of Applied Psychology
Copyright 1993 by the American Psychological Association, Inc.
Walter C. Borman, Mary Ann
Oppler, Elaine D. Pulakos, and
Leonard A. White
This research explores the role of early supervisory experience and cognitive ability in first-line
supervisor performance. Similar to F. L. Schmidt, J. E. Hunter, and A. N. Outerbridge's (1986) study
of nonsupervisors, this research tested structural models hypothesizing relationships among su-
pervisory experience, cognitive ability, supervisory knowledge and proficiency, and performance
ratings, using a sample of 570 second-tour soldiers. The Schmidt et al. model with an additional
ability -> experience path provided the best fit. The significant ability -* experience path was
interpreted as indicating that demonstrated ability contributes to soldiers being given the opportu-
nity to obtain supervisory experience. Experience had a greater impact on supervisor proficiency
than on supervisor knowledge. Ability had a greater impact on supervisor knowledge than on
proficiency. Discussion focuses on the personal characteristics that might be involved in being
assigned supervisory responsibilities.
Somewhat surprisingly, there have been few studies of empiri-
cal relationships between experience on a job and performance
on that job. In a review article more than 10 years ago, Gordon
and Johnson (1982) were able to cite only about six studies
directly relevant to this issue. Furthermore, the results were
equivocal, with evidence of both positive correlations between
nonsignificant correlations between experience levels and per-
formance (Ronan, 1970).
Since the Gordon and Johnson (1982) review, three studies
especially relevant to the experience-performance relationship
have been conducted. In a meta-analysis of Department of La-
bor data, McDaniel, Schmidt, and Hunter (1988) investigated
the relationship between job experience and job performance.
Results showed that the correlation between experience in pres-
ent occupation and supervisory performance ratings (corrected
for attenuation) across 947 samples and more than 16,000 em-
ployees was .32. Correlations were higher for low-complexity
jobs (.39 vs. .28 for high-complexity jobs), possibly because on-
the-job experience is the way these jobs get learned, whereas
Walter C. Borman, Department of Psychology, University of South
Florida, and Personnel Decisions Research Institute, Minneapolis,
Minnesota; Mary Ann Hanson, Personnel Decisions Research Insti-
tute, Minneapolis, Minnesota; Scott H. Oppler, American Institutes
for Research, Washington, DC; Elaine D. Pulakos, Human Resources
Research Organization, Alexandria, Virginia; Leonard A. White, U.S.
Army Research Institute, Alexandria, Virginia.