Peer performance appraisals: the impact of rater competence, rater location, and rating correctability on fairness perceptions. (includes appendix)

"For an increasing number of workers, the [performance review] ritual is changing. From factory floors to office suites, these workers . . . are being reviewed by a jury of their peers, who may know their work better than the boss does" ("Many Employees Get Review by Co-workers," 1993, p. 1-G).

Recent changes in the workplace emphasizing teamwork, autonomous work groups, and total quality management are resulting in a rethinking of the performance appraisal process. A task once undertaken primarily by managers, performance appraisals now involve a variety of individuals, including peers. Indeed, a recent Wyatt Co. survey of 897 U.S. companies found that 15% of the surveyed companies reported using peers as a source of performance information ("More Employees Facing '360-Degree' Evaluations," 1993).

Relatedly, in her 1989 book When Giants Learn to Dance, Kanter notes that in organizations using autonomous work groups, not only do team members often select new members (a form of peer evaluation), they may also allocate performance rewards via peer evaluations of team members. Similarly, in a 1991 book describing the quality programs implemented at Ford, Donald Petersen (former chairman and CEO of Ford) and John Hillkirk contend that to accurately evaluate whether someone is a true team player, it is essential to gather performance information from both the individual's peers and subordinates (p. 67).

This increased interest in using peers as a source of performance information in the popular press has not been matched by a concommitant amount of attention to the process of peer evaluation in the academic performance appraisal literature. This lack of research attention was recently noted by Cardy and Dobbins (1993) who decried the lack of research on 360-degree appraisals and who suggested that "we need to identify factors that affect the acceptance of subordinate and peer ratings" (p. 18).

This deficiency is all the more surprising given that evidence confirming the accuracy of peer appraisals has been available in the academic literature for a number of years (e.g., Wexley & Klimoski, 1984). Indeed, discussions of the performance appraisal process in academic journals seem to assume that appraisals are almost exclusively conducted by managers on their subordinates. As a result, relatively little research exists providing information as to the most effective use of peers in the performance evaluation process.

There are a number of arguments that can be made to support the notion that, in certain situations, peers may be better sources of information regarding employee performance than supervisors. For instance, because peers may have closer and more frequent contact with ratees than supervisors, they may be able to assess a wider range of performance dimensions (Borman, 1974) or to make more precise performance distinctions across ratees (Klimoski & London, 1974). Indeed, in Wexley and Klimoski's (1984) review of the research on peer performance appraisals, they concluded that the information possessed by peers concerning employee performance may in fact be more accurate than that possessed by any other rarer. Moreover, in some contexts, peer appraisals may be a necessity because supervisors or administrators are not capable of accurately evaluating performance due to a lack of knowledge about the individual's particular specialty area.

Although the predictive validity and reliability of peer performance appraisals have been well established (Kane & Lawler, 1978; Lewin & Zwany, 1976; Schmitt, Gooding, Noe, & Kirsch, 1984), the acceptability of peer ratings on the part of ratees appears to be problematic (Cederblom & Lounsbury, 1980; Love, 1981; Medland, Yates, & Downey, 1974). Indeed, one of the major obstacles constraining the use of peer performance appraisals appears to be the problem of ratee acceptance. Thus it appears that research examining the acceptability of various peer rating systems would be valuable in increasing our understanding of the acceptance issues associated with peer appraisals. Rather than approach this research in a shotgun fashion, we suggest that the literature on procedural justice or fairness provides a useful theoretical framework for examining the acceptance problems associated with peer appraisals. Procedural fairness researchers have articulated a number of criteria (or dimensions) by which individuals evaluate the fairness of the procedures to which they are subject. It is reasonable to think that employees will perceive peer appraisals as more fair when administered in ways consistent with fairness criteria.

Leventhal (1980) has proposed several features that should enhance perceived procedural fairness. Those features include the following: procedures that suppress bias, procedures that are correctable (e.g., offer the ratee an opportunity to correct appraisal errors), procedures that are applied consistently across people and over time, procedures that allow representation (or the ability to control the process), procedures that are ethical, and procedures that use accurate information to make the decision. These features have been supported empirically in numerous contexts, with various procedures (for a review, see Lind & Tyler, 1988).

There are a number of fairness criteria that seem germane to the issue of performance appraisals in general. Indeed, the following three empirical studies in the area of supervisory performance appraisals support Leventhal's (1980) proposals. In the first study, when Greenberg (1986) surveyed employees to determine the features contributing to perceptions of performance appraisal fairness, he found the following seven factors were associated with increased fairness perceptions: (a) when ratings are based on actual performance, (b) when salary/promotion recommendations are based on the ratings received, (c) when the supervisor solicits the user's input prior to the evaluation, (d) when two-way communication takes place during the interview, (e) when the ratee has the opportunity to challenge or rebut the evaluation, (f) when the rater is familiar with the ratee's work, and (g) when the rater consistently applies the performance standards.

Consistent with Greenberg's (1986) results, Landy, Barnes, and Murphy (1978) found the following features to be associated with perceptions of fair performance appraisals: the existence of a formal evaluation procedure, a supervisor who is knowledgeable about the ratee's performance, frequently conducted evaluations, an opportunity for the rater to express his or her views, and assistance from the supervisor in helping the subordinate develop plans for improving performance.

And finally, Dipboye and de Pontbriand (1981) reported similar results in that employees rated the appraisal system more positively when the performance criteria were job relevant, when the subordinate had an opportunity to present his or her side during the evaluation, and when the performance evaluation system resulted in plans and goals to improve performance.

The results of these three studies are consistent with Leventhal's (1980) hypotheses regarding the dimensions contributing to perceptions of procedural fairness (accuracy, correctability, consistency, representativeness, low bias, and ethicality). For example, the finding by Landy et al. (1978) that fairness is increased when the supervisor is seen as knowledgeable about the ratee's performance is consistent with Leventhal's accuracy criterion in that the evaluation is more likely to be accurate under such circumstances.

Although it seems likely that these procedural features would also serve to enhance the perceived fairness of peer performance appraisals, there are several differences between supervisory appraisals and peer appraisals that may cause some additional features to be important in influencing the acceptability of peer ratings. First, because peers are (by definition) at the same or similar …