Assessing ADHD in college students: Integrating multiple evidence sources with symptom and performance validity data.

Prior research supports the use of multiple types of evidence from multiple sources when assessing ADHD in adults. However, limited research has examined how to best integrate the resulting set of data into a well-supported diagnostic conclusion. Moreover, clients sometimes overreport symptoms or display low effort on performance tasks, further complicating the interpretation of assessment data. The present study examined self-ratings and observer (e.g., parent) ratings of symptoms as well as performance task data from 514 postsecondary students assessed for ADHD at a university-affiliated clinic. Observer ratings were more reliable than self-ratings and were more likely to be corroborated by other data. The 2 types of ratings showed moderate to large relationships with each other as continuous variables (.32 < r < .52) while agreement around categorical symptom cutoffs was slight or fair (.12 < kappa < .32). Both types of ratings showed only small relationships with a performance test designed to assess ADHD symptoms. Approximately half of the cases in the sample had at least 1 piece of potentially noncredible data (suggesting potential symptom overreporting, inconsistent responding, or inadequate effort). Requiring ratings from multiple informants (as opposed to a single informant) of clinically significant symptoms for a diagnosis substantially reduced the effect of noncredible data, while also reducing the number of diagnoses by approximately half. Implications of these and other findings for practice and future research are discussed.