If you want to estimate the true population average of a normally distributed trait, would it make more sense to use the sample mode rather than the sample average, since the mode won't be affected if say low IQ people are under-represented in the sample for various reasons?
Emil, one thing you might want to consider about Harvard undergraduates is the staggering disadvantage faced by white students. It is roughly known that nearly 20% of Harvard students are of Ashkenazi origin - at least according to a 2000 article (Tilove, 1999). This means that nearly half of white students are Jewish. The real losers of affirmative action are not Asians, but non-Jewish whites...
Great article!
Would you consider doing an update with the new 2023 Supreme Court rulings on Students for Fair Admissions?
If you want to estimate the true population average of a normally distributed trait, would it make more sense to use the sample mode rather than the sample average, since the mode won't be affected if say low IQ people are under-represented in the sample for various reasons?
Emil, one thing you might want to consider about Harvard undergraduates is the staggering disadvantage faced by white students. It is roughly known that nearly 20% of Harvard students are of Ashkenazi origin - at least according to a 2000 article (Tilove, 1999). This means that nearly half of white students are Jewish. The real losers of affirmative action are not Asians, but non-Jewish whites...
https://www1.udel.edu/educ/gottfredson/color/articles/tilove.html