What 85 years of research tells us about screening job applicants...
A recent journal article summarises 85 years of research on the best ways to screen job applicants. The article lists various methods for screening candidates and how effective they are. The closer to 1.0 the number, the more powerful the methodology. Negative numbers suggest the method is counter-productive and actually screens out the best candidates:
0.54 Work sample tests
0.51 Tests of general mental ability
0.51 Structured interviews
0.48 Job knowledge tests
0.44 Job tryout procedures
0.41 Integity tests
0.38 Unstructured interviews
0.31 Conscientiousness tests
0.26 Reference checks
0.18 Job experience
0.10 Years of education
0.10 Interests
0.02 Hand writing analysis
-0.10 Age
Source: Schmidt and Hunter, Psych. Bulletin, 1998, Vol. 124, No. 2, 262-274
Thus, mental aptitude or intelligence tests, tests of personality traits and of job specific knowledge are powerful tools for predicting job success.
Structured interviews are also good. They involve asking every applicant the same questions and scoring responses by some method to yield a score. As such, they are much like tests, but more expensive to administer, requiring interviewer time.
Unstructured interviews are notoriously unreliable. Different interviewers make different judgments of the same candidate. Research on human judgment shows that our judgments of people are distorted by many factors, usually without our awareness.
We rate people of our own ethnic background higher. We rate people who appear pretty or handsome higher. We attribute peoples' success in the past more to their personal attributes and less to the setting (e.g. good supervisor and job conditions) than we should. |