. 2015 Apr 1;11(2):139–147. doi: 10.1089/chi.2014.0110

Table 3.

Logistic Regression Predicting Child TV Watching Behavior on Weekdays by Parental Control and Parental Nurturance, with Parental Efficacy To Restrict Screen Viewing as a Potential Mediator

Parental control	Adjusted^a (with clustering)
Step 1: Outcome=child TV on weekdays^b	OR	95% CI	p
Parental control (C)	0.74	0.58–0.93	0.009
	Pseudo-R², 0.118; p<0.001
Step 2a: Predictor: parental control	Coeff	95% CI	p
(a) Outcome: efficacy to influence screen viewing (A1)	0.25	0.14–0.36	<0.001
	R², 0.033; p<0.001
Step 2b: Mediator on outcome	OR	95% CI	p
Efficacy to influence screen viewing (B)	0.75	0.68–0.83	<0.001
	Pseudo-R², 0.143; p<0.001
Step 3: Outcome=child TV on weekdays^b	OR	95% CI	p
Parental control (C’)	0.80	0.64–1.00	0.056
Efficacy to influence screen viewing	0.77	0.69–0.85	<0.001
	R², 0.150; p<0.001
Mediation statistics:		Bias-corrected 95% CI
Indirect effect	−0.04	−0.06 to −0.02
Proportion of total effect mediated	0.23

Parental nurturance	Adjusted^a (with clustering)
Step 1: Outcome=child TV on weekdays^b	OR	95% CI	p
Parental control (C)	0.95	0.90–1.02	0.139
	Pseudo-R², 0.109; p<0.001
Step 2a: Predictor: parental nurturance	Coeff	95% CI	p
(a) Outcome: efficacy to influence screen viewing (A1)	0.12	0.09–0.15	<0.001
	R², 0.085; p<0.001
Step 2b: Mediator on outcome	OR	95% CI	p
Efficacy to influence screen viewing (B)	0.75	0.68–0.83	<0.001
	Pseudo-R², 0.143; p<0.001
Step 3: Outcome=child TV on weekdays^b	OR	95%CI	p
Parental nurturance (C’)	0.99	0.93–1.05	0.672
Efficacy to influence screen viewing	0.76	0.68–0.84	<0.001
	R², 0.143; p<0.001
Mediation statistics:		Bias-corrected 95% CI
Indirect effect	−0.07	−0.10 to −0.04
Proportion of total effect mediated	0.71

^{^a}

Adjusted for child BMI z-score, IMD, and parental weekday TV viewing.

^{^b}

>2 hours versus 2 hours or less.

TV, television; IMD, index of multiple deprivation; OR, odds ratio; Coeff, coefficient; CI, confidence interval.