4 ms·

I think this is the right step forward, again, but we should work very hard to make the tests as unbiased as they can be. They are not perfect but they are a re

by Ajay-p 3y ago

I think this is the right step forward, again, but we should work very hard to make the tests as unbiased as they can be. They are not perfect but they are a reliable measurement, especially as the creators have gotten better at removing hidden biases.

bushbaba 3y ago
Society is biased, and universities prepare students to be successfull in the real world. While yes we should all remove hidden biases where possible, but we also need to be cognizant of some of those biases to BE successful.
- dragonwriter 3y ago
  > Society is biased, and universities prepare students to be successfull in the real world. So, if any bias, they should admit with counter-to-general-society biases, since those who satisfy the biases of the general society don’t need preparation. (Unless the goal is to just appear to be preparing people for success while doing nothing, in which case having biases aligned with those of the general society is the ideal strategy: pick the people that were going to succeed no matter what, and claim credit for preparing them.)
  - waterhouse 3y ago
    > pick the people that were going to succeed no matter what, and claim credit for preparing them. I do think this is what lots of schools do. I mean, "the best people available" will be roughly identical with "the pool of people that contains the highest fraction of those who would succeed no matter what". And if your alumni are successful, then you claim credit. Why not? (I've occasionally seen people attempt to calculate "value added", but completely accounting for selection effects is difficult.) And, by the way, just because someone would have succeeded no matter what doesn't mean you can't increase their success. It's plausible that some highly-driven highly-competent genius would have ended up becoming a top engineer going to school X, but would have co-founded a billion-dollar company going to school Y, due to some combination of getting a somewhat better education and meeting the right people. So the value proposition may be real.
  - HDThoreaun 3y ago
    The goal of these institutions is not equality. It is not to prepare everyone for success, it's to maximize the success of their graduates.
kelseyfrog 3y ago
All tests are biased, but some biased are useful. Biasing the tests for "better student" sounds obvious, but it's still a bias to be to clear.
- dragonwriter 3y ago
  Its not a bias if it is the thing the test is intended to measure; its a bias if it is something other than what the test is intended to measure.
  - kelseyfrog 3y ago
    That's fair. Thank for encouraging me to look up the definition

by Ajay-p 3y ago

bushbaba 3y ago

Society is biased, and universities prepare students to be successfull in the real world. While yes we should all remove hidden biases where possible, but we also need to be cognizant of some of those biases to BE successful.

dragonwriter 3y ago
> Society is biased, and universities prepare students to be successfull in the real world. So, if any bias, they should admit with counter-to-general-society biases, since those who satisfy the biases of the general society don’t need preparation. (Unless the goal is to just appear to be preparing people for success while doing nothing, in which case having biases aligned with those of the general society is the ideal strategy: pick the people that were going to succeed no matter what, and claim credit for preparing them.)
- waterhouse 3y ago
  > pick the people that were going to succeed no matter what, and claim credit for preparing them. I do think this is what lots of schools do. I mean, "the best people available" will be roughly identical with "the pool of people that contains the highest fraction of those who would succeed no matter what". And if your alumni are successful, then you claim credit. Why not? (I've occasionally seen people attempt to calculate "value added", but completely accounting for selection effects is difficult.) And, by the way, just because someone would have succeeded no matter what doesn't mean you can't increase their success. It's plausible that some highly-driven highly-competent genius would have ended up becoming a top engineer going to school X, but would have co-founded a billion-dollar company going to school Y, due to some combination of getting a somewhat better education and meeting the right people. So the value proposition may be real.
- HDThoreaun 3y ago
  The goal of these institutions is not equality. It is not to prepare everyone for success, it's to maximize the success of their graduates.

kelseyfrog 3y ago

All tests are biased, but some biased are useful. Biasing the tests for "better student" sounds obvious, but it's still a bias to be to clear.

dragonwriter 3y ago
Its not a bias if it is the thing the test is intended to measure; its a bias if it is something other than what the test is intended to measure.
- kelseyfrog 3y ago
  That's fair. Thank for encouraging me to look up the definition