Table 3.

Regression Models Predicting Participation in the Intervention Program, Moderated by Pretest Levels of Observed Couple Negative Communication

Variable	Coparenting negativity			Relationship quality			Parenting positivity			Parenting negativity			Anxiety			Night wakings
	B	95% CI	p	B	95% CI	p	B	95% CI	p	B	95% CI	p	B	95% CI	p	B	95% CI	p
Condition	−0.19	[−0.28, −0.10]	<.001	0.35	[0.14, 0.56]	.001	0.13	[0.00, 0.26]	.058	−0.19	[−0.28, −0.10]	<.001	−0.38	[−1.20, 0.44]	.364	−0.26	[−0.59, 0.06]	.112
Negative Communication	0.14	[0.06, 0.21]	.001	−0.32	[−0.55, −0.09]	.006	−0.20	[−0.31, −0.09]	.001	0.14	[0.06, 0.21]	.001	0.88	[−0.03, 1.78]	.057	0.16	[−0.05, 0.37]	.135
Condition × Negative Communication	−0.13	[−0.23, −0.02]	.016	0.33	[0.02, 0.65]	.038	0.24	[0.08, 0.41]	.004	−0.13	[−0.23, −0.02]	.016	−1.95	[−3.10, −0.79]	.001	−0.45	[−0.85, −0.06]	.025

Note. CI = confidence interval.

All models include controls for site location and pretest characteristics of informant: family income, economic strain, age, child age (for wakings) and social desirability (for parent report). Multilevel modeling was used if data were available from each parent. The moderator was centered before analysis to aid in coefficient interpretation.