Education is largely a waste of time
No.13349878 ViewReplyOriginalReport
Quoted By: >>13349895 >>13349947 >>13349995 >>13350063 >>13350839 >>13353128
http://talytica.com/wp-content/uploads/2016/10/Schmidt-and-Hunter-1998-Validity-and-Utility-Psychological-Bulletin.pdf
Looking at "The Validity and Utility of Selection Methods in Personnel Psychology:
Practical and Theoretical Implications of 85 Years of Research Findings"
This study looks at how different factors affect workplace performance. That is, measures of how much someone will out or underperform once you hire them. If a job pays $50k a year and someone produces output worth $75k a year, they are outperforming, etc.
The study starts by pointing out that these differences are quite large and important. Once you hire someone, the standard deviation of their job performance is approximately 40% of their base salary. This is even after considerable selection effects: if you hire someone, they already fit several selection criteria. That is to say, if you hired EVERYONE, the standard deviation would be much higher.
With that said, what are the most important criteria?
Pic related. Intelligence (GMA tests) are among the most important predictors alongside work sample tests, Job knowledge tests, peer ratings and structured interviews, all at close to 0.5 validity.
By contrast, education scores at a measly 0.1. Close to irrelevant.
Even more interesting is that once you know someone's intelligence, knowing their education is completely irrelevant. intelligence validity = 0.51, intelligence + education = 0.52.
indeed once you know someones intelligence, most other metrics don't have much additional utility.
These days education is one of the biggest factors in getting employed, yet the data suggests that it's a pretty poor measure. Certainly, most of the stuff learned in university is useless later on in life. Education is likely being used as an intelligence proxy, and it appears to be a pretty poor one.
Looking at "The Validity and Utility of Selection Methods in Personnel Psychology:
Practical and Theoretical Implications of 85 Years of Research Findings"
This study looks at how different factors affect workplace performance. That is, measures of how much someone will out or underperform once you hire them. If a job pays $50k a year and someone produces output worth $75k a year, they are outperforming, etc.
The study starts by pointing out that these differences are quite large and important. Once you hire someone, the standard deviation of their job performance is approximately 40% of their base salary. This is even after considerable selection effects: if you hire someone, they already fit several selection criteria. That is to say, if you hired EVERYONE, the standard deviation would be much higher.
With that said, what are the most important criteria?
Pic related. Intelligence (GMA tests) are among the most important predictors alongside work sample tests, Job knowledge tests, peer ratings and structured interviews, all at close to 0.5 validity.
By contrast, education scores at a measly 0.1. Close to irrelevant.
Even more interesting is that once you know someone's intelligence, knowing their education is completely irrelevant. intelligence validity = 0.51, intelligence + education = 0.52.
indeed once you know someones intelligence, most other metrics don't have much additional utility.
These days education is one of the biggest factors in getting employed, yet the data suggests that it's a pretty poor measure. Certainly, most of the stuff learned in university is useless later on in life. Education is likely being used as an intelligence proxy, and it appears to be a pretty poor one.
