>The report also contains a number of other fascinating statistics. For
>example, according to (Schmidt and Hunter, 1998) years of education and
>years of job experience correlate poorly with measures of actual
>on-the-job performance: 0.10 and 0.18, respectively.
Be wary of numbers like this. As you might expect, they vary considerably from job to
job. There are big problems with restriction of range in most studies since, for
example, you don't see many highschool dropouts applying for jobs that require a PhD.
Also, is a correlation of .18 such a poor correlation? Relative to what? If
performance variance is high enough choosing only the top 10% of your applicants by
education can make a big difference in your bottom line. Schmidt and Hunter are famous
advocates of the use of IQ tests, and IQ generally is a better predictor of job
performance than just about any other single factor. But IQ isn't costless to observe
and for some jobs has a correlation with measures of performance that is as low as the
numbers cited above for years of education. Personnel Psychologists aren't economists
and a lot of the stuff they do would make an economist choke. -- Bill Dickens
William T. Dickens
The Brookings Institution
1775 Massachusetts Avenue, NW
Washington, DC 20036
Phone: (202) 797-6113
FAX: (202) 797-6181
E-MAIL: [EMAIL PROTECTED]
AOL IM: wtdickens
