Most employers still depend on graphic type rating scales to appraise performance, but these scales are especially susceptible to several unclear standards, halo effect, central tendency, leniency or strictness, and bias.
Unclear Standards: This graphic rating scale seems objective, but would probably result in unfair appraisals because the traits and degrees of merit are ambiguous. For example, different supervisors would probably define â€œgoodâ€ performance, â€œfairâ€ performance and so on differently. The same is true of traits such as â€œquality of workâ€ or â€œcreativityâ€
There are several ways to fix this problem. There the form specified what was meant by â€œoutstanding,â€ â€œSuperior,â€ and good quality of work. This specificity results in appraisals that are more consistent and more easily explained.
Halo Effect: Experts define halo effect as â€œthe influence of a raterâ€™s general impression on ratings of specific rate qualities. For example supervisors often rate unfriendly employees lower on all traits, rather than just for the trait â€œgets along well with others.â€ Being aware of this problem is a major step toward avoiding it. Supervisory training can also alleviate the problem.
Central Tendency: Some supervisors stick to the middle when filling in rating scales. For example, if the rating scale ranges from 1 to 7, they tend to avoid the highs (6 and &) and lows (1 and 2) and rate most of their people between 3to5. If you use a graphic rating scale, this central tendency could mean that you rate all employees â€œaverage.â€ That may distort the evaluations, making them less useful for promotion, salary, or counseling purposes. Ranking employees instead of using graphic rating scales can reduce this problem, since ranking means you canâ€™t rate them all average.
Leniency or Strictness: Other supervisors tend to rate all their subordinates consistently high or low just as some instructors are notoriously high or low graders. This strictness / leniency problem is especially severe with graphic rating scales, particularly when the firms donâ€™t insist that their supervisors avoid giving all their employees high (or low) ratings. On the other hand ranking forces supervisors to distinguish between high and low performers.
Therefore, if a graphic rating must be used, it may be a good idea to impose a distribution that about 10% of the people should be rated â€œexcellent,â€ 20% good and so forth. In other words, try to get a spread unless, of course, a manager is sure all his people really do fall into just one or two categories.
The appraisal done may be less objective than the superior thinks. One study focused on how personality influenced the evaluations students gave their peers. Raters who scored higher on â€œconscientiousnessâ€ tended to give their peers lower ratings they were stricter in the other words; those scoring higher on â€œagreeablenessâ€ gave higher ratings they were more lenient.
Itâ€™s not just the appraiserâ€™s tendencies but the purpose of the appraisal that causes strictness / leniency. Two researchers reviewed 22 studies of performance appraisal leniency. They concluded that â€œperformance appraisal ratings obtained for administrative purposes such as pay raises or promotions were nearly one-third of a standard deviation larger than those obtained for research or employee development purpose.
In some organizations each attribute is given a weight age and rating on a 9 point scale. The goals set are based on some mutually agreed quantity by the Appraiser and Appraised at the beginning of the appraising period. The numbers of attributes are reasonably fixed at 6 to 9 for the period under consideration. Three variable parameters in Appraisal thus reduce the chance of tendency and bias problems.