Access Type

Open Access Dissertation

Date of Award

January 2012

Degree Type


Degree Name




First Advisor

Mary C. Sengstock


Prisoner recidivism has and continues to impact families and communities. Traditional methods aimed at reducing this phenomenon have had little success in curtailing this problem. One obvious but often overlooked tool that may play a significant role in dealing with this issue is the importance of family relationships. This dissertation quantitatively examines offender's perceptions of the importance of family relations, specifically the relationships with the offender's children, spouse or significant other. These relationships are analyzed to determine their level of impact on prison misconduct and parole recidivism.

Response data from 102 male ex-offenders from the years of 2009 to 2010 are used to test the originating question of this dissertation: To what extent are the relationships between offenders and their families related to prison misconduct or recidivism? Variations in perceptions towards familial importance are ascertained via sub-group analyses. These subgroups analyses use demographic/personal factors (age, race, marital status, educational attainment prior to incarceration, educational attainment during incarceration), and family background/structural factors (residence of children before incarceration and during parole supervision, strength of bond with children and spouse or significant other prior to and during incarceration, and while under parole supervision).

This analysis found that differences in prison misconducts does exist between offenders who had children prior to incarceration, or were married or in a relationship during incarceration and those offenders who did not have children or were not in a relationship during this time period. The results show that offender's who have children, are married or in a relationship are less likely to engage in prison misconduct. However, these relationships are not significant predictors of recidivism.

Some linear regression models indicate that factors or variables such as age, educational attainment prior to prison, composite measures of the quality of the relationship with spouse or significant other help predict prison misconduct, but only the variable of educational attainment during incarceration helps to predict parole recidivism.