“…This fundamental distinction between JA versus JS was re¯ected in the de®nition of JA oered by Harvey (1991, p. 74), which explicitly excluded speculations regarding`job requirements' that are couched in terms of inferred worker KS requirements and hypothetical AO traits. Unfortunately, some authors persist in choosing to blur this distinction by using the term`job analysis' to refer to both the description of work, as well as the process of inferring worker-trait requirements,`competencies,' and similar speculative judgments (e.g., Fleishman and Mumford, 1991;Hughes and Prien, 1988;Lopez, Kesselman and Lopez, 1981). Numerous reasons exist for arguing against this practice (see Harvey, 1991, pp. 75±78 for a more detailed discussion), including the types and magnitudes of inferences required, the cognitive decision making processes involved, the amounts and levels of speci®city of information that must be considered, and perhaps most important, the fundamental dierences in the types of properties being rated (i.e., properties of jobs, expressed in terms of work behaviors and contextual characteristics, in the case of job analysis, versus properties of people, expressed in terms of hypothetical psychological traits, in the case of job speci®cation).…”