To date, the vast majority of research regarding personality in IO Psychology has relied on self-report assessments. Despite support for the utility of third-party assessments, IO Psychologists have only just begun extensive research in this area. Connelly and Ones (2010) conducted a meta-analysis that demonstrated that accuracy of third-party ratings improved as intimacy between the judge and the target grew. This remained true with the exception of predicting behavioral criteria, where non-intimates maintained superior predictability (Connelly & Ones, 2010). This was later contradicted by a recent investigation that found the best predictive validity for third-party assessments when they are taken from personal acquaintances as opposed to work colleagues (Connelly & Hulsheger, 2012). The current study is intended to investigate how the depth of the relationship and breadth of behavioral observations differentially moderate the relationship between third-party personality assessments and accuracy criteria (i.e., self-other overlap, discriminant validity and behavior). Results indicate that both depth and breadth impact accuracy criteria and they do so differentially based on trait visibility and evaluativeness. These findings will be discussed along with practical implications and limitations of the following research.

No description available.