A cross-sectional study of the relationship between parents’ and children’s physical activity

BMC Public Health

Table 3 Tests of potential moderators of the parent–child physical activity relationship as measured by pedometers

Potential moderators	Moderator categories	Bivariate correlations		Test of moderation^a
		n	r	B (SE)^a	r _partial	p-value
Child gender	Girls	289	.30***	−0.12 (.09)	-.06	.20
Child gender	Boys	323	.23***	−0.12 (.09)	-.06	.20
Parent gender	Fathers	97	.27*	0.5 (.11)	.02	.64
Parent gender	Mothers	515	.25***	0.5 (.11)	.02	.64
Gender homogeneity^b	Same gender	326	.22***	−0.07 (.09)	-.04	.44
Gender homogeneity^b	Opposite gender	286	.30***	−0.07 (.09)	-.04	.44
Child weight status	Non-overweight	476	.24***	.11 (.11)	.04	.31
Child weight status	Overweight/obese	136	.29**	.11 (.11)	.04	.31
Parent weight status	Non-overweight	339	.28***	−0.03 (.10)	-.01	.79
Parent weight status	Overweight/obese	273	.20**	−0.03 (.10)	-.01	.79
Weight status homogeneity^c	Same weight status	372	.30***	0.05 (.10)	.02	.63
Weight status homogeneity^c	Opposite weight status	240	.17*	0.05 (.10)	.02	.63
Household income	Low (<$80,000/year)	482	.28***	0.25 (.14)	.09	.07
Household income	High (>$80,000/year)	137	.13	0.25 (.14)	.09	.07
Parent education	No graduate degree	526	.24***	0.38 (.16)	.11	.02
Parent education	Graduate degree	86	.33**	0.38 (.16)	.11	.02
Area-level SES^d	Low (< mean)	336	.22***	0.00 (.00)	.07	.14
Area-level SES^d	High (> mean)	276	.29***	0.00 (.00)	.07	.14

Note. *p < .05, **p < .01, ***p < .001; SES = socioeconomic status; ^apotential moderators were tested one by one by adding an interaction term (parent steps*potential moderator) to a linear regression model where child steps was the criterion variable and parent steps, child and parent gender, child and parent weight status, education, household income, and area-level SES were predictor variables; ^bgender homogeneity was also added to the model as a predictor variable; ^cweight status homogeneity was also added to the model as a predictor variable; ^darea-level SES was tested as a continuous variable

ISSN: 1471-2458