Abstract
Gender's role in self-control measures has been largely neglected. Although studies show that males have lower self-control than females, rarely have researchers questioned whether items used to measure self-control should be used for both groups. This study uses a Rasch rating scale analysis to assess item functioning of Grasmick et al.'s 24-item self-control scale for males and females. Using a sample of young adults, results indicate that 33% of the scale items showed differential functioning or item bias; that is, after controlling for self-control, females found one third of the items to be either more or less agreeable than males. Once biased items were removed from the scale, males, on average, still had lower self-control than females. In addition, after excluding biased items from the scale, the effect of self-control on criminal behavior and other outcomes was similar to the effect found with the full 24-item scale. Suggestions for future research on Grasmick et al.'s self-control scale are offered, and limitations of the current study are discussed.