Performance appraisals look to improve employee performance by measuring and analyzing an employee’s job performance; this process should result in objective, reliable feedback that can be used to develop and motivate employees.

Significance

“Appraisal reliability and validity still remain major problems in most appraisal systems, and new (and presumably improved) appraisal systems are often met with substantial resistance," according to Susan Taylor's article, "Due Process in Performance Appraisals," which appeared in Cornell University's Administrative Science Quarterly.

To be considered reliable, a performance appraisal system must consistently and objectively measure employee performance. The same exact activity and results performed by any employee should be measured and valued exactly the same. Without this reliability, employees will lose trust in the appraisal process and can result in employee backlash.

If employees do not buy in to the performance appraisal system, the system cannot accomplish its main objectives of improving performance and motivating employees. Since most companies tie rewards to appraisal ratings, an unreliable appraisal process will frustrate employees.

Facts

Rater bias accounts for most of the unreliability in the performance appraisal process. Every performance appraisal system consists of a certain extent of rater bias, because human judgment is an unavoidable part of any evaluation process.

Types

The most common and simple types of rater bias include leniency, harshness, central tendency, and similarity. Leniency bias reduces reliability because of the rater's tendency to rate all employees higher than they should be rated; harshness bias effects reliability by rating all employee efforts lower than their actual performance.

Central tendency results in a rating bias that places every employee in the middle of the rating scale. These three biases can be the result of an appraiser’s personality or mood and can pose a significant problem.

Only using a portion of the performance scale increases the chances of unreliability because a rater with a harshness tendency and a rater with leniency bias can result in two equal job performances being rated completely different. The concept of similarity bias mirrors the common rating flaw of favoritism.

Causes

Some other less prevalent and more complex causes of appraisal unreliability include contrast errors, halo and recency error. Contrast errors occur when raters compare employees to other employees rather than against objective standards. Halo bias affects reliability because the overall view of the employee’s worth influences every dimension of performance ratings rather than looking at each section separately and objectively.

Recency error causes severe unreliability in evaluations, but this problem resides in the employee’s mindset rather than a rater bias; employees that know when evaluations occur will make sure they perform their best during those select times only, resulting in recency error.

Considerations

Although removing human judgment in the appraisal process can be impossible, the following techniques will reduce the chances of rater bias. Formally structure the appraisal system and train employees on the process of objective rating. Reliability will improve if the appraisal system has specific guidelines, clear definitions and effective rating scales. Unreliability can result from raters having different training, experience and ideas on how to properly evaluate employees; providing the necessary training can help reduce the variation between raters’ mindsets.