Skip to main content
. Author manuscript; available in PMC: 2020 Jan 1.
Published in final edited form as: Soc Sci Res. 2018 Sep 17;77:130–147. doi: 10.1016/j.ssresearch.2018.09.002

Table 3.

Regression Analysis of Child’s BPI by Migration Status and Other Control Variables (Standard Errors in Parentheses).

Internalizing BPI Externalizing BPI

Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4
Migration status (ref. rural local, both parents)
 Urban local, both parents −0.231 (0.214) 0.009 (0.240) −0.079 (0.240) −0.006 (0.224) −0.012 (0.371) 0.674 (0.409) 0.537 (0.405) 0.337 (0.403)
 Rural-urban migrant, both parents 0.374 (0.387) 0.502 (0.366) 0.358 (0.353) 0.311 (0.341) 0.645 (0.525) 0.855 (0.539) 0.653 (0.524) 0.512 (0.519)
 Rural-urban migrant, absent parent 0.103 (0.497) 0.391 (0.493) 0.156 (0.469) −0.026 (0.436) 0.578 (1.032) 0.843 (1.018) 0.483 (1.020) 0.272 (0.960)
 Left behind, father away −0.130 (0.236) −0.106 (0.244) −0.292 (0.232) −0.319 (0.223) 0.160 (0.457) 0.095 (0.447) −0.158 (0.428) −0.189 (0.414)
 Left behind, mother away −0.039 (0.431) −0.052 (0.475) −0.121 (0.439) −0.217 (0.417) 0.567 (0.901) 0.796 (0.906) 0.697 (0.895) 0.642 (0.934)
 Left behind, both away 0.484* (0.234) 0.829** (0.291) 0.577* (0.271) 0.177 (0.279) 0.785* (0.370) 1.231* (0.486) 0.865 (0.497) 0.486 (0.501)
Child is male 0.011 (0.146) −0.013 (0.146) −0.091 (0.142) 0.248 (0.242) 0.219 (0.242) 0.158 (0.238)
Child’s age −0.055 (0.107) −0.063 (0.105) −0.069 (0.099) −0.730*** (0.166) −0 747*** (0.166) −0.735*** (0.163)
Children’s age squared 0.005 (0.006) 0.005 (0.006) 0.005 (0.006) 0.025* (0.010) 0.025** (0.010) 0.024* (0.009)
PCG is male −0.092 (0.195) −0.101 (0.192) −0.018 (0.195) −0.077 (0.306) −0.070 (0.302) −0.024 (0.316)
PCG’s age −0.007 (0.008) −0.010 (0.008) −0.017* (0.008) −0.022 (0.014) −0.025 (0.014) −0.026 (0.015)
Sibling (ref. no sibling)
 Child has one sibling 0.196 (0.166) 0.153 (0.161) 0.041 (0.157) 0.799** (0.277) 0.764** (0.268) 0.703** (0.268)
 Child has two or more siblings 0.416 (0.328) 0.162 (0.320) −0.250 (0.287) 0.855 (0.460) 0.525 (0.456) 0.207 (0.439)
Child is minority −0.248 (0.300) −0.313 (0.276) −0.357 (0.244) −0.768 (0.509) −0.941 (0.491) −0.983* (0.445)
Family income quartiles (ref. bottom 25%)
 Lower 25% −0.379* (0.180) −0.273 (0.178) −0.206 (0.173) −0.736* (0.334) −0.580 (0.332) −0.541 (0.332)
 Upper 25% −0.457 (0.248) −0.298 (0.243) −0.121 (0.232) −0.910* (0.359) −0.679 (0.351) −0.561 (0.348)
 Top 25% −0.467* (0.232) −0.237 (0.224) 0.039 (0.224) −1.230** (0.391) −0.877* (0.381) −0.679 (0.388)
Region (ref. North/Northeast)
 East 0.206 (0.226) −0.130 (0.217) −0.052 (0.436) −0.413 (0.429)
 South-Central 0.835** (0.253) 0.587* (0.237) 0.924* (0.456) 0.583 (0.441)
 West 1.276*** (0.264) 0.757** (0.266) 1 701*** (0.477) 1.167* (0.485)
PCG’s parenting practices −0.047*** (0.013) −0.030 (0.019)
PCG’s emotional distress 0.228*** (0.022) 0.304*** (0.033)
PCG level of education −0.011 (0.022) 0.049 (0.036)
Constant 14.988*** (0.127) 15.363*** (0.512) 14.933*** (0.509) 14.305*** (0.713) 24.361*** (0.220) 29.291*** (0.922) 28.875*** (0.949) 26.298*** (1.234)
N (number of children) 4,338 4,338 4,338 4,338 4,338 4,338 4,338 4,338
Change in coefficient (left behind by both parents vs. rural children with both parents) between current model and previous model 0.346 (0.184) −0.253*** (0.072) −0.400*** (0.089) 0.446 (0.332) −0.366** (0.117) −0.379** (0.123)
Proportion of total effect mediated by variables in current model (relative to previous model) −2.314 0.769 0.697 −1.641 0.785 0.448
Propensity score matching estimates (left behind by both parents vs. rural children with both parents) 2.090* (1.030) 0.613 (1.011) 3.884* (1.609) 1.033 (1.347)
*

p < .05.

**

p < .01.

***

p < .001.