Table 8 Regression models estimating mechanisms as a function of paternal incarceration in early childhood

From: Chains of Adversity: The Time-Varying Consequences of Paternal Incarceration for Adolescent Behavior

  Model 1 Model 2
Outcome variable b (SE)   b (SE)  
Mother and father separated 1.538 (.103) *** 0.968 (.126) ***
Mother relationship quality − 0.818 (.058) *** − 0.329 (.056) ***
Father relationship quality − 0.655 (.064) *** − 0.270 (.065) ***
Mother in poverty 0.780 (.084) *** 0.213 (.107) *
Father in poverty 1.034 (.084) *** 0.513 (.099) ***
Mother material hardship 0.615 (.076) *** 0.104 (.080)  
Father material hardship 0.695 (.101) *** 0.223 (.107) *
Mother engagement 0.079 (.024) *** 0.028 (.026)  
Father engagement − 0.604 (.059) *** − 0.314 (.062) ***
Mother shared responsibility in parenting − 0.777 (.046) *** − 0.360 (.044) ***
Mother cooperation in parenting − 0.692 (.047) *** − 0.342 (.045) ***
Mother depression 0.431 (.101) *** 0.125 (.122)  
Father depression 0.504 (.131) *** 0.097 (.160)  
Mother perceived social support − 0.674 (.073) *** − 0.167 (.066) *
Father perceived social support − 0.681 (.085) *** − 0.109 (.078)  
Mother residential move 0.692 (.088) *** 0.294 (.101) **
Father residential move 0.949 (.106) *** 0.469 (.124) ***
Mother neighborhood disadvantage 0.886 (.125) *** − 0.010 (.115)  
Father neighborhood disadvantage 0.936 (.140) *** − 0.045 (.128)  
  1. Each row represents a separate dependent variable. Ordinary least squares (OLS) regression models estimate relationship quality, hardship, engagement, shared responsibility in parenting, cooperation in parenting, perceived social support, and neighborhood disadvantage. Logistic regression models estimate separation, poverty, depression, and residential move. Model 1 is unadjusted. Model 2 adjusts for all time-invariant and time-varying control variables. Coefficients and standard errors are pooled across multiply imputed data sets. * p < .05, ** p < .01, *** p < .001