The Terrible, Horrible, No Good, Very Bad InterviewMarch 15, 2016
Recruiters and hiring managers love interviews. They particularly love unstructured interviews where questions change as the conversation evolves. Hiring managers often say when they let the conversation flow, they can get a gut feel for the candidate’s overall job fit. They say they “just know” whether or not someone will succeed. There’s only one problem with that statement. It’s completely false.
In a meta-analysis of 85 years of peer-reviewed research on 19 selection tools, the unstructured interview came in 9th. Specifically, the same authors of that study also found that job interviews can only predict approximately 14% of the variability in employee performance.
Despite its terrible validity, the unstructured interview remains the most common part of the hiring process because it’s free, easy, and requires no skills, knowledge, or training to deliver.
Instead of requesting work samples or using a test of general mental ability (the two most valid selection tools; .54 and .51 respectively), hiring managers say “so, tell me about yourself” and are puzzled when a candidate fails miserably in the first year. Unstructured interviews allow candidates to tailor their answers to what the interviewer wants to hear. They also allow an interviewer to inject bias into the process by asking questions that confirm a preconceived beliefs about a candidate.
In fact, the degree to which bias plays a role in an unstructured interview is particularly troubling as the EEOC consider an interview to be a selection test, and therefore requires that it be validated before use. No hard data exists on how many organizations have studied the validity of their unstructured interviews, but there are probably very few.
The problem is we think we’re excellent judges of character, but we’re not.
For a more successful, reliable, and fair selection process, consider adding more valid predictors of behavior to your interview process. As previously stated, requesting work samples was found to be the most valid (.53) selection tool, followed by using tests of general mental ability (.51) such as the Watson-Glaser™ Critical Thinking Appraisal, RAVEN’S Progressive Matrices™, or Advanced Numerical Reasoning Appraisal™.
When you do conduct an interview, you can increase validity by taking a competency-based structured interview approach. A carefully crafted structured interview that measures specific skills or competencies necessary for the job and is delivered consistently will help reduce bias and evaluate candidates more systematically. Additionally, multiple valid measurements should always be combined during the selection process to get a more accurate measure of a candidate’s abilities.
While interviews are free, easy, and trusted as part of the selection process, there are far better measures of a candidate’s ability that should be incorporated to ensure you hire the best possible candidate for your organization.
To learn more about how organizations are using assessments to select and retain their talent, download our white paper here.
Editor’s Note: Breanne Harris is the Solutions Architect for Pearson TalentLens. She works with customers to design selection and development plans that incorporate critical thinking assessments and training. She has a Master’s degree in Organizational Psychology and has experience in recruiting, training, and HR consulting. She is the chief blogger for Critical Thinkers and occasionally posts at ThinkWatson. Connect with her on LinkedIn and Twitter for more of her thoughts.