Top private research universities - Who wins competition for students in each state?

College Search & Selection
101

<p>

</p>

<p>As p-girl noted, bcos it’s a small state, and, importantly, its 3,000 miles away. A NE student that can be accepted to Stanford will also likely be accepted to one Ivy, at least. The simple fact is that the vast majority of students attend college close to home (within a 6 hour drive). Since any Ivy can be reached in the NE within ~six hours, many MA students would also likely choose Yale over Stanford, Princeton over Stanford, Wharton over Stanford…just bcos they are a LOT closer!</p>

102

<p>Hmm collegehelp I was wondering if you had the stats of the % of HYP students that come from the Northeast (which I assume has the same population of California)? That might show us truly how more ‘diverse’ those unis are compared to Stanford. (I come from halfway across the world so this is of some interest to me)</p>

103

<p>bluebayou-
But, California is just as far from Harvard as Massachusetts is from Stanford. So, in a pairwise comparison like that, distance is equivalent.</p>

<p>superwizard-
I’ll get the answer to your question as soon as I can.</p>

104

<p>Alright, say Harvard and Stanford each have the same class size (1000) so we can ignore differences in class size. Further, let’s say California has 10,000 students willing to forego the UC system and attend a private uni while Massachusetts has 5,000 willing to forego the Mass public system and attend a private uni.</p>

<p>If Harvard attracted 100 Californians and Stanford attracted 20 from Massachusetts, I would NOT conclude that Harvard out-drew Stanford.</p>

<p>But, if I divided the 100 CA freshmen at Harvard by 10,000 I would get a score = .01 for Harvard. And, if I divided the 20 Mass freshman at Stanford by 5,000 I would get a score = .004 for Stanford. Therefore, Harvard out-drew Stanford RELATIVE to the number of students from their respective states willing to attend a private uni.>></p>

<p>But yet go back to my post. All Harvard and Stanford did in this hypothetical case was to open the phone book and randomly select 1,000 names across the US to populate their freshman classes. I set it up precisely that way – that there was no selectivity, just random chance – and your algo concludes that Harvard outdrew Stanford. What does that tell you about your algo?</p>

105

<p>percent of freshmen attending a private school (university or LAC) who attend an in-state private school</p>

<p>TX 79%
MI 78%
CA 66%
IL 64%
WI 60%
NY 57%
WA 53%
PA 50%
FL 49%
VA 46%
NC 44%
MA 34% </p>

<br>

<br>

<p>collegehelp, do you have, for each state, what % of those state’s students who go to 4 year colleges go to public schools within that state, public schools outside that state, private schools within that state, private schools outside that state? That, I think, would be more telling.</p>

106

<p>jazzymom-
Yes, the bigger the difference in the scores between two schools, the more likely it is to be a true, consistent difference. There is bound to be some random fluctuation from year to year. If there were a 10% swing in all the numbers from one extreme to another from one year to the next, a score might change by about 2-4. That is, if the freshman class size (say 1000) increased by 10% and the number of students from the source state (say 10,000) increased by 10% and the number of freshmen who enroll from that state (say 100) decreases by 10%, then the score decreases by about 2.5. But, that would mean all the variables would have to conspire against you at the same time. And, smaller states and smaller schools are more volatile in their numbers. Hence the range 2-4. The smaller things are, the more you have to allow for random fluctuations in the numbers.</p>

107

<p>Pizzagirl-
Pick a couple of states and I will get those numbers for you.</p>

108

<p>

</p>

<p>I wasn’t talking about all of the private schools. I was talking specifically about Stanford, and in particular, whether Stanford favors Californians. In fact, it is probably the opposite that is true - that Californians favor Stanford.</p>

109

<p>“But, California is just as far from Harvard as Massachusetts is from Stanford. So, in a pairwise comparison like that, distance is equivalent.”</p>

<p>That’s true. But, if we are going with sphere of influence between Harvard and Stanford then we should at least only compare those pulled from within say, 300 miles of Stanford that went to Harvard (deleting all the LA, Orange County and San Diego students for instance) and add in all the students that live within 300 miles of Harvard that are going to Stanford (which would include all the students from NYC as well as New Jersey and for that matter, Philadelphia.)</p>

<p>I’m guessing your numbers would be much, much different.</p>

110

<p>My understanding of this metric is that it’s not designed to measure appeal irt to distance per se. So Stanford doesn’t get magnetism points for attracting students from S. D. because the point of the pairing of uni’s is to compare how magnetic S is to R.I. students versus how magnetic Brown is to CA students. San Diego residents can attend Brown and add to that magnetism score but they don’t count for S because distance traveled is not the issue as much as the amount of pull it takes to get students to cross the country. Or perhaps the amount of regional “presence” a uni has in its competitor’s state. </p>

<p>IRT Harvard and Stanford, it’s fair to say that Harvard outdraws (wins) over Stanford relative to the number of those students from their respective states who choose to travel across the country to attend a high-prestige college. But NE residents don’t have to go far to have a range of choices.
Those students who live in the NE have several top-20 universities to pick from without going far from home. </p>

<p>The pull to Stanford has to be extremely strong to overcome both distance and the Ivy League magnet(s) in their home region. For CA students wanting a high-prestige, private, top-20 college, though, there are only two (looking only at your list) and one is rather specialized. The CA resident who wants a PRIVATE alternative to Stanford has no choice but to go East and that means that those colleges have a wind at their backs in competition with S in this exercise. Not to mention the magnetism of the Ivy label.</p>

111

<p>

</p>

<p>Yes, of course, the commute is the same. But, the fact is that those in MA have several other equally good (Ivy) opportunities besides Cambridge, including New Haven, Princeton, etc. And, my assumption is that kids from the NE who apply and are accepted to Stanford will also apply to and will be accepted at an Ivy college, and likely attend them instead, just bcos they are closer to home. But, Stanford has no private peer on the west coast; for that matter, no private peer west of the Mississippi.</p>

<p>cross-posted with Jazzymom, whose explanation is more clear.</p>

112

<p>jazzymom and bluebayou-
True, there are not many elite private universities in CA other than Stanford. There are Caltech and USC, which is huge. There is also a fantastic public system in CA. But, as you say, for students seeking a private uni, many probably have to look out of state.</p>

<p>On the other hand, there are plenty of elite private universities between CA and New England that are closer to CA. There are other universities that have only one or two elite universities, such as Washington U. In contrast to Stanford, Washington U. is a very successful competitor with New England schools.</p>

<p>What you are doing is speculating about the reasons why so many schools out-draw Stanford. It is just that: speculation (but a very good observation, nonetheless). </p>

<p>Do the Stanford rejects really all go to the Ivy League? I don’t think so.</p>

<p>If it is true (and we can’t know for sure) that, for many CA students it is either Stanford or New England, then that is just part of the explanation why most private universities outside CA proportionally draw more students from CA than the CA private unis draw from other states. Regional supply and demand are part of the equation.</p>

113

<p>BB: Acknowledged that Stanford has few academic peers outside of MA, CT, and NJ (HYP&MIT). But the Midwestern privates do okay against Stanford’s lure to their region’s students. According to the data presented, Chicago, Notre Dame, Northwestern and WashU all show competitive strength against Stanford’s “pull.”</p>

<p>Collegehelp: Do you have data that would indicate the relative magnetism of top colleges for international students? I can guess which ones lead, but it might be interesting to see if Stanford wins the magnetism pull for top international students, particularly from Asia, who might factor distance in their decisions.</p>

114

<p>Top students from Asia, in my observation, hold the Ivies (particularly HYP) in extremely high regard (perhaps higher than your average American).</p>

<p>That is not to say that Stanford doesn’t do well in Asia, it most certainly does, just that the fact that Stanford may be physically closer vs. Harvard/MIT is generally a non-factor. HYPSM in general is very strong indeed in Asia.</p>

115

<p>collegehelp:</p>

<p>You are turning my point around. The question is where do Mass residents go when also accepted to Stanford, bcos that is your analysis. I was just speculating that many stay close to home and head for New Haven or NJ simply bcos an Ivy pedigree is extremely important in the NE. And, yes, I forgot WashU is “west” of the Mississippi River, but not by much. :slight_smile: </p>

<p>However, please note that WashU is NOT part of the HYPSM acronym, so is not a peer of Stanford. I know many Calif parents who will pay for a top private (HYPS), but would take a UC instead of WashU bcos of the price differential, and bcos Cal Engineering is nearly equal to Stanford and MIT.</p>

<p>And, no, USC, ain’t even close.</p>

116

<p>“But yet go back to my post. All Harvard and Stanford did in this hypothetical case was to open the phone book and randomly select 1,000 names across the US to populate their freshman classes. I set it up precisely that way – that there was no selectivity, just random chance – and your algo concludes that Harvard outdrew Stanford. What does that tell you about your algo?”</p>

<p>Pizzagirl-
Random selection would mean that Harvard and Stanford were equivalent in “selectivity” and that would be exactly the conclusion from my algorithm if there were 5 times as many California students as Massachusetts students (100 CA vs 20 MA). My algorithm should work perfectly.</p>

117

<p>Carnegie Mellon win-loss record ¶</p>

<p>Carnegie Mellon University 6.2
Harvard University 5.6</p>

<p>Carnegie Mellon University 7.0
California Institute of Technology 4.5</p>

<p>Carnegie Mellon University 9.7
Yale University 7.2</p>

<p>Carnegie Mellon University 15.3
Georgetown University 11.6</p>

<p>Carnegie Mellon University 3.9
Emory University 5.1</p>

<p>Carnegie Mellon University 3.5
University of Chicago 9.6</p>

<p>Carnegie Mellon University 3.5
Northwestern University 5.8</p>

<p>Carnegie Mellon University 8.3
University of Notre Dame 11.2</p>

<p>Carnegie Mellon University 14.5
Johns Hopkins University 14.6</p>

<p>Carnegie Mellon University 6.2
Massachusetts Institute of Technology 5.4</p>

<p>Carnegie Mellon University 2.6
Washington University in St Louis 3.9</p>

<p>Carnegie Mellon University 4.5
Dartmouth College 8</p>

<p>Carnegie Mellon University 15.3
Princeton University 10.3</p>

<p>Carnegie Mellon University 9.7
Columbia University in the City of New York 9</p>

<p>Carnegie Mellon University 9.7
Cornell University 12.5</p>

<p>Carnegie Mellon University 6.3
Duke University 9</p>

<p>Carnegie Mellon University 38.2<br>
University of Pennsylvania 32.5</p>

<p>Carnegie Mellon University 10.9
Brown University 8.1</p>

<p>Carnegie Mellon University 4.0
Vanderbilt University 3.9</p>

<p>Carnegie Mellon University 5.7
Rice University 3.2</p>

<p>Carnegie Mellon University 7.0
Stanford University 2.8</p>

118

<p>But yet go back to my post. All Harvard and Stanford did in this hypothetical case was to open the phone book and randomly select 1,000 names across the US to populate their freshman classes. I set it up precisely that way – that there was no selectivity, just random chance – and your algo concludes that Harvard outdrew Stanford. What does that tell you about your algo?"</p>

<p>Pizzagirl-
Random selection would mean that Harvard and Stanford were equivalent in “selectivity” and that would be exactly the conclusion from my algorithm if there were 5 times as many California students as Massachusetts students (100 CA vs 20 MA). My algorithm should work perfectly.>></p>

<p>Your algo changes, however, when you explicitly restrict it to those students who are attending private unis. If the California system suddenly decreased in quality tomorrow and the students who would otherwise go there decided to go elsewhere, that would affect your denominator greatly, but it wouldn’t necessarily speak to the magnetism of either Harvard or Stanford. Isn’t the better denominator overall population of the state, or overall population of college-age students in the state?</p>

119

<p>Put another way, your algo is very dependent on the quality of the public unis in the states under consideration. In a situation where one state has a public uni that is very good and pulls out lots of top students and one state has a lesser quality one, you are inherently changing the game beyond just simple size-of-state pull.</p>

120

<p>Pizzagirl-
Don’t forget that this is a tug-of-war, magnet versus magnet. California public schools exert a magnetism of their own to keep students in-state. Harvard (and other OOS privates) have to overcome the attraction of CA publics. So, if the CA public system decreased in quality, yes the denominator would increase but so should the number of CA students who meet Harvard standards, and then the number of Harvard students from CA might increase.</p>

<p>If the CA public system disappeared entirely, the number of CA students going to privates would increase, the number of CA students attending Harvard would also increase, and my algorithm would work fine.</p>

<p>And, yes, it should be more than just a simple size-of-state pull. It is the pull relative to the number of students who are available to be pulled (i.e., the total number from the state attending private universities).</p>