Unlike the OP it's at the hirer's discretion how much weight they put in the score and it's made clear to us there's a definite pattern that more nervous candidates & those without English as a natural language score less well. We ask them to come into the office, only in rare occasions does someone do it remotely.
Most of my candidates in software positions score 40-70% and have done well in the company, I've had a number in the 0-20% bracket of whom I took 1 based on his good performance in other aspects of the interview - he's not turned out to be a great hire relative to the others. A handful in the company get the 95%+ scores and they've all done well - I think many are among the best performers in our department suggesting some kind of particular capability.
The output of the test is a % score with breakdown in a few catagories plus a 2-3 page personality profile. I showed mine to a few ex colleagues, it said probably 8 things like "will have days where he's uncommunicative or seems unfriendly and days he's warmer" which everyone pointed out as eerily true about me (and the negative aspects I can therefore work on!) but it read a little like a horoscope in that you can probably find truth in it and it's up to you to what extent you forgive the paragraphs that are very untrue, although presented in a scientific way.
Honestly, I think it's pretty helpful as accompanying information and would use it again at another job
Yeah, I think that describes approximately 1 of out every 1 persons.
It would ideally be descriptive vs. prescriptive, so you could uncover things like "we get mainly recent-college-grads applying, but our strongest employees are those who joined mid-career", or "we have very few minorities who apply, but those who apply and are offered positions tend to be strong and stay with the company a long time".
Pretty meaningless unless you are hiring at Google scale, though.
Many big corporates have psychometricians on staff. At least one even has a department dedicated to psychometric testing.
There's plenty of demand for skilled developers so I don't understand why people put themselves through the humiliation. I expect that a lot of skilled people self-select out of these sorts of psychometric processes if they can, to the detriment of potential employers.
I also wouldn't be surprised if many of the tests contain an element of uncorrected racial/cultural bias, and the psychometric industry has been exploiting the inability of the government to enforce the law (Employment Equity Act), which makes discriminatory testing illegal.
Bingo! The way it actually works is you organize everything with the assumption that applicants will fail 20% of tests. A different failure mode each time, but the odds are excellent if you give an applicant ten tests, and they fail 1 in 5, every applicant will fail at least one. Then you remove all the undesirables (skin color, sexuality, that kind of thing) and declare the fact that the bosses nephew being the only guy to fail the 360 interview with the janitor turns out to be not relevant, so lets hire him.
In the original article the only reason the 1% test result mattered In That Particular Case is the HR person didn't like that asian's nationality, or race, or religion, or something like that. Guarantee if I was the candidate it would be swept under the rug (unless the HR lady hates me because of my race, politics, religion, of course)
Why would the psychometric industry want to exploit such a lack of enforcement by vending a discriminatory test? Do you think their customers are interested in using tests that give bogus results?
The Employment Equity Act doesn't make discriminatory testing illegal, it legalizes discrimination.
I don't know if the results matter.