An Integration of Biographical Data and Personality Research Through Sherwood Forest Empiricism: Robbing From Personality to Give to Biodata.

Archival data was used to assess the effectiveness of incorporating biodata scoring into extant personality measures. Personality and biodata theories were briefly reviewed and several commonalities were noted. Hypotheses were developed for Conscientiousness, Emotional Stability, Agreeableness, and Openness to Experience. The first three dimensions were expected to predict absenteeism and all four were expected to predict turnover, and empirical scoring was expected to increase predictive validity. Data were collected from three samples (school bus drivers, bus drivers, and law enforcement) and three different personality measures were scored using England\u27s (1971) vertical percent method. Results showed that two of the three samples did not produce significant correlations between personality and two criterion measures. Only the correlation between Conscientiousness and absenteeism (r =.13) for the law enforcement sample was similar to Barrick and Mount\u27s (1991) results. Personality profiles for high and low absenteeism employees, and stayers and leavers were compared, but the profiles did not contain the hypothesized elevations. Biodata scoring improved some correlations, but inconsistently, for law enforcement employees. Biodata scoring for this sample was most useful for predicting turnover. Correlations for the other samples were small and insignificant. Whereas biodata scoring provided incremental validity over the personality scales, the results were inconsistent and therefore did not support the hypotheses. Further, profiles developed through biodata scoring diverged from those developed via personality scoring. The interpretation of biodata profiles was difficult due to possible changes in scale meaning due to scoring. The overall component of personality and biodata profiles were compared and it was shown that sometimes the biodata overall component explained incremental variance. Limitations of the study included using incumbents versus applicants, a lack of specificity of the criterion measures, and an inability to compare the results from the three samples. Future studies are needed to determine possible applicant-incumbent differences for personality and biodata measures and determine possible changes in the structure of personality tests caused by empirical scoring. These results provide disconfirming evidence for personality research and support the need to reconsider the situational specificity hypothesis