A performance appraisal in most organizations follows a relatively predictable and well-known methodology. First, the managers ask the employees to describe their accomplishments and comment on their areas of improvement. The managers then assess the performance of the employees.
The underlying assumption in performance management is a belief that when managers fill an assessment form, their evaluation will be objective and fair. Every manager attempts to be completely meritocratic and objective during performance reviews. However, bias in performance appraisal inevitably creeps in, rendering the process ineffective and unfair.
Most human decisions are inherently biased and based on beliefs and intuition rather than facts or logic. Nobel Laureate Daniel Kahneman described this aspect in his book ‘Thinking, Fast and Slow.’
Not surprisingly then, despite our best efforts, biases creep into the performance evaluation process.
A bias is “an error in an assessment that occurs when a person’s conscious or unconscious prejudices influence the person’s assessment of another person.” Bias in performance appraisal causes the marking up/down of a person. It can be especially harmful in high-stakes decisions involving promotions, compensation, hiring, etc.
Organizations need to be mindful of the existence of human biases and devise strategies to guard against these biases affecting their performance review process.
Some common biases influencing the performance evaluation processes
- Recency bias
Recency bias causes people to focus on the most recent incidents rather than on the total period of appraisal. If an employee recently won a large order, it will mask an otherwise average/below-average performance. Likewise, if an employee recently messed up a deal, it will adversely affect the judgment of an otherwise good year of performance.
It is vital to collect/render regular employee feedback throughout the year to counter the effect of recency bias. It will ensure that a manager has data on performance for the entire period of appraisal. Reviewing this data during performance appraisal will help the manager arrive at an objective performance assessment rather than judging the employee based only on the most recent events.
- Primacy bias
This bias in performance appraisal causes managers to privilege first impressions of an individual and base their overall assessment on this impression rather than on performance over the entire assessment period.
Guarding against primacy bias is quite similar to guarding against recency bias. One needs to maintain a dossier on the performance of the individual being reported upon. Inputs on performance across the appraisal period prevent first impressions from being given undue weightage in performance appraisals.
- The halo/horn effect
The halo/horn effect causes individuals to anchor their assessment on one good/bad event.
The most common occurrence of the bias in performance appraisal due to the halo effect manifests itself in the assumption that a person with good grades in college will make a good employee. Likewise, assuming that a person with poorer college grades will make a bad employee is the horn effect at play.
Managers can avoid this bias in performance appraisal if they base their evaluation on varied aspects of employee performance. For example, while managers assess individual performance, it is also essential to evaluate the individual’s performance as a contributor to a team. In addition, an individual should not only undergo scrutiny for technical awareness but also for meeting the agreed deadlines.
A rounded assessment of the individual ensures that one standout positive or negative trait does not overshadow the overall evaluation and cause bias in performance appraisal.
- Centrality bias
This commonly occurring bias in a performance appraisal causes appraisers to rate people at the center of a rating scale.
This tendency to take the neutral option is not desirable in performance reviews when individuals must take a clear stand on employee performance. When managers do not take a stand and rate everyone in the middle, it becomes difficult to separate the high-performing individuals from the average ones.
The organization can overcome this bias in performance appraisal by changing the design of the performance rating scale. A simple option would be to replace the five-point scale with a four-point scale. It will require assessors to take a clear stand.
- Leniency bias
This bias in performance appraisal arises from the goodness of the human heart. It prompts managers to give favorable assessments even to employees they know have significant room for improvement.
If a manager rates every employee at 4 on a 5-point scale, objectivity is lost, and it becomes impossible to differentiate between good, bad, or average employees. It also makes it difficult to shortlist individuals for promotions, pay raises, etc. This aspect can cause the top-performing team members to feel disgruntled.
Managers can best overcome leniency bias by changing the rating scale design. The rating scale should have a broad spread that allows managers to rate employees accurately. For example, the rating scale could have an ‘above average’ rating as a half-point and a 'superior performer’ as the top rating.
- Likeness bias
It is also known as the ‘similar-to-me’ bias. This bias causes one to give a higher rating to individuals with backgrounds, interests, and skills matching one’s own.
This bias can gradually make the workplace less inclusive and reduce organizational diversity. This bias in performance appraisal results in individuals preferring to hire people from schools/ colleges that they attended or to rate such individuals higher than the others.
A sound strategy to prevent this bias from gaining hold is by requiring managers to be specific in their assessments. When managers are required to be precise, they cannot fall back on stereotypes, and the assessment becomes fair and objective.
- Idiosyncratic bias
This bias in performance appraisal causes managers to rate employees high on skills they themselves are not good at. On the other hand, they tend to rate individuals low on skills they excel. These personal eccentricities of the manager end up having a significant impact on the assessment.
It is estimated that half the variance in assessment arises out of the eccentricities of the rater and affects the rating more than actual performance. Rater bias skews performance appraisals significantly. It ends up providing information on the rater than the person undergoing evaluation.
Culture Amp, the employee experience platform, recommends some solutions for raters to overcome this bias. They advise that the raters, while undertaking assessments, consider factors such as - I would hire this person again, if this individual resigned, I would try my best to retain them, etc.
- Confirmation bias
It is a tendency that causes interpretation of any new information presented in ways that confirm our prior beliefs. It is somewhat similar to primacy bias but tends to affect individuals deeper.
Confirmation bias causes us to agree with people with similar tastes and beliefs and disagree with those without. Although a very natural human inclination, it distorts performance data. If you believe that someone is a poor performer, you will tend to interpret all the person’s actions in a manner confirming your initial assumption.
Following scientists is an effective strategy to overcome this bias in performance appraisal. Scientists test their hypotheses by looking for facts that disprove their assumptions. When you have a particular impression about someone, seek evidence that proves a trait opposite to your assumption. Also, when collecting feedback, pay special attention to feedback that contradicts your beliefs.
- Gender bias
When assessing women, raters have a tendency to rate them with a focus on their personality and attitudes. On the contrary, when rating men, they are evaluated for performance and behavior. Such gender bias diminishes growth opportunities for women and contributes to widening the pay gap.
It is best to tackle gender bias in performance appraisal by having a structured appraisal system. A structured approach forces raters to assess individuals on events, performance, and impact rather than on intangibles such as personality, style, etc.
- Sampling bias
This bias, also known as the “law of small number bias,” incorrectly leads to the thinking that a small sample represents the properties of the whole population.
To understand this better, consider a team in which everyone is a super performer. If you evaluate only one person, you assume that the rest will not be as good and end up rating the others lower. When considering performance ratings at the organizational level of the organization, the group tends to get average ratings, even though each member in the group is exceptional.
Overcoming this bias requires talent calibrations and uniform understanding across the organization of what ratings mean. For example, ‘above average’ should mean the same to everyone in the organization. However, it is easier said than done and requires organization-wide coordination to succeed.
For a performance appraisal system to be effective, it needs to be free of biases. The first step to being bias-free is accepting that all of us often unknowingly have prejudices. Some tests, such as the implicit association test, help uncover human biases.
For effective human resource management, leverage systems, processes, and technology tools to guard against biases and enable superior decisions. Consistent criteria for performance appraisal and an open mind are the best enablers of an objective and fair performance appraisal system.