Page 52 - Psychological Management of Individual Performance

general cognitive ability and job knowledge 31
success and skill acquisition criteria represent level of performance attainment brought
about by changes in behavior (Ackerman, 1988; Pearlman, Schmidt, & Hunter, 1980).
Eight meta-analytic studies examined ability–job performance relations (Bobko,
Roth, & Potosky, 1999; Callender & Osburn, 1981; Dunnette, 1972; Schmidt, Hunter, &
Caplan, 1981; Schmitt, Gooding, Noe, & Kirsch, 1984; Schmitt, Rogers, Chan,
Sheppard, & Jennings, 1997; Vinchur, Schippmann, Switzer, & Roth, 1998; Vineberg &
Joyner, 1982), four meta-analytic studies investigated ability relations to both job perfor-
mance and training outcome criteria (Callender & Osburn, 1981; Hunter & Hunter, 1984;
Pearlman et al., 1980; Schmidt et al., 1980), and one study examined ability relations to
skill acquisition (Colquitt, LePine, & Noe, 2000). As shown in Table 2.1, results of each
meta-analytic study shows a positive relationship between individual differences in
general cognitive ability and work-related outcome criteria.
Further examination of results shown in Table 2.1 suggests that that the size of validity
coefficients obtained varies as a function of the criterion used. Validity coefficients also
vary as a function of the corrections that are made. In general, with the exception of
findings obtained by Colquitt et al. (2000; r = .38), validities tend to be highest for pre-
diction of training criteria (ranging from .54 to .91), followed by job proficiency ratings
(.52 and .73), followed by overall job performance/supervisory ratings (ranging from .25
to .32). The higher predictive validities for training measures compared to overall, multi-
dimensional job performance measures suggests that individual differences in general
cognitive ability may account for more variance in performance when performance is
defined in terms of skill acquisition or job proficiency (see, e.g., Ackerman, 1994, for a
discussion of this issue).
The accumulated meta-analytic findings indicate that individual differences in general
cognitive ability can be expected to account for variance in performance across a wide
range of jobs. From a scientific perspective, however, the size of the estimated true-score
correlations (ranging from .25 to .91) suggests that the overall meta-analytic findings
tell only part of the story. That is, the variability of predictive validities for ability on
performance criteria suggests that ability influences on performance may be moderated
by unspecified factors (such as task demands) and may accrue through multiple pathways.
During the past decade, several new streams of research have emerged that more
precisely delineate the scientific basis for observed ability–performance relationships.
Below, we discuss three of these recent research streams: (1) theory and research in-
vestigating ability–performance relations as a function of task demands, (2) studies
investigating ability, job knowledge, and job performance relations, and (3) theory and
research conceptualizing adult intelligence in terms of process and knowledge, rather
than in terms of a general ability factor.
The first contemporary line of inquiry addresses the notion that job demands may mod-
erate the strength of the relationship between general cognitive ability and performance
(Gottfredson, 1986). Building upon an integration of cognitive, information processing
theory and ability theory, Ackerman (1988), for example, proposed that in tasks char-
acterized largely by consistent information processing components, general ability and
broad content abilities should be most strongly associated with the initial, novel stage
of task performance (such as during training) and less associated with performance in
intermediate and late stages of task performance (such as after substantial job tenure). As
such, for jobs with predominantly consistent task requirements, general ability should