. 2018 Jan 1;27(5):1594–1604. doi: 10.1007/s10826-017-0987-y

Table 3.

Mixed effects analysis of behavioral health, economic hardship, and labor market outcome changes in the control, partial and full intervention groups in the Building Wealth and Health Network RCT, Philadelphia 2014–2015

Assessment period^a
Measure	Baseline	9 mo.	9 mo. vs Baseline	12 mo.	12 mo. vs Baseline	15 mo.	15 mo. vs Baseline	Group^b
	LSM	LSM	p-value	LSM	p-value	LSM	p-value	p-value
Adult depressive symptoms
Control group	10.66	9.55	0.3998	11.55	0.5134	12.83	0.1349
Partial intervention	9.02	9.46	0.3689	9.97	0.9712	11.36	0.9215	0.4098
Full intervention	9.78	10.36	0.3085	10.26	0.8036	8.65	0.0640	0.0154
Adult self-efficacy
Control group	31.89	29.05	0.0589	30.36	0.3212	31.10	0.6293
Partial intervention	29.59	29.15	0.2237	30.89	0.1438	32.15	0.1064	0.5903
Full intervention	31.90	32.98	0.0388	32.39	0.2950	32.62	0.4537	0.4170
Child’s developmental risk
Control group	0.10	0.31	0.0680	0.10	0.9507	0.09	0.8882
Partial intervention	0.23	0.29	0.5741	0.14	0.4239	0.17	0.6031	0.5575
Full Intervention	0.11	0.19	0.4568	0.28	0.1321	0.27	0.1612	0.1883
Hardship index
Control group	2.39	2.16	0.5590	2.60	0.6115	2.27	0.7868
Partial intervention	2.42	2.12	0.4690	1.97	0.2457	2.04	0.3479	0.6663
Full intervention	2.59	2.56	0.9557	1.86	0.0640	2.19	0.3268	0.8783
Employment
Control group	0.12	0.43	0.0068	0.61	<0.0001	0.38	0.0384
Partial intervention	0.08	0.56	0.2511	0.57	0.9751	0.48	0.3508	0.4867
Full intervention	0.04	0.43	0.5951	0.44	0.5637	0.51	0.1570	0.3413
Earnings^c
Control group	2.40	2.26	0.5998	2.18	0.4307	2.45	0.8720
Partial intervention	2.14	2.14	0.6773	2.18	0.4300	2.17	0.9578	0.0793
Full intervention	2.01	2.25	0.2471	2.37	0.0857	2.43	0.2853	0.8769

^a All values are least squares means. Statistically significant p values at p < 0.10 are shown in bold

^b Group difference at month 15. The excluded category is the control group

^c The mixed effects model was estimated in the log of earnings, but the least squares estimates of earnings are reported for readability. The p values reported in this section are from the log of earnings mixed effects analysis