Describe the different types of rater errors.

Question

What will be an ideal response?

Answer 1

There are many possible sources of error in the performance appraisal process. One of the major sources is the rater. Although completely eliminating errors is impossible, making raters aware of potential errors and biases helps to reduce them. The different types of rater errors are as follows:Recency and primacy effects: The recency effect occurs when a rater gives greater weight to recent events when appraising an individual's performance. Examples include giving a student a course grade on the basis of only the student's performance in the last week of class or giving a drill press operator a high rating even though the operator made the assigned quota only in the last two weeks of the rating period. Another time-related issue is the primacy effect, which occurs when a rater gives greater weight to information received first when appraising an individual's performance.Central tendency, leniency, and strictness errors:Ask students, and they will tell you about professors who tend to grade easier or harder. A manager may develop a similar rating pattern. A rater who gives all employees a score within a narrow range in the middle of the scale (i.e., rate everyone as "average") commits a central tendency error, giving even outstanding and poor performers an "average" rating. Rating patterns also may exhibit leniency or strictness. The leniency error occurs when ratings of all employees fall at the high end of the scale. To avoid conflict, managers often rate employees higher than they should. This "ratings boost" is especially likely when no manager or HR representative reviews the completed appraisals. The strictness error occurs when a manager uses only the lower end of the scale to rate employees.Rater bias: When a rater's values or prejudices distort the rating, rater bias occurs. Such bias may be unconscious or quite intentional. For example, a manager's dislike of certain ethnic groups may cause distortion in appraisal information for some people. Use of age, religion, seniority, sex, appearance, or other "classifications" may also skew appraisal ratings if the appraisal process is not properly designed. A review of appraisal ratings by higher-level managers may help correct this problem.Halo and horns effects: The halo effect occurs when a rater scores an employee high on all job criteria because of performance in one area of the assigned work responsibilities. For example, if a worker has few absences, the supervisor might give the worker a high rating in all other areas of work, including quantity and quality of output, without really thinking about the employee's other characteristics separately. The opposite is the horns effect, which occurs when a low rating on one characteristic leads to an overall low rating.Contrast error: Rating should be done using established standards. One problem is the contrast error, which is the tendency to rate people relative to others rather than against performance standards. For example, if everyone else performs at a mediocre level, then a person performing only slightly better may be rated as "excellent" because of the contrast effect. But in a group where many employees are performing well, the same person might receive a lower rating. Although it may be appropriate to compare people at times, the performance rating usually should reflect comparison against performance standards, not against other people.Similar-to-me/different-from-me errors: Sometimes raters are influenced by whether people possess characteristics that are the same as or different from their own qualities. For example, a manager with an MBA degree might give subordinates with MBAs higher appraisals than those who have only earned bachelor's degrees. The error reflects measuring an individual against another person (the rater) rather than measuring how well the individual fulfills the expectations of the job.Sampling error: If the rater has seen only a small sample of the person's work, an appraisal may be subject to sampling error. For example, assume that 95 percent of the reports prepared by an employee have been satisfactory, but a manager has seen only the 5 percent that had errors. If the supervisor rates the person's performance as "poor," then a sampling error has occurred. Ideally, the work being rated should be a broad and representative sample of all the work completed by the employee.