size
In previous 12 months, exposed to information on
Relationships	17.82	20.84	− 3.02	0.680	1.000	221
Parenting	68.24	44.60	23.65	0.010	0.043	220
Child health care	57.28	52.63	4.65	0.614	0.999	221
Education related services	29.44	28.80	0.64	0.930	1.000	220
Career Counseling or job training	18.54	22.05	− 3.51	0.624	0.976	220
Methods of birth control	84.70	67.22	17.48	0.024	0.114	217
Percent correct on assessments of knowledge of contraception
Condoms	58.55	61.09	− 2.54	0.470	0.920	220
Birth control pills	45.08	55.14	− 10.06	0.050	0.144	220
IUDs	35.38	34.87	0.51	0.904	1.000	220
Other methods	37.11	36.70	0.41	0.923	1.000	220

Treatment and control group means are regression adjusted. Impacts on binary outcomes are estimated using a two-stage least squares model instrumenting for program participation using treatment status, with standard errors adjusted to account for heteroskedasticity. Regressions control for all baseline characteristics and stratum in addition to treatment status. All p-values are based on a two-sided test, and adjusted p-values control for the familywise error rate using the method in Hothorn et al. (2008). Sample sizes differ across outcomes due to missing outcome data. Source: Baseline survey and 12 month follow-up survey

^*Significantly different from zero at the .10 level after adjusting for multiple comparisons, two-tailed test

^**Significantly different from zero at the .05 level after adjusting for multiple comparisons, two-tailed test

^***Significantly different from zero at the .01 level after adjusting for multiple comparisons, two-tailed test