Item response theory and computer adaptive testing of the sexual knowledge scale of the sexual knowledge and attitude test in a college sample

Abstract

ABSTRACT Our study addresses the limited availability of well-validated scales for assessing general sexual knowledge among college populations. To fill this gap, we examined the psychometric properties of a 41-item sexual knowledge scale derived from the Sexual Knowledge and Attitudes Test – Adolescents (SKAT-A) in a sample of young adults aged 18–25 (N = 1,291). We employed classical test theory (CTT) procedures, followed by item response theory (IRT) and computerized adaptive testing (CAT), to refine the SKAT-A knowledge scale. Both CTT and IRT analyses identified six items for removal due to poor discrimination and difficulty parameters. A confirmatory factor analysis supported a well-defined unidimensional latent trait of sexual knowledge. The results of CAT simulations using dynamic item administration demonstrated the scale’s measurement precision, with moderate test reliability and relatively low standard errors. On average, there was a 54.3% reduction in the number of items administered without compromising scale reliability. This study concludes that the SKAT-A efficiently assesses a unidimensional trait of sexual knowledge in college-attending young adults. It highlights that only a subset of items from the full test bank is necessary to achieve this, providing a practical and reliable tool for assessing sexual knowledge in this population.