Page 53 - Psychological Management of Individual Performance

32 ability and non-ability predictors of job performance
be most highly related to individual differences in performance during the initial stage of
task practice or performance, but decline in predictive validity for performance during
later stages of practice and experience.
In contrast, tasks that involve inconsistent information processing components are
expected to show little substantial change in ability–performance relations across phases
of skill development. That is, for (inconsistent) tasks in which stimulus–response asso-
ciations change (precluding the proceduralization of task components) individual dif-
ferences in general cognitive ability are predicted to remain strongly associated with
performance throughout phases of skill acquisition. Empirical evidence to support these
predictions is provided by Ackerman and his colleagues (Ackerman & Kanfer, 1993;
Ackerman, Kanfer, & Goff, 1995) in the context of air traffic controller training. Sim-
ilarly, Ree and Earles (1992) summarized research evidence indicating the pervasive
efficacy of general ability in predicting initial job performance in nearly every occupa-
tional domain. However, Ree and Earles (1992) suggest that individual differences in
general cognitive ability predicts job performance regardless of job/task requirements. To
the extent that individuals may proceduralize large components of job performance, and
to the extent that job performance ratings assess job proficiency (rather than judgments
involving contextual or adaptive job requirements), Ackerman’s theory implies that the
development of job expertise moderates the cognitive ability–performance relation.
A second stream of relevant research has focused on individual differences in job
knowledge as the proximal ability-related determinant of job performance. In particular,
Hunter (1983) proposed that general cognitive ability should be related to performance
in two ways: (1) through the extent to which the job calls for general mental activities,
such as planning, memory, and reasoning, and (2) through the extent to which an indi-
vidual masters the requisite job knowledge. Hunter (1983) further defined requisite job
knowledge along two dimensions: (1) knowledge of technical information about objects
and concepts required to do the job, and (2) knowledge of processes and judgmental
criteria required for efficient action on the job.
Relevant to this hypothesis, we identified two reanalyses utilizing validity generaliza-
tion procedures (Dunnette, 1972; Hunter, 1982, as reported in Hunter & Hunter, 1984) to
investigate the relationship between individual differences in job knowledge and perfor-
mance criteria. Results obtained in both studies found large-sized estimated true-score
relations between job knowledge and job performance (r = .51) and performance ratings
(r = .48).
In a meta-analytic-derived path model using military and non-military samples, Hunter
(1983) found that the estimated true-score correlation between job knowledge and job
performance was .67. Results obtained in the path model further indicated that general
ability exerts a direct effect on job knowledge, which in turn influences job performance.
Interestingly, Hunter (1983) also noted that supervisory ratings of performance were bet-
ter predicted by individual differences in job knowledge than by general cognitive ability.
Further evidence to support Hunter’s emphasis on job knowledge as a mediator of
ability–job performance relations is provided in a cross-sectional study by Schmidt,
Hunter, and Outerbridge (1986) investigating the impact of job experience, ability, and
job knowledge on performance. Schmidt et al. (1986) found that job knowledge was
the strongest predictor of job performance (r = .74), with both ability and job experi-
ence influencing job performance primarily through job knowledge. In a second related
study, Schmidt, Hunter, Outerbridge, and Goff (1988) examined the relative influence