It would be naive to assume that all evaluators will impartially appraise their subordinates. The evaluator or raters biases include
Errors of leniency are caused by the tendency of the lenient rater to it most of the ratees on the higher side of the scale, while the tough rater places them on the lower side of the scale.
This is so because every evaluator has his own value system, which acts as a standard against which appraisals are made.
Relative to the true or actual performance an individual exhibits, some evaluators mark high and others low. The former is referred to as positive leniency error and the latter: as negative leniency error (strictness error).
When evaluators are positively lenient in their appraisal, an individual’s performance becomes overstated. Similarly, a negative leniency error understates performance, giving the individual a lower appraisal.
If the same person appraised all individuals in an organisation, there would be no problem. Although there would be an error factor, it would be applied equally to everyone.
The difficulty arises when we have different raters with different leniency errors making judgements.
The “halo effect” is a tendency to allow the assessment on one trait to influence assessment on others.
According to Bernardin and Beatty, halo effect is a “tendency to rate high or low on all factors due to the impression of a high or low rating on some specific factors”.
This arises when traits are unfamiliar, ill defined and involves personal reactions. One way of minimising the halo effect is appraising all the employees by one trait before rating on the basis of another trait.
The central tendency error refers to the tendency of not using extreme scores on the judgement scale; most of the rates are clustered in the middle.
According to Bernardin and Beatty, central tendency is “the reluctance to make extreme ratings (in either direction); the inability to distinguish between and among ratees; a form of range restriction”.
Raters who are prone to the central tendency error are those who continually rate all employees as average.
They follow play safe policy because of answerability to management or lack of knowledge about the job and person he is rating or least interest in his job.
This type of rating will create problems, especially if the information is used for pay increases.
The rater’s personal prejudice can influence the objectivity of performance appraisals. If the rater dislikes an employee, he may rate him very poorly.
If the evaluator knows that a poor appraisal could significantly hurt the employee’s future (particularly opportunities for promotion or a salary increase) the evaluator may be reluctant to give a realistic appraisal.
The raters generally remember the recent actions of the employee at the time of rating. If a favourable action has taken place recently, the employee will be given a high rating. Otherwise he will be given a poor rating if an unfavourable action has taken place recently.
2. Opportunity Bias:
This results when the amount of output is influenced by factors beyond the control of employees. Some employees have better working conditions, supportive supervisors, more experienced co-workers and hence their output may be greater than others working on identical tasks.
3. Group Cohesiveness:
Cohesive groups with high morale can produce more than less cohesive groups with low morale.
4. Knowledge of Predictor Bias:
A rater’s knowledge of the performance of an employee on predictors can influence his appraisal ratings.
An employee who topped in the selection list might leave the impression that he is the best among the employees and hence, the rater may rate him as ‘good’ when his performance is moderate.
5. Similarity Error:
When evaluators rate other people in the same way that the evaluators perceive themselves, they are making a similarity error.
Based on the perception that evaluators have of themselves, they project those perceptions on others.
For e.g.: the evaluator who perceives himself as aggressive may evaluate others by looking for aggressiveness.