How fake science misleads managers

J. Antonakis

2018-04-14

Abstract:Key questions that managers often ask me include: Which is the best way to select personnel for various positions? What do you think of the MBTI (Myers-Briggs Type Indicator)? Does intelligence matter? Largescale scientific studies tell us that validated psychometric tests (e.g., The Wonderlic Intelligence test or the NEO-PI personality test) work extremely well in predicting future work performance, as do work sample tests. Structured interviews, where the manager asks predetermined work-specific questions, which are posed in the same way to all candidates, work rather well too. How does science decide what is valid? First, a test (or method) must measure what it claims to measure. Second, a test must predict future work performance. Third, a test that measures something unique must be shown to do better than competing tests. These claims must be scrutinized by the scientific community and the results published in reputable journals. In the medical field the usefulness of a medicine is gauged by predicting health status of patients and a new medicine’s efficacy is compared to known treatments. However, medicines are vetted by an independent body, and practitioners are certified to prescribe medicines. The practice of management, unfortunately, does not work in this way, and selection tests are not approved by an independent body. I often see invalid selection methods being used by companies. They may outsource the selection function to a consulting company or use an in-house person (who might not have the right training). One of the most popular selection tools is the MBTI, which is as good as a horoscope1. Why? It was developed by two individuals who were not psychologists and had no training in psychometrics. Next, the instrument purports to measure “types” or classes of individuals. However, there is no theory to explain the validity of the types, no

Psychology,Business