Optimize Hire pre-employment tests evaluate candidates on traits proven by decades of research to predict success at work. Below, we outline the best and worst predictors of job performance based on the research of the top industrial and organizational psychologists in the world. Sometimes people are surprised by these findings and it’s common for companies to use bad predictors to evaluate candidates. If this is you, consider re-evaluating your process and adopt a strategy to screen candidates based on data-driven solutions.
Best Predictors of Job Performance
1. Cognitive ability
Cognitive ability is consistently the best predictor of job performance across all job types, levels and industries. Cognitive ability covers a wide variety of aptitudes including spatial reasoning, logical reasoning, verbal skills, computational skills, and analytical skills. A relatively short cognitive ability test can accurately predict aggregate employee success for most jobs, but a longer version may be more helpful for higher level hires.
2. Growth mindset
Growth mindset is the idea that people can always improve. Someone with a high growth mindset is willing to learn and make an effort to make themselves better. This has been studied extensively by Dr. Carol Dweck at Stanford, who has shown that high growth mindset is predictive of employee success. Like cognitive ability, growth mindset is predictive across the gamut of jobs, skill-levels and industries.
3.Conscientiousness & emotional stability
If you’ve ever heard of the popular Big Five Personality Skills test, these traits will sound familiar. Of those five commonly assessed personality traits, conscientiousness and emotional stability are by far the most effective at predicting an employee’s ability to perform well in a given role. Conscientiousness indicates someone who is aware of his or her surroundings, while emotional stability indicates someone’s ability to effectively deal with negative emotions and move forward after failure. Predicting these traits is extraordinarily helpful to managers who want employees to be able to take constructive criticism, learn from mistakes and improve with time.
Worst Predictors of Job Performance
1. First impressions
First impressions can make or break many interviews and hiring decisions. Managers are known to rely on a “gut instinct” to decide whether or not someone is right for the job, but it’s often an unreliable way to predict a candidate’s future performance. According to this article by Dr. Daniel Kahneman, you can only trust your gut if your conditions meet the following qualifications:
You’re in a predictable environment
You have regular practice
You have immediate feedback on your judgment
While many hiring managers may feel that interviews are a predictable environment and that they have regular practice as interviewers (both of these are debatable in their own right), they won’t get feedback on their judgments until months down the line, when a hired candidate has had a chance to be trained, settle in, and actually get some work done. In hiring, it’s much better to rely on research and data. Optimize Hire scores are calculated based on decades of psychological research and tested constantly to understand the most predictive traits of successful employees
2. Similarities between the interviewer and candidate
Research shows that hiring managers are extremely likely to hire someone who reminds them of themselves - especially if they don’t use another unbiased metric to measure potential employee success. Even if that hiring manager is a really great worker, the strongest teams are made up of a group of people with diverse strengths, ideas, and working styles.
3. Brain teasers and deliberately tricky questions
Brain teasers and deliberately difficult questions are actually quite poor predictors of job performance. While brain teasers sound similar to a cognitive ability test, more straightforward cognitive ability questions, like those developed by Dr. Shane Frederick at Yale, are actually correlated with future employee success. Trick questions have not been found to have any meaningful connection.