Hacker News Reader

Comment by SCAQTony | Hacker News Reader

"...To get more women into their ranks, many orchestras use blind auditions where musicians perform behind a [curtain]..."

http://www.theguardian.com/women-in-leadership/2013/oct/14/b...

This has helped orchestras in preventing gender/racial discrimination reasonably well. It's very fair and reasonable too because one either hits the note or they don't.

Now, what if every student applicant to elite colleges and universities were given a randomized number and their acceptance therein was solely based upon academic achievement, psychological testing and a resume, is that the "fair" way to do it as well?

I think it would be a fast curtain to "hive mind" creating less opportunities for the underprivileged no matter what their race, gender or otherwise.

View on HN

Yeah, I don't think such a system would be at all ideal. For one thing, I think kids' essays actually are quite useful in getting a sense of what they care about (not clear what "psychological testing" could replace that), and some kids are going to mention their gender / race / socioeconomic background.

There is no higher ideal than a pure meritocracy. We want to allocate the best educational resources to the people most able to put them to use benefiting all humanity. I'm not willing to compromise on this principle. Admitting the less qualified because of their ethnic groups or gonads is simply antithetical to the entire project of civilization and progress.

Hm. So would you be okay with admitting someone who had grown up with less money and scored less high on math tests because he hadn't had access to the best teachers?

The end result is that he is not as good as someone who had money and good teacher.

The author can try to rationalize the fact that she is the female token all she wants. The fact is 'potential' is worthless, what matters is realization.

I doubt you really believe that. Eg, if I told you for sure that the kid with the good teacher was going to be less good than the kid with a bad teacher a year from now, surely you would choose the kid with the bad teacher. In the same way that you would choose the star athlete with an injured leg because he would be excellent once he had healed.

How would you judge that, though? And how much 'potential' is actually realised by the place you go to - faculty and facilities influence outcome (access to supercomputers, for example) and maybe the only thing distinguishing a good student from stanford and someone from maryland is the bias in faculty: http://advances.sciencemag.org/content/1/1/e1400005. In that case, most potential candidates are virtually similar, and some sort of affirmative action is required to balance out the gender pool.

Yes. You can't distinguish that case from a lack of innate ability, and history is full of examples of poor people doing well in academia and elevating themselves. Standardized testing, not affirmative action, is the great equalizer.

Standardized testing is hugely influenced by unfair factors. Like the fact that if I'm rich, I can pay for my kids to get better tutoring which boosts scores by hundreds of points. Or the fact that stereotype threat may keep minorities from doing well on tests. Or that tests can be culturally biased against people from certain backgrounds.

I don't care. Our goal should not be to equalize rewards to effort, but match talent to opportunity. It doesn't matter where that talent comes from. Trying to account for these "unfair factors" is itself hugely unfair, creates perverse outcomes, and in the end, reinforces our biases instead of eradicate them. Even if you begin these measures with the best intentions, they decay to favoritism and resentment. Fair testing has legitimacy. Your ad-hoc patches do not.

So instead of giving these positions to the people most qualified and capable of making the most out of them based on the metrics we have chosen to measure success. We should give them to people because of their upbringing or gonads? Regardless if they are the most qualified person to fill that position?

So the student (let's call them Sally) was afforded a better education, and using their superior knowledge are more qualified for a position. So Sally should be told "Sorry, this spot is being given to Sam because they are an ethnic minority."?

Remind me how that is fair for Sally and not benevolent racism towards Sam?

When I went to college, I knew some kids who had trained their whole life to be test-taking machines. These kids got great grades in school. They sat in the front row. "Is this going to be on the test?" was their favorite thing to say. They did exactly what was asked of them, and they did it well, although a lot of the time, that's all they did.

On the other hand, there were other kids who worked on VR rigs in their spare time, designed solar cars, helped launch linux distros, created early ISPs, and invented NCSA Mosaic. On the whole, these kids probably didn't get as good of grades as the kids in the first group.

A world of pure standardized testing would have selected more from the first group and less from the second. I think this would have been a shame.

You don't go to college to build your dreams. You go to college to help build someone else's.

This [0] is one of the best videos about this topic.

[0] - https://www.youtube.com/watch?v=y_ZmM7zPLyI

Standardized testing is not a great equalizer. It is a great way to maintain the status quo though. Paying for GRE classes can raise your GRE scores. Historically, IQ tests have "proven" that whites are smarter than other races.

> Historically, IQ tests have "proven" that whites are smarter than other races.

Why the scare quotes? On average, whites score a standard deviation higher than blacks on IQ tests, and East Asians score half a standard deviation higher than whites. IQ is strongly heritable (0.5-0.8) and is strongly correlated with lifespan and income.

These are facts. They do not disappear merely because they disagree with your worldview. If you start with an incorrect premise, there's no limit to how wrong your conclusions can be.

Because the link between IQ tests and intelligence is more tenuous than most people understand it to be.

Whatever you think IQ tests measure, it correlates incredibly well with life outcome.

Yeah...and of COURSE that's because IQ tests measure something inherent like intelligence! It isn't because people who do well on IQ tests tend to be wealthy and well-educated, which produces better life outcomes.

Let me be very clear. When I was 15 I took the SAT and got every single question right. That was undoubtedly in part because a) I grew up in one of the wealthiest counties in the country b) my parents got me private math tutoring from the time I was 7 c) they read to me from an early age d) they sent me to private school e) I could go on, but you get the point. If you do well on these tests, good for you! But the idea that they are unaffected by confounding factors like income, education, and stereotype threat is absurd. These false beliefs that we can just give people a test to measure true talent worry me in light of Silicon Valley's incredible lack of diversity.

> It isn't because people who do well on IQ tests tend to be wealthy and well-educated, which produces better life outcomes.

If IQ tests worked as you described, then in studies of twins going adopted into different households, we should see life outcomes correlated more with foster family than with each other. It turns out that twin life outcomes are very closely correlated with each other and strongly with IQ tests than they are to rearing family.

Your beliefs are the ones that are false, sorry.

The largest factor in educational outcomes is not any in-school factor, but rather the socioeconomic status of a child's parents.

Use the educational system to reproduce the privilege of social elites and then launder it as "meritocracy".

What could go wrong.

Reply on news.ycombinator.com