This paper discusses the development, initial psychometric assessment, and
field test of a measure of general job performance. Six hundred and forty student
participants with employment experience rated the extent to which 120 items designed
to represent Campbell’s (1990a) six non-task specific job performance factors were
relevant for measuring performance in their jobs. Scale refinement procedures were
used to reduce the scale to 59 items. Additional analyses examined the differences in
perceived dimension relevance across various job-types represented in the sample. In a
field test of the instrument, 287 U.S. Air Force enlisted technicians personnel and their
immediate supervisors completed the same measure with respect to the technicians'
jobs. The technicians and supervisors provided ratings of both item relevance for the
technicians'jobs and the technicians' actual performance on each item. Initial analyses
of the field data focused on assessing the fit of various conceptual models of
performance to the relevance data. Models tested included a one-factor model, a sixfactor
model, and a hierarchical model with six first-order factors and a second-order
general performance factor. Across the dimension relevance data provided by both
technicians and supervisors, the six-factor model provided the best fit to the data.However, the fit of the model was only marginal at best. Additional analyses of the
field data examined (1) the similarity between incumbents' and supervisors' ratings of
dimension relevance, (2) the similarity between incumbents' and supervisors' ratings of
incumbents'job performance, and (3) relationships between ratings of performance and
dimension relevance for both incumbents and supervisors. Incumbents' ratings of
dimension relevance were generally higher in magnitude and only weakly related to
those provided by their supervisors. Similarly, incumbents' self-ratings of job
performance were higher in magnitude and weakly associated with those provided by
their supervisors. Significant relationships between dimension relevance ratings and
performance ratings were obtained for both technicians' ratings and supervisors' ratings.
The relevance/performance relationships were significantly stronger for the ratings
provided by technicians. Implications of the findings and limitations of the present
study are discussed.