Validating a General Maltreatment Measure for Use with Incarcerated Youth

Background: Maltreatment is a risk factor for sexual and non-sexual delinquency. However, delinquency research tends to utilize maltreatment measures that were developed with community samples. More needs to be done to establish the utility and factor structure of existing measures with incarcerated youth to determine their validity. The goal of the present study was to determine whether a validated maltreatment measure (Maltreatment Abuse Chronology of Exposure; MACE; Teicher & Parigger, 2015) retained the factor structure of the distinct maltreatment type subscales. Methods: Data were collected via paper and pencil surveys from 136 incarcerated youth at 7 residential treatment and community corrections facilities in a Midwestern state. Exploratory factor analysis was used to explore whether the validated factor structure would be upheld with an incarcerated sample. In the original validation with a maltreated community sample, six subscales were found: 1) Emotional Neglect, 2) Non-Verbal Emotional Abuse, 3) Parental Physical Maltreatment, 4) Parental Verbal Abuse, 5) Physical Neglect, and 6) Sexual Abuse. The MACE items were scored on a 5-point Likert scale with response options 0=Never 1=Rarely 2= Sometimes 3=Often 4=Very Often, and included items like Someone in your family hit, beat, pushed, grabbed, shoved, slapped, pinched, punched, or kicked you. Eigenvalue size and magnitude of variance explained were used to compare potential number of factors when all 48 scale items were free to load. Promax rotation and maximum likelihood extraction were used in accordance with best practice for ordinal data when factors are permitted to correlate. In the next iteration, 4- and 5-factor models were tested against one another for goodness of fit, percent of variance explained, theoretical consistency, and parsimony. Results: The 4-factor model explained a cumulative percent of 55.314% of variance in responses and was accepted as the best representation of MACE scale performance with incarcerated youth. The 4 factors that emerged were: 1) Physical/Emotional Abuse (15 items); 2) Sexual Abuse (8 items); 3) Neglect/Household Dysfunction (15 items); and 4) Family Emotional Dynamics (10 items). Using the new factors, binary logistic regression analyses were run to determine whether any of the factors were associated with sexual vs. nonsexual delinquency outcomes. Findings revealed that youth with higher scores on the Physical/Emotional Abuse and Sexual Abuse subscales had greater odds of disclosing a sexual offense (B = 1.061, p = .021 and, B = 1.083, p = .021 respectively). Sexually vs. non-sexually offending youth did not differ Neglect/Household Dysfunction or Family Emotional Dynamics. Conclusions: The current study reveals the importance of validating community measures with incarcerated youth samples prior to utilizing existing subscales, as these youth are distinct from community and maltreated samples. Future research should prioritize both the use of measures developed with sexual and non-sexual offending youth populations and set the expectation that validation analyses be conducted prior to relying on measures developed for populations distinct from incarcerated samples.

Teicher, M. H., & Parigger, A. (2015). The ‘Maltreatment and Abuse Chronology of Exposure’(MACE) scale for the retrospective assessment of abuse and neglect during development. PLoS One, 10(2), e0117423.