| This is a much better study than this recent one on HN that only looked at publication count, because it has six metric rather than one: https://news.ycombinator.com/item?id=35699889 There's been a lot of discussion about diversity and equity in the sciences, but fundamentally, efforts to predict what young individuals are best suited to academic work based on their identitarian profile are foolish and doomed to failure. The best approach is to take the largest possible pool of candidates, i.e. eliminate all restrictions on who gets to enter the pool, and then proceed to put them through the same series of challenges and see who makes it through each successive stage, while also offering the same level of support to each individual. This 'salmon swimming upstream' approach doesn't attempt to predict outcomes based on race/gender/class/etc., it instead selects neutrally based on ability and effort. Given that only a small fraction of the overall population is going to have the necessary combination of mental ability and dedicated interest that it takes to do painstaking scientific research, you want to start with the largest pool of candidates possible, and this is why limiting that pool to members of one group is a very bad idea if you value scientific progress. This makes clear the point at which anti-discrimination policies should be applied: everyone who wants to should get to enter the competition and have equal initial support. Of course, there are issues here with academic development in the K-12 pre-college years, related to parental involvement and cultural expectations, economic disadvantages and quality of local schools, etc. but some of the responses - cutting back on algebra for 8th graders in California etc. - are just idiotic. Incidentally, nepotism is probably more of a problem today, i.e. tenured professors may use political maneuvering and social networking to get their grad students and postdocs jobs as academic professors even though they aren't necessarily the best candidates for those positions. |