<p>Newmassdad absolutely, one point makes no difference, but past 4-5 years is definitely relevant, would not you agree? Of course, if we take it to the extreme, 300 years ago, Yale was the worst college in the US (since there were just two!!) So I would not use the old data to make a quick judgment about anything.</p>
<p>Now on to the hard fact, I think the Kansas table is published in 2006. The number of Princeton winner is 64, which would suggest the data is a compilation of at least 16 years (assuming Princeton has 4 winners every year for nearly two decades straight, which is probably not true, and thus the actual age of the data is probably 20 years old more likely.) I therefore would not use this table to make an argument that NU # of winner for this year is a statistical artifact. So to make it relevant to your argument, which is that NU has been poor in the national fellowship performance, without waiting for Sam, I used your website to look up the # of winners in the past 4 years of NU, Caltech, MIT (since we talked about these 3 in the thread), Chicago (since it is involved in the thread somehow), and Harvard (as a reference for excellence, no pun intended.)</p>
<p>2009, NU 3, Caltech 4, MIT 4, Chicago 2, Harvard 2
2008, NU 3, Caltech 3, MIT 3, Chicago 2, Harvard 3
2007, NU 3, Caltech 1, MIT 2, Chicago 4, Harvard 4
2006, NU 1, Caltech 2, MIT 3, Chicago 3, Harvard 2</p>
<p>Total NU 10, Caltech 10, MIT 12, Chicago 11, Harvard 11 I dont think NU barely made the list per your earlier comment. Caltech for sure would not agree :-P</p>
<p>Anyway, I agree with your other points that NU just has gotten better in the past few years. But just because the number may have been sub-par in the early days (may be the money went to football team for Rose Bowl instead of fellowship), I would not rush to the conclusion so easily.</p>
<p>FYI, I think if my source is correct, about 70-80% of Goldwater at NU is ISP</p>
<p>brebeuff I have no idea why # of Goldwater matters. I am sort of playing devils advocate ha ha. We can start a new thread on statistical data :-)</p>