It says that bullshit like Myers Briggs is misused by laymen. That implies that actual psychometric tests done by experts can be correctly used.
The real headscratcher though is that either Canonical thinks psychometrics work, but still bother with all the other nonsense; or they don't think they work but still leave it in in such a long hiring process...
For example in an adaptive test, question are trialed in order to determine how discriminating they are (i.e. which side of an ability estimate correctly answering places you), the likelihood of error in responding (maybe how confusing the item is or some other interference), likelihood of guessing, etc. It is essentially a statistical discipline with some grounding in psychology, but all of the pyschometricians I worked with considered themselves primarily statisticians.
I've never understood psychometry to be a theory about personality, though I suppose you could apply it to such a theory. Maybe its a case of the same term applied to two things?
> I think that quote says the opposite of what you seem to imply it does.
It says that the use of personality tests in employment procedures is a misuse.
In any case, "personality tests" is not the same as psychometrics. Just like astrology being fraudulent doesn't imply astronomy being unscientific, Myers Briggs being bullshit says nothing about the validity of psychometrics.
It's astonishing. If this line of reasoning was used against other fields of science, many people here would call it out immediately, but psychology is apparently always fair game.
(NB: I don't believe psychometrics are very good indicators in the hiring process.)