“…In order to improve the FIML estimation's accuracy, eight auxiliary variables were included in the regression models as potential correlates of missing data: 1) child's SSBD rank, 2) child's sex, 3) Spanish-speaking parent, 4) current marital status, 4) parent's education level, 5) estimated annual household income, 6) estimated annual household income, 7) number of children in the household, and 8) parental distress as reported on the Parenting Stress Index -Short Form (PSI-SF; Abidin, 1995). Since there was a higher rate of missing data from parents, researchers included auxiliary variables in the models which have demonstrated to be predictive of subsequent dropout from the study (Beauchaine, Webster-Stratton, & Reid, 2005;Herman et al, 2012;Reinke et al, 2012) and which indicate higher levels of stress in families or might be considered as potential barriers between families and research staff (e.g., Spanish-speaking participants). Inclusion of these variables is recommended as part of an overall analysis strategy since they increase statistical power, reduce bias, and improve the plausibility that the data were missing at random without altering the interpretation of parameter estimates (Collins, Schafer, & Kam, 2001;Enders, 2010).…”