Revisiting the "Definitive" Test of Self-Control Theory

Laura J. Dugan, Georgia State University
Dean Dabney, Georgia State University
Brenda Sims-Blackwell, Georgia State University

Nearly a decade after its publication, Grasmick, Bursik & Tittle's (1993) article still stands among the hallmark empirical tests of Gottfredson & Hirschi's (1990) "general theory of crime." When referencing the theory, textbook authors and policy-makers alike draw heavily upon what is described by many as the "definitive" test of theory. in this paper, we subject this respected test of the theory to close criticism observing that the data were not well suited to the OLS regression analysis that was employed. Namely, we begin with the premise that a truncated distribution int he depe4ndent variable significantly jeopardized the validity of the findings. We access the same data set and enlist the same measures that were used in the 1993 article. However, our replication effort is built around the use of a more suitable Ordinal Logistic model. Our findings provide significantly less support for the theory's central assertion that the interactive effect of low self-control and criminal opportunity predict deviant outcomes. Theoretical implications of this replication are presented and a series of new conceptual directions are forwarded.

(Return to Program Resources)

Updated 05/20/2006