Publication Date



Background: The presence of callous-unemotional (CU) traits designates an important subgroup of antisocial youth at risk for severe, persistent, and impairing conduct problems. As a result, the fifth revision of the Diagnostic and Statistical Manual includes a specifier for youth meeting diagnostic criteria for Conduct Disorder who show elevated CU traits. The current study evaluated the DSM-5 criteria using Item Response Theory (IRT) analyses and evaluated two methods for using a self-report measure of CU traits to make this diagnosis. Methods: The sample included 2257 adolescent (M age = 15.64, SD = 1.69 years) boys (53%) and girls (47%) from community and incarcerated settings in the United States and the European countries of Belgium, Germany, and Cyprus. Results: IRT analyses suggested that four- or eight-item sets from the self-report measure (comparable to the symptoms used by the DSM-5 specifier) provided good model fit, suggesting that they assess a single underlying CU construct. Further, the most stringent method of scoring the self-report scale (i.e. taking only the most extreme responses) to approximate symptom presence provided the best discrimination in IRT analyses, showed reasonable prevalence rates of the specifier, and designated community adolescents who were highly antisocial, whereas the less stringent method best discriminated detained youth. Conclusions: Refined self-report scales developed on the basis of IRT findings provided good assessments of most of the symptoms used in the DSM-5 criteria. These scales may be used as one component of a multimethod assessment of the ‘With Limited Prosocial Emotions’ specifier for Conduct Disorder.


Institute for Learning Sciences and Teacher Education

Document Type

Journal Article

Access Rights

ERA Access

Access may be restricted.