The Add Health psychopathy scale: Assessing its construct validity

Abstract Objectives This study examined the construct validity of a measure of psychopathy derived from Add Health data. Methods Two recruitment sites were used, resulting in 446 undergraduate participants in sample 1, and 332 undergraduate participants in sample 2. In addition to the measure from the Add Health data, participants completed three well-validated self-report measures of psychopathy, a measure of general personality, and measures of externalizing behaviors (e.g., crime and aggression). Results Factor analyses failed to confirm a one-factor solution consistent with the original Add Health measure; instead, a five-factor solution fit the data best. The Add Health measure failed to demonstrate convergent and criterion validity. Alternate coding and subscales showed some improvement; however, content and construct validity remained lacking. Conclusions The Add Health measure is not a valid measure of psychopathy and its continued use is not recommended.