A new breakthrough for performance appraisal

Abstract

Most performance appraisal techniques developed in recent years have one thing in common: they all attempt to assign quantitative values to the judgments of raters. Such an approach is used because many job-related behaviors in modern organizations, especially for white-collar managerial and service-type jobs, simply cannot be measured with any so-called hard objective criteria (for example, the number of widgets produced or sold). Unfortunately, the traditional quantitative rater judgments are greatly influenced by many sources of bias or error. For example, there may be problems with leniency, central tendency, and bias or halo. Such errors reduce the effectiveness of the performance appraisal and expose it to charges of discrimination. Recently there has been some theoretical discussion of the advantages of behaviorally based performance appraisals. For example, William J. Kearney, writing in the June 1976 issue of Business Horizons, described the advantages and costs of such appraisal instruments and prescribed guidelines for their application. Because this form of appraisal is relatively new, what has been lacking are studies which systematically analyze its effectiveness. Recently the authors developed a behaviorally anchored rating scale (BARS) fora state agency, assisted in the actual implementation of the technique and most important, systematically evaluated:its effectiveness in comparison to a graphic rating scale that was being used to appraise personnel. This. article gives a brief background of the BARS technique, goes through the steps of how the scales are developed and implemented, and presents the data, analysis and conclusions on the effectiveness of BARS. Recently, behavioral anchored rating scales, sometimes called behavior expectation scales, have emerged as a rating technique that may represent a substantial improvement over traditional rating approaches. In particular, the BARS technique attempts to develop valid criteria for performance evaluation. Although there are slight variations, the technique involves three basic steps: