Comparative Analysis of the Relative Validity for Subjective Time Rating Scales.

摘要

The report summarizes the results of two separate investigations aimed at determining the inherent accuracy of derived job descriptions. Since the accuracy and validity of occupational data may be differentially affected by the rating scale format employed in gathering the data, the first phase of the research reported incorporates hypothetical job descriptions from which accurate criterion data could be generated. The second phase of the research necessitated the development of an occupational survey instrument specifically designed for Air Force basic trainees. Actual time spent measures supplied by supervisors was found to have sufficient reliability for use as a criterion in this phase of the investigation. Criterion comparison (CRICOM) or error values were used with an analysis of variance design to determine the relative validities for subjective time rating scales. In general, five-point relative scales were found to be inferior to the other scales used in this study. It was further established that job incumbents can use scales of greater complexity (e.g., a wider range of response options) than had previously been indicated in the literature. The discrepancies in estimates of absolute time or percentage values previously reported were confirmed, but the inaccuracies within this approach were found to relate only to the absolute raw values. When these absolute values were treated as relative indices, no significant differences in the validity of the derived job descriptions were universally obtained.