Performance Rating Errors

Types of Rating Errors

Common rating errors include leniency errors, severity errors, central tendency errors, restriction of range, halo errors, and recency errors. Leniency errors occur when a rater gives employees higher ratings than their actual performance warrants. For example, supervisors who rate all their average performing subordinates as “above average” on all the performance dimensions could be accused of a leniency bias. For most organizations, leniency is a very common type of rating bias; according to performance appraisal researchers Murphy and Cleveland, it is not unusual to discover that approximately 80% to 90% of all employees are rated as “above average.”

In contrast to leniency errors, severity errors transpire when a rater evaluates employees more unfavorably than their true performance justifies. Using the preceding example, a supervisor with a severity bias would rate his or her average performing coworkers as “below average.” Both severe and lenient ratings are thought to be present when the mean of all ratings given by a particular rater differs substantially from the true mean level of performance. Thus, a rater with a low mean rating is believed to be overly severe and a rater with a high mean rating is believed to be overly lenient.

Similarly, if ratings cluster around the midpoint of a performance scale, central tendency errors are assumed present. That is, a rater with a central tendency bias gives few if any employees very high or very low ratings. Central tendency, severity, and leniency errors are constant misjudgments across individuals. Thus, a rater does not have a severity bias if he or she rates only one person lower than his or her performance warrants.

...