BEHAVIORAL OBSERVATION SCALES AND THE EVALUATION OF PERFORMANCE APPRAISAL EFFECTIVENESS
指出行为观察量表（BOS）概念基础的严重缺陷，并说明其解决方案会使BOS与其他方法无异；进而讨论评估多项目评定量表时误差资本化的特殊问题及正确处理方法。
An earlier article by the present authors (Bernardin and Kane, 1980) pointed out a serious flaw in the conceptual basis of Behavioral Observation Scales (BOS). The present article explains this flaw in more detail and shows that its solution would make BOS indistinguishable from other methods already in existence. The focus then shifts from the conception of appraisal methods to their evaluation with the presentation of a discussion of the special problems of error capitalization that arise in the use of item analysis in the development of multi‐item rating scales. This discussion proceeds to describe the correct approach to removing the effects of error capitalization in the evaluation of the psychometric properties of such rating scales.