= 27d567bc2d48d9f40e9ccc20ea370b15)
Eugine_Nier comments on Rationality Quotes April 2012 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (858)
Unfortunately, it seems to me that most of the information that "race" provides is screened off by various things that are only weakly correlated with race, and it also seems to me that our badly-designed hardware doesn't update very well upon learning these things. For example, "X is a college graduate, and is black" doesn't tell you all that much more than "X is a college graduate"; it's probably easier to deal with this by having inaccurate priors than by updating properly.
Which are generally much harder to observe.
Um, Affirmative Action. Also tail ends of distributions.
I was under the impression that AA applied to college admissions, and that college graduation is still entirely contingent on one's performance. (Though I've heard tell that legacy students both get an AA-sized bump to admissions and tend to be graded on a much less harsh scale.)
Additionally, it seems that there's a lot of 'different justification, same conclusion' with regards to claims about black people. For instance, "black people are inherently stupid and lazy" becomes "black people don't have to meet the same standards for education". The actual example I saw was that people subconsciously don't like to hire black people (the Chicago resume study) because they present a risk of an EEOC lawsuit. (The annual risk of being involved in an EEOC lawsuit is on the order of one in a million.)
A quick google search isn't giving me an actual percentage, but I believe that students who're admitted to and attend college, but do not graduate, are still significantly in the minority. Even those who barely made it in mostly graduate, if not necessarily with good GPAs.
One of the criticisms of colleges engaging in "AA" type policies is that they often will put someone in a slightly higher level school (say Berkeley rather than Davis) than they really should be in and which because of their background they are unprepared for. Not necessarily intellectually--they could be very bright, but in terms of things like study skills and the like.
There is sufficient data to suggest this should be looked at more thoroughly. In general it is better for someone to graduate from a "lesser" school than to drop out of a better one.
Okay, but if not everyone graduates from college, and the point of admissions is to weed out people who'll succeed in school rather than wasting everyone's time, then how does a college degree mean anything different for a standard graduate, a legacy graduate, and an affirmative-action graduate? (Note that the bar is lowered for legacy graduates to the same degree as affirmative-action graduates, so if you don't hear "my father also went here" the same way as "I got in partly because of my race", then there's a different factor at work here.)
In the extreme case where being above a given level of competence deterministically causes graduation, you're correct and AA makes no difference; the likelihood (but not necessarily the prior or posterior probability) of different competence levels for a college graduate is independent of race. In the extreme case where graduation is completely random, you're wrong and AA affects the evidence provided by graduation in the same way as it affects the evidence provided by admission. Reality is likely to be somewhere in between (I'm not saying it's in the middle).
It depends on the actual distribution of legacy and AA graduates.
I'd say that the point of admissions is less to weed out people who'll succeed from people who'll waste the school's time than to weed out people who'll reflect poorly on the status of the school. Colleges raise their status by taking better students, so their interests are served not by taking students down to the lower limit of those who can meet academic requirements, but by being as selective as they can afford to be. Schools will even lie about the test scores of students they actually accept, among other things, to be seen as more selective.
I think it's more a case same observations, different proposed mechanisms.