Maternal Parenting Stress Following Paternal or Close Family Incarceration: Bayesian Model-Based Profiling Using the HILDA Longitudinal Survey

Abstract

Objectives To: (1) examine the existence and extent of heterogeneity in efects of paternal incarceration (PI) or close family incarceration (CFI) on maternal parenting stress; (2) identify key variables related to the efect of PI and CFI on maternal parenting stress. Methods Using data from HILDA, an Australian longitudinal survey, we investigate changes in maternal parenting stress for mothers who experienced either PI or CFI. There were 15 demographic and stress-related explanatory variables. Using Bayesian profle regression, we examine the average changes in maternal parenting stress after incarceration compared to the mother’s average level of parenting stress in prior waves, simultaneously with model-based clustering to characterise the profles of mothers having a diferent degree of change. Results Three profles of mothers were identifed: (1) A small decrease in parental stress levels (n=112); (2) No measurable average change in parental stress levels (n=46); (3) A small increase in parental stress levels (n=117). Only for the second cluster did the 95% posterior credible intervals for the means include zero as a plausible value. The estimated means for clusters 1 (decrease) and 3 (increase) did not overlap and are clearly separated. Conclusions Neither PI nor CFI helped profle mothers. Thus, research should examine wider family incarceration efects on children and caregivers. Prior adversity, wellbeing and family demographics contributed to the cluster profles. Parenting stress is heterogeneous and improved methods are needed to disentangle the efects of incarceration from other contextual, recent and cumulative adverse events in people’s lives.