Table 3.

Multilevel Lagged Models Predicting Self-Reported Daily Number of Physical Health Symptoms (N = 114)

Variable	Model		Model with CGG Interaction
	Coefficient (SE)	Effect Size	Coefficient (SE)	Effect Size
Within-Person Level 1
Intercept	2.93 (0.43)**	.43	2.87 (0.49)**	.42
Day	−0.04 (0.03)	.09	−0.03 (0.03)	.09
Previous-Day # Child Behavior Problem	0.02 (0.05)	.03	0.02 (0.06)	.02
Previous-Day # Stressors	0.05 (0.07)	.05	0.03 (0.08)	.02
Previous-Day Emotional Support	−0.20 (0.11)+	.16	−0.32 (0.17)*	.20
X CGG Linear	----------	--	0.28 (0.19)	.12
X CGG Curvilinear	----------	--	−0.38 (0.18)*	.19
Between-Person Level 2
Maternal Age	0.02 (0.02)	.09	0.02 (0.02)	.11
Maternal Education	−0.30 (0.11)*	.27	−0.30 (0.10)**	.27
Additional Children with Disability	0.03 (0.26)	.01	0.02 (0.24)	.01
Partner Status	−1.36 (0.52)*	.25	1.36 (0.54)*	.24
Mean # Child Behavior Problem	0.61 (0.18)**	.32	0.60 (0.18)*	.31
Mean # Daily Stressors	0.83 (0.39)*	.21	0.81 (0.38)*	.21
Mean Emotional Support	−0.55 (0.77)	.07	−0.52 (0.68)	.09
CGG linear	0.25 (0.21)	.12	0.28 (0.20)	.14
CGG curvilinear	−0.03 (0.10)	.03	−0.03 (0.06)	.06

Note. Effect size was calculated with the following equation: r = sqrt [t²/(t² + df)]; small effect: r >.10, medium effect: r > .24 and large effect r >.37 (Kirk, 1996). df = degrees of freedom ranged from 105 to113. CGG curvilinear was computed as the square of centered CGG. We re-ran models to also include household income (coded as 1 to 14, starting at < $10K and going up by $10 to $20K intervals to $160K+) in Level 2. The pattern of effects remained the same. CGG = cytosine-guanine-guanine.

⁺p ≤ . 09.

^*p ≤ .05.

^**p ≤ .01.