Bias in interviews and the importance of ability testing

According to Susan J. Herman, author of Hiring Right, A Practical Guide, the interview is the most frequently used selection tool and the least valid and reliable. Validity is the degree to which a selection tool like testing or interviewing measures what it is supposed to measure. The closer a statistic is to 1, the more valid it is. Yet interviews have an average validity of 0.14 while ability testing has an average validity of 0.53.

Ten interviewers interviewing the same applicant can easily come up with ten different perspectives and different reasons for liking or not liking the applicant. Research on organizational behaviour provides a lot of good information on rater bias, including the following:

1. Social construction of reality: People as a group can create their own reality where they perceive some parts of the world and ignore other parts. If everyone else in a social situation believes in a certain reality, a newcomer to the situation is likely to as well.

2. Nearness: Nearness of one person to another can lead observers to perceptually group the persons together.

3. Similarity: Persons who are alike tend to be perceived as part of a group.

4. Perceptual continuity: People tend to see events as related over time, even if they aren’t.

5. Perceptual closure: A missing piece of a picture, story or action can be supplied by the person doing the perceiving.

6. Figure-ground: People tend to see one figure on a background of other people (e.g., a teacher in a classroom situation).

7. Fundamental attribution error: Tendency to underestimate influence of the situation and overestimate influence of individual characteristics on behaviour.

8. Actor-observer effect: People tend to see their own behaviour as due more to internal and stable traits.

9. Self-serving bias: Tendency to give the benefit of the doubt to oneself; people are ready to take credit for success and make excuses for failure.

10. Perceptual defense: Inability to perceive that which is threatening to the perceiver.

11. Stereotyping: Assignment of traits and characteristics to an individual based on that person’s membership in a larger group (e.g., age, sex, race, culture, socio-economic status, types of jobs that should be held by various groups in the population, etc.).

12. Expectancy effect: Self-fulfilling prophecy: if I expect poor performance, I will get poor performance. People tend to see what they expect to see.

13. Halo effect: Letting one overall impression colour specific interpretations. Allowing the rating of an individual on one trait or characteristic to colour ratings on other traits or characteristics.

14. Similar-to-me effect: Tendency to give more favourable evaluations to persons who are similar to the rater in terms of background or attitudes.

15. Just-like-Alfred effect: Just like the “similar-to-me” effect, but in this case, the rater gives more favourable rating to persons who are similar to someone the rater considers ideal.

16. Knowledge of predictor bias: Rater gains knowledge of some factor that may predict the success of the ratee, and then rates the person to confirm this prediction.

17. Projection: Tendency for people to see their own traits and characteristics in others, even if those traits are not present.

18. Causality perceptual error: Tendency to see events as related, or caused, even when there is no connection between them

19. Leniency: Tendency to perceive the performance of one’s ratee as especially good.

20. Harshness: Tendency to perceive the performance of one’s ratee as especially ineffective.

21. Central tendency: Assigning most ratees to a middle-range performance category; not using the extremes of the rating categories.

In addition to the factors identified in organizational behaviour literature, the following two factors also have the potential to create bias:

22. Revenge and retribution: The rater seeks revenge or retribution as a result of some prior incident or perceived slight.

23. Weighting error: Weights are not pre-assigned for decision making, allowing for subjective response.

So with all of these opportunities for bias, why is the interview the most frequently used selection tool?

According to Susan Herman’s research, some people say that the hiring decision should be based on the “chemistry” between the interviewer and the interviewee. Some interviewers value high grades in college, others think grades don’t matter as much as extracurricular leadership. Some believe that assertiveness is the most important quality, whereas others find a cocky applicant offensive. Some think that those who move jobs every few years are on the fast track, whereas others believe that they are unstable and lack loyalty. Some interviewers want to learn a lot about one facet of the applicant, whereas others want to know about many facets. Some interviewers like to know as little as possible about the applicant beforehand; others like to know as much as they can. Some people like to focus on a person’s qualities, and others like to focus on a person’s past record. Some interviewers judge most by an applicant’s physical attractiveness, others by the applicants non-verbal behaviour.

In other words, the subjective rather than the objective often tends to rule when it comes to the use of selection tools.

There is much in Herman’s book for those interested in best practices for recruitment and selection. Susan J. Herman, Ph.D. is Professor of Management at the University of Alaska Fairbanks, and Professor Emerita of Management at Keene State College.

Amery Boyer
The Human Element, just a different way to manage