Article,
Introduction: Job analysis Accuracy versus consequential validity
Journal of Organizational Behavior, (2000)
Abstract
There has been a tendency in industrial/organizational psychology to treat job analysis ratings as
representing objective data, with relatively little concern with bias or inaccuracy. This point/
counterpoint exchange is highly critical of this practice, and discusses the nature of accuracy and
validity of such ratings.
Juan I. Sanchez and Edward L. Levine take the point in arguing that traditional measures of
job analysis accuracy (interrater agreement and comparison with a standard) are of limited value.
Rather they suggest we should evaluate job analysis methods against criteria based on how they
are used or on their `consequential validity'.
Frederick P. Morgeson and Michael A. Campion agree with Sanchez and Levine, using their
Counterpoint paper to expand on the issue of accuracy. They provide a detailed discussion of
accuracy from a psychometrics perspective, and oer a model of the inference process required
for conducting job analysis.
Robert J. Harvey and Mark A. Wilson take an opposing view. They argue strongly that if job
analyses are properly conducted and focus on concrete, speci®c behaviours, it is possible to
demonstrate meaningful accuracy.
These three papers, written by well-known scholars in the area, provide a thought-provoking
analysis of job analysis practices and ratings. The issues raised are important for anyone wanting
to collect useful job analysis data for either practice or research, and will hopefully stimulate
needed work in this area.
