The Effect of Childhood Maltreatment on Adult Criminality: A Tobit Regression Analysis

• Published in: Child maltreatment Vol. 8; no. 2; pp. 129 - 137
• Main Authors: Grogan-Kaylor, Andrew, Otis, Melanie D.
• Format: Journal Article
• Language: English
• Published: Thousand Oaks, CA SAGE Publications 01.05.2003; Sage; SAGE PUBLICATIONS, INC
• Subjects: Abused children; Adult; Adult and adolescent clinical studies; Adults; Age Factors; Antisocial behavior; Arrests; Biological and medical sciences; Child; Child Abuse; Child abuse & neglect; Child Abuse - psychology; Child Abuse - statistics & numerical data; Child Neglect; Childhood; Childhood Factors; Crime; Crime - psychology; Crime - statistics & numerical data; Criminality; Data Collection; Demography; Humans; Information; Juvenile Delinquency - statistics & numerical data; Medical sciences; Psychology. Psychoanalysis. Psychiatry; Psychopathology. Psychiatry; Race; Regression Analysis; Sexual abuse; Social behavior disorders. Criminal behavior. Delinquency; Time Factors; United States of America; Victimology; child maltreatment; tobit regression analysis; adult criminality predictors; Human; Child abuse; Risk factor; Prediction; Sex; Early traumatism; Regression analysis; Victimology; Social behavior disorder; Age; Antecedent; Criminology
• ISSN: 1077-5595; 1552-6119
• DOI: 10.1177/1077559502250810

This article reports on the results of an analysis of a data set containing information on 667 nonmaltreated and 908 maltreated children. The data also contain information on whether the study subjects were arrested in early adulthood. Because adult arrests are an imperfect and censored measure of antisocial behavior, tobit regression analysis was used to examine the effect of the subjects’ experiences of child maltreatment on later arrests while controlling for those subjects’ demographic characteristics. The analysis finds that children’s age, race, and sex and experiences of child neglect all have an impact on subsequent adult arrests. However, physical abuse and sexual abuse do not emerge as statistically significant predictors of arrests in this model. The study also illustrates a method for the decomposition of tobit coefficients to extract more information from them.