December 21, 2018

The Biases That Fooled Us

By Swathi Dhar, Juanjo Cardona

5 min

AssessmentsMulti-IndustryRecruitment

A review of the most common biases and rating mistakes when interviewing candidates, and how to minimize them. 

When interviewing candidates we want to adhere to standard and consistent evaluations across the board. However, often we incur in biases – consciously or unconsciously – that result in potentially poor hiring decisions as well as in unequal opportunities for candidates.

 

Biases undermine the consistency and fairness intended with structured interviews. Orchestras in the U.S. were mostly composed of men in the 70s, with the top five having fewer than 5% of female musicians. To address such a gender imbalance, orchestra conductors started holding blinded auditions where musicians would play hidden behind a sheet. Consequently, gender, race, personal connections, or reputation stopped accounting for anything in auditions. The only thing that mattered was the music that came from behind the sheet. The proportion of women playing in the larger orchestras has grown by a factor of five since blinded auditions became the norm – a great leap, though they still make up for only one-quarter of the musicians.

 

On the other hand, rating errors include, for instance, giving all candidates high ratings or all low ratings.

 

Most Common Biases and Rating Mistakes in a Recruitment Interview

 

Attribution bias

Categorizing people helps individuals navigating our social world more efficiently. To prevent from brain overload, we make sense of a large number of stimuli by sorting those into buckets, into categories. This categorizing frees up mental resources for other tasks. For example, our stereotype for the elderly is what causes us to speak loudly in their company, even if a particular individual does not suffer from deafness.

 

In the context of an interview, the risks for an interviewer is to end up paying more attention to actions that are more consistent with the stereotype, than to actions that contradict it, resulting in less accurate evaluations (increases the risk of getting false positives).

 

“Similar to me” / Confirmation bias

We have a natural tendency to like others who are similar to us in various ways, whether it is because they studied at the same institutions we did, support the same basketball team, or have similar interests.

 

In the context of an interview, the risk is to bestow higher ratings than they actually deserve to candidates who appear to be similar to the interviewer.

 

Halo effect

The tendency to like (dislike) everything about a person, based only in very limited information. Taking just a very limited piece of information we build unfounded associations, including things that we have not observed. Because someone shows she is such a great (poor) speaker, we assume she might be also a great (poor) creative person or a great (poor) problem solver.

 

In the context of an interview, the halo effect increases the weight of first impressions, causing the interviewer to build associations that are not funded in observations.  It might induce higher ratings on, let’s say, Problem Solving just because the candidate has a high score in verbal communication (Interpersonal Skills), irrespective of the candidate’s performance on Problem Solving. It also increases the likelihood to dismiss information that might contradict the first impression.

 

“Whether it is an attribution (“Similar to me”) or halo effect bias, the best course of action is to avoid jumping to conclusions too fast”

 

Rating mistakes: Strictness, Leniency, and Central Tendency

Rather than mistakes per se, the following tendencies devoid evaluations of any discriminatory value since they all represent a propensity to give all interviewees similar ratings.

 

Strictness refers to the tendency to give low ratings to all candidates. On the opposite extreme, we have leniency: a tendency to give high ratings to all candidates.

 

As opposed to leniency and strictness – which are about extreme ratings – interviewers might present a tendency to rate all competencies at the middle of the scoring scale. This rating mistake is referred to as central tendency.

 

How to Minimize Them?

 

How to minimize biases? Whether it is an attribution (“Similar to me”) or halo effect bias, the best course of action is to avoid jumping to conclusions too fast. When evaluating, interviewers should concentrate on the responses given by the candidate rather than on the outward characteristics or personality of the candidate; hold back from considering any non-performance related factors. A re-examination of the scores to the candidate based on the hand notes interviewers took during the interview might also help to reduce biases.

 

Another way to tackle biases is to conduct interviews using a panel of interviewers, rather than being led by just one interviewer. This is a way to even out individual judgments. Yes, there might be individual errors, but since all panelists share a common basis when all judgments are averaged, the average usually is accurate. However, a panel might not work if all panelists share the same bias and/or if panelists evaluations influence one another – that is if assessments from different panelists are correlated.

 

How to minimize rating errors? By understanding the competencies assessed and comparing the behaviors observed in the interview with the behaviors used to establish the proficiency-level ratings for each competency.

 

When there are doubts on whether to award a high (low) score, interviewers need to understand such score does not indicate perfect (complete lack of) performance. It means, in the high score cases, that the interviewee demonstrated more of the competency than it is generally common; equally, in low score cases, it means she did not show much of the competency in her responses. 

 

 

Conclusion

 

Avoiding biases is a tricky business. The use of heuristics to simplify cognitive processing power comes as second nature to us. It permeates across a wide range of situations in our daily lives, and recruiting is no exception.


The ancient Greek aphorism and Delphic maxim “know thyself” might come in handy, but probably is not enough. Knowing yourself is only a necessary condition. Borrowing a metaphor from James Williams’, this is more akin to have only a handgun when you have an army of tanks advancing upon you. Holding your ground under these circumstances it is simply not a feasible endeavor (1).


In addition to self-awareness about our natural tendency towards these cognitive shortcuts, perhaps the best tool we can apply is not to make our minds right out of the bat. As mentioned previously, wait until the interview is over, refer back to our notes, and proceed to assess the candidate once you have put some distance and detached yourself from any of the potential biases that might have approached you. Yes, no one said it would be easy.

 

(1) Williams, James (2018). Stand Out of Our Light. Cambridge University Press. UK. May 2018. Page 112.

 

 


 

Swathi Dhar

Regional Manager (Operations) at Direct HR

L: English, Chinese

T: +86 21 6010 5010

E: s.dhar@directhr.cn

 

Juanjo Cardona

Editor at ChinaHRnews.com

L: English, Spanish

T: +86 21 6010 5000

E: j.cardona@directhr.cn