Bias
Become a Better Interviewer
Interviews are poor predictors of actual job performance. Here's how to improve.
Posted June 13, 2013
Interviews are the most common personnel selection tool and the most heavily weighted factor in hiring decisions. Unfortunately, interviews are also notoriously inaccurate predictors of actual performance on the job, with decades of industrial and organizational psychology research indicating that the average validity of a typical unstructured interview is around 20%. In many cases, flipping a coin would be a more robust method for selecting among job candidates. With the cost of a bad hire estimated at one year’s total compensation, individuals and organizations can benefit greatly by learning how to boost their hit rate.
Problems with interviews
Here are some of the reasons why interview-based hiring decisions are so often inaccurate:
Intrinsic limitations of the interview:
Interviews are situation-specific “samples” of behavior that often do not generalize to job performance. When we meet job candidates in an interview context, most of us are prone to mistake their “state” in that particularly stressful and artificial situation, as instead being reflective of their “traits.” So rather than perceive that a candidate is overly formal and reserved because he or she is being interviewed, we tend to see the person as having those personality attributes in every context. There is also a higher degree of subjectivity in interviews than in other selection tools, and the two main purposes of the interview- assessment and recruitment- often interfere with one another.
Interviewer biases:
Some of the common cognitive biases that interviewers frequently demonstrate include:
- Leniency: rating all candidates favorably
- Stringency: rating all candidates unfavorably
- Central tendency: not differentiating between candidates
- Contrast effect: evaluating candidates in comparison to others
- Halo effect: one good or bad attribute determines the entire evaluation
Other factors that unduly influence interviewers’ judgment include:
- Physical attractiveness of candidate
- Perceived similarity with candidate
- Stereotypes on the basis of gender, age, national origin, ethnicity, education, work experience, etc.
- Incorrect assumptions about non-verbal behavior
Interviewer errors:
In addition to interviewer biases that influence perceptions and evaluations, interviewers are also susceptible to making errors in how they conduct interviews and process information, for example:
- Making judgments too quickly
- Gathering insufficient information by not asking tough or probing questions when necessary
- Over-weighting negative information
- Spending interviews confirming first impressions
Making interviews better
So what can be done? Here are some specific suggestions based on well-established findings in organizational psychology:
Preparing for the interview
Too often, candidates arrive at the office and the assigned interviewers have not prepared sufficiently, or at all. Here’s what can be done in advance of the candidate’s arrival:
- Clearly define the role that the candidate is being interviewed for
- Specify the necessary knowledge, skills, abilities, attributes
- Review the candidate's file in advance
Structuring the interview
The best way to compare candidates to one another is to ensure that their experiences in the interview are as similar as possible. To create a standard experience:
- Interview different candidates in the same location for the same amount of time
- Ask a consistent set of job-relevant behavioral or situational questions
- Use the same criteria and rating scale for all candidates
Conducting the interview
The conduct of the interview also matters. Here are some general suggestions:
- Set the candidate at ease and establish rapport
- Make a statement at the beginning of the interview to set expectations
- Listen carefully and actively - use the 80/20 rule – do not interrupt
- Take notes throughout the interview or not at all in order to avoid “shaping” responses
- End with a clear statement about next steps
In implementing the above, it’s helpful to strike the right balance:
- Sticking to the protocol versus adapting for the individual candidate
- Letting the candidate speak freely without letting the conversation drift
- Being friendly versus probing insufficiently
- Comprehensiveness versus redundancy
- Being encouraging but not biasing responses
Ask the right questions
In addition to standardizing the conduct of the interview, it’s also crucial to standardize the content of the interview:
- Ask the same questions of all candidates to ensure consistency and enable comparison
- Follow-up and probing questions can vary when appropriate
- Ask one question at a time
- Use open-ended rather than closed ended or leading questions
- Don't ask questions which encourage candidates to present strengths as weaknesses
Don’t ask the wrong questions
It is illegal to treat candidates differently on the basis of, to ask questions about, or to make employment decisions based on factors such as:
- Age
- Gender
- Race
- Religion
- Marital, family or residential status
- Place of birth, country of origin or citizenship
- Arrest record
- Disabilities
- Health
Use consistent ratings and rating scales
As described earlier, it’s imperative to standardize the ways in which information is gathered across different interviews. Equally important is to standardize how information gets evaluated. Here are some suggestions:
- Rate all candidates on the same criteria using the same scale
- Use separate rating scales for each criterion
- Be mindful of biases in making ratings
- Make ratings as soon as possible after the interview
- Discuss ratings of candidates with other interviewers as soon as possible after interviews
In conclusion, getting interviews right matters a lot. Ideally, employment interviews should be part of an evolving, integrated candidate evaluation system embedded in an evolving, integrated Human Capital system. The guidelines above, when implemented properly, can increase the accuracy of job interviews and spare the organization the time, money and energy that gets wasted when interviews are not predictive.