Page 62 - Psychological Management of Individual Performance

progress, problems, and prospects for future research 41
in jobs involving substantial social interaction (e.g., sales) than in jobs that do not de-
mand much social interaction (e.g., writing). Barrick et al. (2000), for instance, show the
unique predictive validity of agreeableness for teamwork performance ratings, and extro-
version for sales/service performance ratings. However, recent studies (e.g., Ackerman &
Rolfhus, 1999; Rolfhus & Ackerman, 1999) also provide evidence to suggest a negative
relationship between extroversion and intelligence.
Although research evidence does not support the notion that openness to experience
makes a substantial contribution to prediction of variance in job performance, meta-
analytic research on personality–training outcome relations suggests that openness to
experience does provide strong predictive validity in learning/training environments.
In these environments, where skill acquisition typically involves volitional effort and
intrinsic motivation, an individual’s propensity for new learning is positively related to
training outcomes.
Findings with respect to the predictive validity of neuroticism for job performance
are more problematic. Although emotional stability, conceptualized in terms of general
emotional adjustment, has long been regarded as an important factor upon which to rule
out applicants for stressful or dangerous jobs, results obtained in meta-analytic studies
suggest only weak predictive validity for job performance using non-clinical personality
measures. Further consideration is needed regarding how individual differences in emo-
tional adjustment, anxiety, or emotional regulation may influence work behavior and job
performance.
The success of research in the ability and personality domains has encouraged I/O
researchers to reconsider the validity of alternative predictors, such as biodata and inter-
views. Although these measures are commonly used, and recent meta-analyses suggest
that such measures may capture important variance in the job performance criterion,
very little is known about the predictor space assessed by such measures. Incremental
predictive validity studies suggest that biodata and structured interviews assess valid
variance in the criterion, beyond that of general cognitive ability and broad personality
tests. Further research to investigate the unique aspects of the predictor domain tapped
by alternative measures, and how these aspects relate to job performance represents the
next important step in this area.

PROBLEMS
Despite the formidable progress in person prediction of job performance over the past
15 years, a number of theoretical and empirical issues remain. These issues are discussed
next.
Thefirstproblempertainstoconstructmismatch,ortheextenttowhichabilityandnon-
ability predictor tests correspond to criterion measures of job performance. Cronbach
(1949) was among the first to suggest that ability and non-ability measures differ in
terms of their construction and purpose for prediction. In particular, ability measures,
interviews and test-based training outcome measures, are designed to provide measures
of maximal performance, that is, an assessment of what an individual can do under con-
ditions of maximal effort. In contrast, non-ability personality measures and measures of
job performance generally provide assessment of an individual’s typical dispositional
tendencies and performance, respectively; that is, what an individual does under repre-
sentative conditions.