Page 59 - Psychological Management of Individual Performance

38 ability and non-ability predictors of job performance
Goldstein,Payne,Davison,&Gilliland,2000;Huffcutt&Arthur,1994;Hunter&Hunter,
1984; McDaniel, Whetzel, Schmidt, & Maurer, 1994; Reilly & Chao, 1982; Wiesner &
Cronshaw, 1988; Wright, Lichtenfels, & Pursell, 1989). Results of meta-analytic studies
indicate estimated true-score validities for interviews ranging from .06 (Cortina et al.,
2000) to .37 (Huffcutt & Arthur, 1994; McDaniel et al., 1994). Findings by Cortina
et al. (2000) further indicate that the predictive validities for interviews tends to be higher
using structured interviews (r = .27) compared to unstructured interviews (r = .06). But
the reasons for the observed relationship between interview scores and job performance
remain relatively poorly understood. Schneider and Schmitt (1986) suggested that in-
terviews may be particularly useful for predicting individual differences in personality
traits (e.g., conscientiousness). Moreover, it is important to make sure that the variance
captured by these procedures is not the same as that captured by general ability and per-
sonality traits. That is, to move the field forward, these tests must provide incremental
predictive validity, when considered along with extant general ability and personality
variables.

BIODATA
Biodata, or background measures, contain items that are typically selected on the basis
of their conceptual and empirical relation to the job under consideration. In addition to
potential assessment of cognitive abilities and personality traits, biodata items may assess
values, attitudes, skills, work style, and previous work experiences (see, e.g., Mumford &
Stokes, 1991, Mumford & Whetzel, 1997 for reviews). Biodata items may include, for
example, questions about job behaviors (e.g., how many days per week do you typically
stay late to finish a work assignment?), questions about reactions in particular situations
(e.g., how often have you asked coworkers to help you complete a difficult assignment?),
or items assessing work attitudes (e.g., how important has it been for you to consistently
produce a high quantity of work?).
Over the past 15 years, the predictive validity of biodata measures has been examined
in several empirical and meta-analytic studies (Bobko et al., 1999; Schmitt et al., 1997;
Vinchur et al., 1998). Results of the meta-analytic studies indicate estimated true-score
and predictive validities for overall job performance ranging from .24 to .52 (Schmitt
et al., 1997; Vinchur et al., 1998). Similar to interview measures, a number of researchers
have noted that the predictive validity of biodata measures may be due in part to the
extent to which these measures capture variance associated with individual differences in
cognitive abilities and personality dimensions (Mount, Witt, & Barrick, 2000). However,
the inclusion of biodata items aimed at other non-ability factors, such as interests, values,
and attitudes, suggests that biodata predictors may assess unique and valid aspects of the
predictor space not captured in ability or personality measures.

INCREMENTAL PREDICTIVE VALIDITY

A central question for those interested in the predictive validity of biodata and interview
measures pertains to the extent to which these measures provide incremental validity,
beyond that of general cognitive ability and personality traits, for training and job per-
formance. In this section we discuss these relationships and results of recent studies