Yikes, US News dropped us to 10th

[YourLandlord]

Here we are.

Actually it's the other way around. Larger endowment means higher salaries for faculty, lower student to faculty ratios in the classroom, etc, etc down the line.

Well, right, a large endowment is very very very important. But Harvard has been raising and investing money for >350 years. Harvard and Yale are the 2 oldest schools, and they have the 2 largest endowments. These things go hand in hand -- they have a lot of money because they established excellence long long ago and have been benefiting from that for a long time.

[rthomas]

Here we are.



Well, right, a large endowment is very very very important. But Harvard has been raising and investing money for >350 years. Harvard and Yale are the 2 oldest schools, and they have the 2 largest endowments. These things go hand in hand -- they have a lot of money because they established excellence long long ago and have been benefiting from that for a long time.

Cool. The age thing is quite interesting. Can you make a regression between enowment and ranking. And a regression between age and ranking.

And this/last year every endowment took a huge hit. How does that play?

[YourLandlord]

Cool. The age thing is quite interesting. Can you make a regression between enowment and ranking. And a regression between age and ranking.

And this/last year every endowment took a huge hit. How does that play?

We probably wouldn't expect anything wilder than a linear or exponential relationship between this type of data. I've also changed from school name to actual US News score, to account for "true" margins between schools for this type of analysis.

Based on R^2, Age is not a good direct predictor of US News Score.
Based on R^2, Endowment is not a great direct predictor of US News Score, but might be a slight indicator, if only because HYP are so so old.

[As for endowment hits, some got hit more than others, but I think all are down. I don't think that will have a clear effect on rankings here, as this "2010" ranking is really pulled from 2008 data. It will probably have more of an effect on schools with moderate endowments (which are more crucial than Yale's $22Bln ridiculousness of a source to tap) over the next few years if they can't fund professorships. That data will be interesting, but hard to draw conclusions from for another few years (and even then it will be hard to pinpoint effect).]

[YourLandlord]

One more -- Endowment as a function of age.

Given a simple linear relationship, age is not a good predictor of endowment.

However, this is because endowments for the non-HYP schools are clumped and appear fairly independent of age. I maintain that HYP have such large endowments because a) they are really old and b) they were really prestigious a long time ago. These two factors have allowed them to acquire funds for a long time -- the rich get richer, if you will.

[YourLandlord]

LOL at these "sophisticated statistical techniques"

http://www.usnews.com/blogs/college-...surprises.html

stunning, stunning insight!

[cspan37421]

The US News & WR ranking of colleges is, methodologically speaking, horse puckey. It was when we peaked, and it is now. Don't worry.

[rthomas]

We probably wouldn't expect anything wilder than a linear or exponential relationship between this type of data. I've also changed from school name to actual US News score, to account for "true" margins between schools for this type of analysis.

Based on R^2, Age is not a good direct predictor of US News Score.
Based on R^2, Endowment is not a great direct predictor of US News Score, but might be a slight indicator, if only because HYP are so so old.

[As for endowment hits, some got hit more than others, but I think all are down. I don't think that will have a clear effect on rankings here, as this "2010" ranking is really pulled from 2008 data. It will probably have more of an effect on schools with moderate endowments (which are more crucial than Yale's $22Bln ridiculousness of a source to tap) over the next few years if they can't fund professorships. That data will be interesting, but hard to draw conclusions from for another few years (and even then it will be hard to pinpoint effect).]

70% of the variation is due to endoment size.

[YourLandlord]

70% of the variation is due to endoment size.

HYP really mess up this data set, as they are such outliers in all categories -- age (except P), endowment, and US News score.

[Jarhead]

It should be obvious. We dropped in the standings, because Coach K is spending too much time as the Olympic coach.

[bjornolf]

Yea, sorry, the up trending blip in ranking was all my fault. Seriously, though, the boarding school I attended went from 7th to 3rd while I was there, then fell back to around 6 or 7 in the years after I left. Duke did the same thing. It was ranked around 10th when I was picking a college, went up into the top 5 while I was there, and now is leveling back down. I just seem to have that effect on schools. I'm just that . Joking, of course. I'm sure it was all just a big coincidence.

[YmoBeThere]

I'd love to see year-by-year rankings. They go back to at least the late 1980s. It wouldn't be hard to research using those library thingies I vaguely remember we used to use before internet was invented.

The highest Duke ever ranked was third, in 1994 and/or 1995. Which by YourLandlord's thinking would have reflected stuff that went on in 1991-92. What happened at Duke in 1991 and 1992... someone remind me.

I don't think it a coincidence that some of us posting here attended at that time. Just sayin'

[cspan37421]

Dance, my little puppets! - Mort Z.

[johnb]

Third when they used my class's stats, but that was the "most academically accomplished income class ever" (or until that point)...

I know what you mean, but it's such a juicy slip that it made me wonder if there weren't some secret admissions formula that takes into account parental income accomplishment.

[YmoBeThere]

70% of the variation is due to endoment size.

And who said size didn't matter?