Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purposes: Grasmick et al. (1993) developed a scale to test the assumption made by Gottfredson and Hirschi (1990) that self-control is a one-dimensional trait. The scale contains total of 24 items, four questions form each of the six categories (impulsivity, simple task, risk seeking, physical activities, self-centeredness, and temper.). Although reliability and validity of this scale have been well established, there are mixed findings about construct validity – factorial validity – and invariance across social categories such as gender, age, and race/ethnicity. The purposes of this study are 1) to test whether the six dimensions identified in the theory are concurrent indicators of Low Self-Control across ethnic groups and to confirm it as a significant predictive variable to juvenile delinquency.

Methods: This study utilized data from Center for the Study of Crime, Delinquency, and Social Control at the University of Oklahoma, which conducted the ‘High School Delinquency Survey' with 1724 participants consisting of three ethnic groups: Native Americans (n=382), Caucasians (n=1,122), and African Americans (n=220). Using the LISREL 8.54, a series of analyses tested first the original six-factor measurement model and then full model including two ascribed demographic controls (age and race) in an SEM with one dependent variable: self-reported juvenile delinquent behaviors (such as interpersonal offense, property damage, and vandalism).

Results: From preliminary analyses, it had good reliability on overall scale (alpha =.86) along with acceptable level on subscales (alphas = .73 ~ .80). Consistent with prior investigations, the current confirmatory factor analyses indicated that the six-factor model was the most accurate model (÷˛/df =1436.77/237, RMSEA =.056, CFI=.91, IFI=.91, GFI=.93) in comparison with alternative models such as one-factor model (÷˛/df =7843/252, RMSEA=.13, GFI=72) and second-order model (÷˛/df =1639.68/246, RMSEA=.059, GFI=92). It was also found that both Native and African Americans are more likely to have lower self-control than Caucasians and there is a significant interaction effect between age and race on low self-control. Multiple group invariance tests revealed that Native American model had better goodness of fit than other two groups.

Discussion: The study findings confirmed that the scale has six distinct dimensions and suggest that the “Temper” dimension of the scale is most useful in predicting juvenile delinquent behaviors. Because several items exhibited bias across race groups, it would benefit from more precise and culturally sensitive wording to discriminate and better reflect each dimension. Taken together, this study provides an empirical evidence of the robustness of self-control in relation to form of delinquency. Implications of the findings are discussed in terms of future theory and research.

Reference: Grasmick, H.G., Tittle, C.R., Bursik, R.J., & Arneklev, B.J. (1993). Testing the core empirical implications of Gotteredson and Hirschi's general theory of crime. Journal of Research in Crime and Delinquency, 30(1), 5-29