Peer-reported personality problems of research nonparticipants: Are our samples biased?

Peer nominations of pathological personality traits were collected on 1442 freshmen participating in a study regarding personality and 283 students who initially provided consent to participate but failed to show up for the assessment. Ten peer-based personality disorder scales and eight IIP-64 scales were entered into two separate multiple logistic regression procedures to predict the probability of nonparticipation. There was a significantly higher probability of participation if peers nominated someone as having more histrionic, obsessive–compulsive, self-sacrificing, and intrusive/needy characteristics. Students were significantly less likely to participate if peers nominated them as being higher on narcissism or non-assertiveness. Results suggest it may be more difficult to obtain sufficient numbers of people high in narcissistic traits than individuals with other personality traits. Researchers may need to employ novel strategies to recruit individuals with narcissistic traits for experimental studies.