Skip to main content
. 2023 Mar 13;11:992539. doi: 10.3389/fpubh.2023.992539

Table 5.

Propensity score matching estimation of the effects of living without parents at the ages of 0–6/7–12/13–15/16–18 on individual physical and mental health in adulthood, China, 2018.

Method Nearest neighbor Radius Kernel Local linear regression
0–6 years Physical health (ATT) −0.215*** −0.201*** −0.192*** −0.203***
(−3.92) (−4.04) (−3.89) (−3.19)
Mental health (ATT) 0.264*** 0.276*** 0.277*** 0.278***
(4.25) (4.79) (4.84) (3.97)
7–12 years Physical health (ATT) −0.145*** −0.151*** −0.142*** −0.152***
(−2.75) (−3.15) (−2.99) (−2.46)
Mental health (ATT) 0.259*** 0.234*** 0.233*** 0.234***
(4.56) (4.41) (4.42) (3.59)
13–15 years Physical health (ATT) −0.132*** −0.112*** −0.110*** −0.118**
(−2.89) (−2.70) (−2.67) (−2.18)
Mental health (ATT) 0.174*** 0.164*** 0.163*** 0.170***
(3.56) (3.59) (3.61) (3.06)
16–18 years Physical health (ATT) −0.132*** −0.112*** −0.110*** −0.118**
(−2.89) (−2.70) (−2.67) (−2.18)
Mental health (ATT) 0.174*** 0.164*** 0.163*** 0.170***
(3.56) (3.59) (3.61) (3.06)

ATT means the average treatment effect on treatment. T-statistics are reported in parentheses. The element number of the nearest-neighbor matching with a caliper was 4, the radius was set to 0.01 in radius matching, and kernel matching and local-linear matching used default kernels and bandwidth.

***

p < 0.01,

**

p < 0.05.