12 ms·
Here's the problem: you cannot guarantee equal outcomes across intersectional groups. Period. Not going to happen. A hundred years of psychometric research back
by core-questions 6y ago
Here's the problem: you cannot guarantee equal outcomes across intersectional groups. Period. Not going to happen. A hundred years of psychometric research backs this up; there are simply differences that cannot be papered over by adjusting metrics and applying equalizing coefficients.
A highly effective education system that really gets the best out of every person involved is also going to be selective, it is going to group people into different classes based on ability, and the resulting output is going to be highly offensive to the militant people who right now are literally doing their best to take over American cities.
So no, we can't fix this problem. We can spend more on it, but ultimately all of that money is just IOUs against a decaying system that demands everyone be equal, even if that lowest common denominator means all but the super-rich end up living in a favela within a few decades.
- comex 6y agoIn short, you're claiming that some races or genders are inherently more capable than others, independent of all 'nurture' factors? If so, you're wrong, but feel free to clarify what "hundred years of psychometric research" you're referring to.
- core-questions 6y agoHow about you show me an IQ study with a big cohort that shows otherwise? You're the one asserting something that doesn't line up with anyone's experience with zero proof.
- candiodari 6y agoNo we are claiming that we need $4309582309239 extra spending on the Governor's wife's tennis partners' baker's niece's program to totally-not-put-all-those-dollars-back-into-the-governors-electoral-budget. That will fix it. What I mean to say is that 99.9% of those programs are just abusing the system (even if 10% of them started out with honest intentions).