Skip to main content
. 2018 Aug 6;6:217. doi: 10.3389/fped.2018.00217

Table 4.

Mediation effect of caregiver-child interaction on the association between family childcare types and child conduct problem behaviors.

Characteristics Conduct problem behaviors (Dependent variable, Y) Mediation Effect
Model 2: X → Y Model 3: X+M → Y Relative indirect effect (95% CI) Proportion of mediation, %
Coefficient (95% CI) Coefficient (95% CI)
Y TAKEN AS A CATEGORICAL VARIABLE
Family childcare types (Independent Variable, X)
Parents together (Ref) Ref Ref
Mother alone (D1) 2.53 (1.27–5.04)** 2.36 (1.18, 4.72)* 0.055 (0.014, 0.093) 8.9%
Mother with others (D2) 2.55 (1.52–4.27)*** 2.52 (1.50, 4.22)*** 0.019 (−0.005, 0.039) NA
Grandparents (D3) 2.18 (1.01–4.71)* 1.86 (0.85, 4.04) 0.133 (0.064, 0.192) 23.4%
Changing caregivers (D4) 2.33 (1.47–3.69)*** 2.22 (1.40, 3.51)** 0.053 (0.025, 0.077) 9.0%
Caregiver-child interaction (Mediator, M) 0.45 (0.34, 0.60)***
Y TAKEN AS A CATEGORICAL VARIABLE:
Family childcare types (Independent Variable, X)
Parents together (Ref) Ref Ref
Mother alone (D1) 0.059 (0.019, 0.099)** 0.051 (0.011, 0.091)* 0.008 (0.003, 0.014) 13.7%
Mother with others (D2) 0.063 (0.037, 0.088)*** 0.060 (0.034, 0.085)*** 0.003 (−0.0002, 0.006) NA
Grandparents (D3) 0.043 (0.002, 0.085)* 0.024 (−0.018, 0.065) 0.019 (0.013, 0.027) 44.9%
Changing caregivers (D4) 0.073 (0.051, 0.094)*** 0.065 (0.044, 0.086)*** 0.008 (0.005, 0.011) 10.8%
Caregiver-child interaction (Mediator, M) −0.116 (−0.136, −0.096)***

Model was adjusted for confounders including child age, gender, parental education level, parental age at pregnancy, marital status, and family income.

‡Relative indirect effect represented the indirect effect on Y via M of being in group Di relative to the reference group.

*

P < 0.05,

**

P < 0.01,

***

P < 0.001; Bold fonts denote statistical significance at 0.05 level. NA: The indirect effect was not statistically significant.