Abstract
The recent attempt to introduce hypersexual disorder into the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), has increased empirical scrutiny of the construct. Consensus on its definition and underlying structure remains elusive. Whereas some conceptualizations favor a categorical latent structure, others speculate that hypersexual behavior is dimensional. Research on the latent structure of hypersexual behavior, however, has been sparse. This is unfortunate, because determination of the latent structure can contribute to more accurate assessment, diagnosis, and understanding of etiological process. To date the only study on hypersexuality's latent structure found consistent evidence of a dimensional structure for males but less clear results for females. In the present study the Multidimensional Inventory of Development, Sex, and Aggression (MIDSA), a self-report, contingency-based inventory, was administered to 1,146 college students. Four indices of hypersexual behavior and six indices of sexual compulsivity were analyzed, using three taxometric methods (mean above minus below a cut [MAMBAC], maximum covariance [MAXCOV], and latent mode factor analysis [L-Mode]). Evidence supported a dimensional latent structure for hypersexuality in male and female samples. Future assessments of hypersexuality must focus on adequate reliability and discriminant validity across the continuum of sexual behavior rather than on attempts to differentiate between arbitrarily developed diagnostic categorizations.