<p>
On this point we simply disagree; self-selection is clearly a factor, but I believe quality undergrad instruction is generally a prerequisite to gaining admission to and then succeeding in a PhD program.</p>
<p>I copy these tables from posts by interesteddad, who distills them from IPEDS and NSF data. It would be helpful if I changed the 1000 to percentage. 
I’ve read that the original purpose of the IPEDS data was to show that LAC graduates are not at a disadvantage compared the research U BA/BS grads; it’s not meant to be a ranking. My impression is that the 25+ schools in the lists are more or less considered to all be excellent schools. Other schools provide quality biochem undergrad instruction, as others have posted. The more info, the better, I think.</p>