Bob Sutton - 85 years of research tells us the best way to evaluate potential employees:
Posted by Eric Barker
Just the interesting stuff.
Viewed 5628 times
Favorited
1 times
Comments (3)
Oct 26, 2009
jjjjn said...
Of course, where IQ tests are illegal or not used, years of education, interests, and major will weakly correlate with IQ and hence should show up as valid predictors of performance.
Oct 26, 2009
Chri said...
This paper was unable to account for the fact that education has, in most studies, already been used as a screening device, thus it will rightfully be much weaker when used a second time. What I mean by this is that if you consider a particular job, let's say college professor, and you look at the pool of applicants, you will see a very narrow range of educational attainment. Thus, you are looking for variation in a very narrow band, which just does not exist enough to provide value. By the vary nature of each job, people have self selected into the applicant pool, giving you a non-random collection of education levels. If we forced a truly random sample of people to apply for a particular job, such that the education levels of the applicants truly covered the full range, we would likely find education to be a significant predictor of success.
Oct 26, 2009
Bob Sutton said...
Chri,Fair enough! But of course in life people are not randomly assigned to conditions, so they had to use the naturally occurring variation. But some of the findings where there is variation, age for example, that don't predict at all are pretty interesting. Nonetheless, no study can answer everything and you have a good point,
Bob
