Attention-Deficit/Hyperactivity Disorder (ADHD) is a chronic disorder associated with long-term impairment across multiple life domains, including academic, occupational, social, and psychological (Barkley, Murphy, & Kwasnik, 1996; Harpin, 2005; Wilens, Biederman, & Spencer, 2002). Increasing numbers of college students are presenting to health centers and counseling programs with complaints of inattention, distractibility, and restlessness, underscoring the critical need for reliable and valid methods for evaluating ADHD in this age group (e.g., U.S. Government Accountability Office [GAO], 2009). There is a growing body of research suggesting that the Barkley Adult ADHD Rating Scale—IV (BAARS-IV; Barkley, 2011) is a reliable tool for assessing ADHD in adults (e.g., Becker, Marshall, & McBurnett, 2014), yet there are a number of limitations and omissions in existing data. The current study investigated the psychometric properties of the BAARS-IV in a clinic-referred sample of 607 college students between the ages of 18 and 25. First, the factor structure of ADHD was examined by comparing traditional two-factor and three-factor models of ADHD with bifactor models of ADHD using confirmatory factor analyses (CFA). The correlated three-factor model consisting of separate dimensions of inattention, hyperactivity, and impulsivity was considered the best representation of the ADHD constructs for this sample. Next, a multidimensional graded response model (GRM) based on item response theory (IRT) was applied to the data. The results of the GRM indicated that most items of the BAARS-IV showed adequate discrimination of their respective latent traits (i.e., Inattention, Hyperactivity, and Impulsivity) and functioned best when measuring participants with average levels of the latent traits. Of the 18 items, only seven items measure sub-clinical levels of their respective latent trait; however, two of these items provided relatively less information. The remaining five items (one hyperactivity and four impulsivity) had good discrimination and difficulty parameters, providing useful information at sub-clinical levels of the latent traits. The final aim of this study was to examine the items for potential differential item functioning (DIF) by gender and to estimate the size and impacts of detected DIF. The results revealed significant DIF for two inattention items, two hyperactivity items, and one impulsivity item. The findings from the current study suggest that some items of the BAARS-IV may have more clinical utility than the others in the assessment of ADHD in college students.
