TABLE 2.

Effects of practice elements profiles and program elements on intervention outcomes: Standardized regression coefficients, for short (T1) and longer term (T2)

	Social contacts a		Parenting stress a		Internalizing problems a		Externalizing problems a
	T1	T2	T1	T2	T1	T2	T1	T2
	β	β	β	β	β	β	β	β
Practice elements profiles
Explorative/supportive
Combined	.062	−.064	.065	−.019	−.017	.031	−.087	.077
Action oriented	−.002	−.066	.075	.043	.003	.033	−.002	.093
Program elements
Number of visits	−.013	−.177**	.075	.071	.004	.057	.057	.057
Duration of visits	−.075	−.074	.062	.099	.019	.129	.015	.054
Telephone contacts	.145*	.146	−.113*	−.131**	−.027	−.188*	−.008	−.140
Intervision	.031	.091	−.072	.016	−.183*	−.042	−.123*	−.028
Supervision	−.158*	−.147	.032	−.051	−.062	−.004	−.088	−.018
Consultation	−.006	−.137*	−.047	−.007	.052	.069	−.020	.004

In these analyses we controlled for the propensity score and baseline scores on social contacts, parenting stress, internalizing, and externalizing problems

Profile 1 (explorative/supportive profile) was regarded as the reference group.

*p < .05, **p < .01, ***p < .001.

^a To make maximum use of the available data, we used the full information maximum likelihood (FIML) estimation procedure as implemented in Mplus to deal with missing data on the four outcome measures. Original sample sizes for social contacts were T0 = 349, T1 = 224, and T2 = 140; for parenting stress T0 = 382, T1 = 252, and T2 = 142; and for both internalizing and externalizing problems T0 = 276, T1 = 196, and T2 = 91.