An Assessment of Biodata Predictive Ability across Multiple Performance Criteria

Although selection tests are typically validated against only a single criterion, calls in the personnel selection literature urge the examination of selection devices across multiple criteria. This study examined biodata criterion-related validity across multiple performance criteria in a sample of 6,036 automated systems controller applicants. The performance measures examined included a paper and pencil knowledge test, behavioral ratings of simulation performance, and a paper and pencil test of simulated problems. The results suggested that the biodata instrument was useful in predicting performance across a number of criteria. The use of a shortened biodata scale captured 89 % of the corrected criterion validity obtained from the original 142-item biodata instrument. Practical implications of these findings are discussed. ____________________________________________________________________ Biographical data (biodata) is a paper-and-pencil personnel selection technology in which applicants are asked about previous life experiences that are presumed to influence their personal development. Items are usually constructed in multiple-choice format and optimally weighted to predict criteria of interest (Mumford & Owens, 1987; Owens, 1976). Biodata can be likened to a more standardized version of a paper and pencil behavioral interview in that it focuses on applicants ’ past experiences. Each applicant is presented with each biodata item in an identical manner and is given the same set of possible response options from which to choose to best describe either the frequency or magnitude of his or her previous experiences. Meta-analytic reviews report average biodata cross-validities between r =.30 and.40 (Beall, 1991; Hunter &amp