Abstract
This study examined the item functioning of the Inventory of Callous-Unemotional Traits (ICU) in an ethnically diverse sample 1,190 of first-time justice-involved adolescents (mean age = 15.28 years, SD = 1.29). On elimination of 2 items, the total ICU score provided a reliable (internally consistent and stable) and valid (correlated with and predictive of measures of empathy, school conduct problems, delinquency, and aggression) continuous measure of callous and unemotional (CU) traits. A shortened, 10-item version of the total scale, developed from item response theory (IRT) analyses, appeared to show psychometric properties similar to those of the full ICU and, thus, could be used as an abbreviated measure of CU traits. Finally, item analyses and tests of validity suggested that the factor structure of the ICU reported in a large number of past studies could reflect method variance related to the ICU, including equal numbers of positively and negatively worded items. Specifically, positively worded items (i.e., items for which higher ratings are indicative of higher levels of CU traits) were more likely to be rated in the lower response categories, showed higher difficulty levels in IRT analyses (i.e., discriminated best at higher levels of CU traits), and were more highly correlated with measures of antisocial and aggressive behavior. On the basis of these findings, we recommend using the total ICU as a continuous measure of CU traits and do not recommend continued use of the subscale structure that has been reported in multiple past studies.