3 ms·

Here are the actual statistics https://www.aamc.org/data/facts/applicantmatriculant/157998/factstablea24.html https://www.aamc.org/data/facts/applicantmatricula

Here are the actual statistics https://www.aamc.org/data/facts/applicantmatriculant/157998/factstablea24.html https://www.aamc.org/data/facts/applicantmatriculant/157998/... This isn't a false claim.

kenjackson 11y ago
But the data isn't broken down into schools. There is probably (although I don't know for sure), a difference in accept rate between a tier-1 vs bottom-tier medical school. They should control for that too. At least I didn't see that in the data.
danharaj 11y ago
There's lies, damned lies, and statistics. Without knowing the breakdown by school, the result can easily be the result of an effect like Simpson's paradox[1]. The AEI source even acknowledges that Simpson's paradox prevents one from drawing conclusions (also please note the rhetorical quotation marks): > The AAMC doesn’t provide acceptance data by individual medical school, so we can’t conclude that any of the four medical schools at public universities in Michigan (University of Michigan, Michigan State, Wayne State and Oakland University) are practicing illegal “affirmative discrimination” or “racial profiling” in admissions But quickly pivots, pretending like this doesn't matter: > "Based on national data, is there any conclusion other than the obvious one – that US medical schools are granting special preferences for admissions on the basis of race for certain preferred minority groups (blacks and Hispanics) over other non-preferred minority groups (Asians) and whites" Yes. Simpson's paradox. The data is consistent with schools being discriminatory in favor of whites if white people tend to apply only to selective schools, for example. This is why I don't bother with the playschool polemical bullshit that comes out of AEI. They ignore basic facts about statistical inference in order to push their views. [1] https://en.wikipedia.org/wiki/Simpson%27s_paradox https://en.wikipedia.org/wiki/Simpson%27s_paradox P.S. Looks like I got roped into reading their blog post anyway. I guess my plan to save my time and patience failed.
- remarkEon 11y ago
  -> There's lies, damned lies, and statistics. That's probably one of the worst sentences that's ever been uttered in the english language, because it lets you dismiss an argument by assuming that because it involves math it must be manipulated and therefore bullshit.
- huac 11y ago
  That's not what Yule-Simpson implies. They're missing the data breakdown by school, so you can't make the conclusion that somehow they're lying instead. Again, they don't make a conclusion and you can't either. In the absence of that school-centered data, it is reasonable to come to their conclusions on the national data set.

Here are the actual statistics https://www.aamc.org/data/facts/applicantmatriculant/157998/factstablea24.html https://www.aamc.org/data/facts/applicantmatriculant/157998/... This isn't a false claim.

kenjackson 11y ago

But the data isn't broken down into schools. There is probably (although I don't know for sure), a difference in accept rate between a tier-1 vs bottom-tier medical school. They should control for that too. At least I didn't see that in the data.

danharaj 11y ago

There's lies, damned lies, and statistics. Without knowing the breakdown by school, the result can easily be the result of an effect like Simpson's paradox[1]. The AEI source even acknowledges that Simpson's paradox prevents one from drawing conclusions (also please note the rhetorical quotation marks): > The AAMC doesn’t provide acceptance data by individual medical school, so we can’t conclude that any of the four medical schools at public universities in Michigan (University of Michigan, Michigan State, Wayne State and Oakland University) are practicing illegal “affirmative discrimination” or “racial profiling” in admissions But quickly pivots, pretending like this doesn't matter: > "Based on national data, is there any conclusion other than the obvious one – that US medical schools are granting special preferences for admissions on the basis of race for certain preferred minority groups (blacks and Hispanics) over other non-preferred minority groups (Asians) and whites" Yes. Simpson's paradox. The data is consistent with schools being discriminatory in favor of whites if white people tend to apply only to selective schools, for example. This is why I don't bother with the playschool polemical bullshit that comes out of AEI. They ignore basic facts about statistical inference in order to push their views. [1] https://en.wikipedia.org/wiki/Simpson%27s_paradox https://en.wikipedia.org/wiki/Simpson%27s_paradox P.S. Looks like I got roped into reading their blog post anyway. I guess my plan to save my time and patience failed.

remarkEon 11y ago
-> There's lies, damned lies, and statistics. That's probably one of the worst sentences that's ever been uttered in the english language, because it lets you dismiss an argument by assuming that because it involves math it must be manipulated and therefore bullshit.
huac 11y ago
That's not what Yule-Simpson implies. They're missing the data breakdown by school, so you can't make the conclusion that somehow they're lying instead. Again, they don't make a conclusion and you can't either. In the absence of that school-centered data, it is reasonable to come to their conclusions on the national data set.