<p>p-girl:</p>
<p>Excellent follow-on. I had missed the fact that the algo assumes that all Mass residents that go to, say Tulane, are H material…[not dumping on Tulane – Scott Cowen is extremely impressive for he has done for that school]</p>
<p>this might be interesting and relevant (and more complicated)</p>
<p>percent of freshman from each state who go to different types of institutions
private universities
public universities
private LACs
public associates (community colleges)</p>
<p>Institution Type California
pub uni pct 34.5%
pri uni pct 7.6%
pri LAC pct 2.5%
pub asoc pct 55.4%</p>
<p>Institution Type Connecticut
pub uni pct 32.5%
pri uni pct 26.6%
pri LAC pct 12.5%
pub asoc pct 28.3%</p>
<p>Institution Type Georgia
pub uni pct 44.7%
pri uni pct 5.3%
pri LAC pct 6.7%
pub asoc pct 43.3%</p>
<p>Institution Type Illinois
pub uni pct 34.3%
pri uni pct 15.4%
pri LAC pct 9.1%
pub asoc pct 41.2%</p>
<p>Institution Type Indiana
pub uni pct 59.0%
pri uni pct 9.9%
pri LAC pct 8.7%
pub asoc pct 22.4%</p>
<p>Institution Type Maryland
pub uni pct 33.6%
pri uni pct 12.4%
pri LAC pct 8.9%
pub asoc pct 45.2%</p>
<p>Institution Type Massachusetts
pub uni pct 30.1%
pri uni pct 24.0%
pri LAC pct 14.4%
pub asoc pct 31.5%</p>
<p>Institution Type Missouri
pub uni pct 41.5%
pri uni pct 12.5%
pri LAC pct 6.0%
pub asoc pct 40.0%</p>
<p>Institution Type New Hampshire
pub uni pct 38.3%
pri uni pct 23.0%
pri LAC pct 16.9%
pub asoc pct 21.8%</p>
<p>Institution Type New Jersey
pub uni pct 30.3%
pri uni pct 22.6%
pri LAC pct 6.9%
pub asoc pct 40.2%</p>
<p>Institution Type New York
pub uni pct 33.1%
pri uni pct 25.5%
pri LAC pct 6.9%
pub asoc pct 34.4%</p>
<p>Institution Type North Carolina
pub uni pct 39.4%
pri uni pct 3.9%
pri LAC pct 8.0%
pub asoc pct 48.7%</p>
<p>Institution Type Pennsylvania
pub uni pct 40.3%
pri uni pct 17.2%
pri LAC pct 10.8%
pub asoc pct 31.6%</p>
<p>Institution Type Rhode Island
pub uni pct 35.2%
pri uni pct 19.1%
pri LAC pct 7.3%
pub asoc pct 38.4%</p>
<p>Institution Type Tennessee
pub uni pct 46.7%
pri uni pct 8.3%
pri LAC pct 7.6%
pub asoc pct 37.4%</p>
<p>Institution Type Texas
pub uni pct 35.8%
pri uni pct 7.5%
pri LAC pct 3.0%
pub asoc pct 53.7%</p>
<p>Institution Type Virginia
pub uni pct 44.1%
pri uni pct 7.0%
pri LAC pct 8.5%
pub asoc pct 40.4%</p>
<p>Pizzagirl-
Regarding your post #121…
“The students who are available to be pulled are NOT just merely those students who wound up at private unis.”</p>
<p>But, I think the best measure of the number of students who might go to a private uni IS the number who actually go to a private uni. These are exactly the right denominator. These are the students that no public university, LAC, or community college could attract. in the country could attract. These are the students effectively available to go to a private uni.</p>
<p>And, yes, the denominator increases when I exclude the students who went to publics. LACs, and community colleges but it does so for every state and is equivalent. The denominator SHOULD be smaller than the total number of HS grads in a state and smaller than the total number who go to college.</p>
<p>University of Notre Dame win-loss record (IN)</p>
<p>University of Notre Dame 3.8
Harvard University 6.3</p>
<p>University of Notre Dame 8.4
California Institute of Technology 6.2</p>
<p>University of Notre Dame 7.5
Yale University 7</p>
<p>University of Notre Dame 11.0
Georgetown University 9.2</p>
<p>University of Notre Dame 8.0
Emory University 11.9</p>
<p>University of Notre Dame 19.3
University of Chicago 16.8</p>
<p>University of Notre Dame 19.3
Northwestern University 25.7</p>
<p>University of Notre Dame 92.7</p>
<p>University of Notre Dame 8.3
Johns Hopkins University 2</p>
<p>University of Notre Dame 3.8
Massachusetts Institute of Technology 9.2</p>
<p>University of Notre Dame 15.2
Washington University in St Louis 21.7</p>
<p>University of Notre Dame 5.1
Dartmouth College 4.9</p>
<p>University of Notre Dame 6.5
Princeton University 15.1</p>
<p>University of Notre Dame 4.2
Columbia University in the City of New York 1.9</p>
<p>University of Notre Dame 4.2
Cornell University 2.9</p>
<p>University of Notre Dame 8.7
Duke University 7.8</p>
<p>University of Notre Dame 11.2
Carnegie Mellon University 8.3</p>
<p>University of Notre Dame 11.2
University of Pennsylvania 6.2</p>
<p>University of Notre Dame 4.8
Brown University 4.5</p>
<p>University of Notre Dame 11.9
Vanderbilt University 21.6</p>
<p>University of Notre Dame 11.1
Rice University 9.3</p>
<p>University of Notre Dame 8.4
Stanford University 5.6</p>
<p>California Institute of Technology win-loss record (CA)</p>
<p>California Institute of Technology none?
Harvard University 19.2</p>
<p>California Institute of Technology 35.8</p>
<p>California Institute of Technology 3.3
Yale University 18</p>
<p>California Institute of Technology 2.8
Georgetown University 10.9</p>
<p>California Institute of Technology 2.6
Emory University 4.8</p>
<p>California Institute of Technology 6.4
University of Chicago 9.3</p>
<p>California Institute of Technology 6.4
Northwestern University 11.3</p>
<p>California Institute of Technology 6.2
University of Notre Dame 8.4</p>
<p>California Institute of Technology 10.7
Johns Hopkins University 10.8</p>
<p>California Institute of Technology none?
Massachusetts Institute of Technology 14.3</p>
<p>California Institute of Technology 8.7
Washington University in St Louis 9</p>
<p>California Institute of Technology 6.1
Dartmouth College 10</p>
<p>California Institute of Technology 6.5
Princeton University 16</p>
<p>California Institute of Technology 4.6
Columbia University in the City of New York 12.8</p>
<p>California Institute of Technology 4.6
Cornell University 9.1</p>
<p>California Institute of Technology 41.6
Duke University 8.3</p>
<p>California Institute of Technology 4.5
Carnegie Mellon University 7</p>
<p>California Institute of Technology 4.5
University of Pennsylvania 10.6</p>
<p>California Institute of Technology none?
Brown University 15.2</p>
<p>California Institute of Technology 5.4
Vanderbilt University 3.2</p>
<p>California Institute of Technology 9.3
Rice University 6.4</p>
<p>California Institute of Technology 35.8
Stanford University 51.9</p>
<p>Emory University win-loss record (GA)</p>
<p>Emory University 5.7
Harvard University 7</p>
<p>Emory University 4.8
California Institute of Technology 2.6</p>
<p>Emory University 4.9
Yale University 8.5</p>
<p>Emory University 8.3
Georgetown University 11.6</p>
<p>Emory University 151</p>
<p>Emory University 5.4
University of Chicago 6.7</p>
<p>Emory University 5.4
Northwestern University 8.7</p>
<p>Emory University 11.9
University of Notre Dame 8</p>
<p>Emory University 11.6
Johns Hopkins University 6.9</p>
<p>Emory University 5.7
Massachusetts Institute of Technology 6.7</p>
<p>Emory University 7.8
Washington University in St Louis 10.7</p>
<p>Emory University 7.8
Dartmouth College 8.7</p>
<p>Emory University 5.1
Princeton University 10</p>
<p>Emory University 6.3
Columbia University in the City of New York 7</p>
<p>Emory University 6.3
Cornell University 4.4</p>
<p>Emory University 22.1
Duke University 19.9</p>
<p>Emory University 5.1
Carnegie Mellon University 3.9</p>
<p>Emory University 5.1
University of Pennsylvania 7.7</p>
<p>Emory University 7.0
Brown University 5.7</p>
<p>Emory University 33.9
Vanderbilt University 32.3</p>
<p>Emory University 9.0
Rice University 16.5</p>
<p>Emory University 4.8
Stanford University 6.8</p>
<p>
</p>
<p>collegehelp - do you have the stats for Michigan? thanks!</p>
<p>Regarding your post #121…
“The students who are available to be pulled are NOT just merely those students who wound up at private unis.”</p>
<p>But, I think the best measure of the number of students who might go to a private uni IS the number who actually go to a private uni. These are exactly the right denominator. These are the students that no public university, LAC, or community college could attract. in the country could attract. These are the students effectively available to go to a private uni.>></p>
<p>Let me try this another way.
Suppose that you’re looking down at a map of the United States. And a bird goes by and drops blue paint on the top 5% of students in every state in the union. There will be different numbers of blue dots in every state, because obviously some states are more populated than others – lots of blue dots in Texas, very few in Montana … but for every state, it’s the top 5% of their student population. With me so far?</p>
<p>Now, ignore the rest of the US and focus on MA and CA. CA has 5 times the population of MA, so there are five times as many 5%-kids in CA as in MA. Let’s say that there are 5,000 blue dots in CA and 1,000 in MA, for the sake of argument. </p>
<p>Harvard wants to “magnetize” the 5,000 blue dots in CA - the top 5% students in CA.<br>
Stanford wants to “magnetize” the 1,000 blue dots in MA - the top 5% students in CA. </p>
<p>What you did in your algorithm was that you took some of the 5,000 CA blue dots and you removed them, because those kids WOUND UP at UCLA / Berkeley. Even though they ARE part of the caliber of student that Harvard wants to magnetize. </p>
<p>And what you also did in your algorithm was that you took the 1,000 MA blue dots and you ADDED on, let’s say, 1,000 red dots in MA, representing those kids who weren’t H / S caliber but who wound up at Syracuse or Tulane or other private unis not of H / S caliber. </p>
<p>In other words, you’re now measuring Stanford’s ability to magnetize against a WIDER pool - not just students who were H / S caliber, but students who AREN"T H / S caliber. Of course Stanford’s ability to magnetize against them will be lower … half of those students aren’t going to ever be “Stanford-worthy” in the first place! How many of those red dots are going to be magnetized to Stanford? None, since they’re not qualified. But your algo still counts them as being part of the denominator, JUST because they go to private unis.</p>
<p>You’re using “goes to a private uni” as a proxy for “quality of student.” And it’s a poor proxy.</p>
<p>PizzaGirl:</p>
<p>Methinks data logic is a lost cause here, might as well type on another thread. :D</p>
<p>And, yes, the denominator increases when I exclude the students who went to publics. LACs, and community colleges but it does so for every state and is equivalent. >></p>
<p>Do you see why it’s not equivalent? Hint: It’s in the data you posted above, where you indicated for various states what % of the students go to publics, private unis, private LAC’s and CC’s.</p>
<p>Answer:</p>
<p>Obviously, c-help don’ “see” it, or s/he would have stopped posting pages ago…</p>
<p>About my force-field model of college admissions:</p>
<p>The denominator I used in the calculation is the best choice. There is some noise in the calculation but not much. I think many good questions have been raised and incisive criticism has been offered by some really insightful posters but I think the method I used is valid albeit a little imprecise. It provides useful information. It successfully answers a particular question.</p>
<p>There are many reasons why a student might chosses to attend a college. Combined, these reasons account for a school’s magnetism.
Possible reasons:
cost
distance
offers a major you like
urban/suburban/rural environment
prestige
perceived quality of education
climate
social life
campus beauty
(and others)</p>
<p>My algorithm controls for cost and distance leaving the other attractive forces. I am first trying to answer the question: Of students who are available for recruitment at private unis, how well do private universities compete for available students?</p>
<p>Which students are avalable for recruitment? They would first be qualified HS grads who are not price-sensitive and who would be willing to travel the required distance. There is no need to speculate about who these students might be. We know exactly who they are. They are the students who don’t go to in-state schools and who are willing to travel a certain distance.</p>
<p>By comparing schools home state with home state, I controlled for distance. I controlled for cost by selecting students for the denominator who are willing to pay the cost of private education. There are some students in each state who go to their state school for reasons other than price but price and distance are almost always big factors. Students who attend public unis in CA for reasons other than price and distance are matched in MA by students who attend their public unis (e.g. U Mass Amherst) for reasons other than price and distance, and so the calculation works fine.</p>
<p>Every year, a certain number of students graduate from HS in CA but not not all of them are effectively available for attending Harvard because they simply don’t go to college. So, number of HS grads doesn’t belong in the denominator. </p>
<p>Of those who go to college, some are destined for an in-state public. The in-state public has the advantages of low cost and short distance. We know exactly the number who go to the instate publics (universities and community colleges). The community college students are probably not Harvard material. In CA and MA the percent of students who go to public unis is about the same. The difference is that in CA, far more students go to Comm Coll rather than a private. Harvard would have to overcome price sensitivity; CA students are in effect given roughly $40,000 for attending an in-state public. Students who are price-sensitive have their shoes glued to California public soil by a $40K payoff and are effectively not available to Harvard.</p>
<p>Similarly, some students are destined for private LACs perhaps because they prefer the style of education. LAC versus university is a pretty big chasm and students who go to LACs are looking for something different than a university can offer. So, LAC students are effectively not available to Harvard. If they cross-applied to private uni and private LAC but went to an LAC, it probably means they ultimately prefer LACs. Similarly, many students who cross-apply with publics and privates but attend a public demonstrate their ultimate preference for publics. Students also go where they are not rejected when chosing among cross-applications but this phenomenon is probably equivalent in both comparison states.</p>
<p>Among the students who go to private unis, not all are really Harvard material. But that’s ok. Since we are using ratios, the only thing necessary is that the proportion of CA students attending private unis who are Harvard material is roughly the same as the proportion of MA private uni students who are Stanford material. The calculation should work fine.</p>
<p>Furthermore, the number of students who might enroll at a private uni is proportionally limited by the size of the freshman class. The use of class size as a second denominator controls for class size.</p>
<p>Regarding the relative concentration of private unis in the Northeast: This might increase the number of Mass students who attend privates but my data in an earlier post shows that the percent of northeastern state students attending public unis is about the same as CA. So, it doesn’t seem to happen. What’s more, the concentration of private unis in the northeast competes with Harvard for students from CA who are destined for NE private unis. This concentration of private unis in the NE actually makes it harder for Harvard to recruit CA students and offsets any larger denominator coming from MA to some extent.</p>
<p>I want to point out again that Washington U in Missouri actually has a 7% higher percent of students attending in-state public unis than CA but competes very well for students from northeast privates. Like Stanford in CA, Wash U is the only elite private uni in MO (Caltech excepted due to size and tech nature).</p>
<p>number of state-to-state “wins” out of 21</p>
<p>Washington U in St. Louis 17.5
U of Chicago 16
Brown 14
Dartmouth 14
U of Notre Dame 14
Harvard 13.5
Cornell 13
Northwestern 13
Yale 11
MIT 11
carnegie Mellon 11
Columbia 10
Princeton 10
University of Penn 10
Emory 8
Vanderbilt 8
Johns Hopkins 7
Duke 6
Georgetown 6
Stanford 4
Rice 3
Caltech 3</p>
<p>collegehelp:</p>
<p>Those figures would indicate either a brain drain away from CA and the South (and Texas) of high caliber students ---- or that the brain drain is in fact away from privates and towards publics in the CA, TX, Southern states.<br>
Particularly considering that in the state-by-state data, it’s clear that the bulk of students choose to stay in or near their home state when choosing one of the privates from your list. </p>
<p>How come some of the colleges just have “#value” next to them instead of a number in the state lists?</p>
<p>“My algorithm controls for cost and distance leaving the other attractive forces.”</p>
<p>Well, no. The distance from Palo Alto (Stanford) to LA is farther than the distance from Boston to Philadelphia. Using state lines for sphere of influence will make universities sitting in tiny states look better. </p>
<p>“In CA and MA the percent of students who go to public unis is about the same.”</p>
<p>Actually, I’m not sure I’d consider 30% the same as 34%. But I am absolutely sure 10% versus 30% is not the same. And that is the difference between those going to a private institution in California versus those going to a private institution in Mass. And I believe that is much more relevant to these types of comparisions than the number of those attending 4 year public institutions. Especially given your assertion:</p>
<p>“Among the students who go to private unis, not all are really Harvard material. But that’s ok. Since we are using ratios, the only thing necessary is that the proportion of CA students attending private unis who are Harvard material is roughly the same as the proportion of MA private uni students who are Stanford material. The calculation should work fine.”</p>
<p>
actually, collegehelp, if you look at your data it does. CA has a much higher percentage of students going to “public associate” colleges than any of the NE states. Now obviously these shouldn’t count since they aren’t going to 4 year universities, but it means that when you look at the percentage of students going to 4-year universities who attend publics, CA is much higher. The easy way to see this is, as UCDalum pointed out, that the percentage attending privates is much higher in MA and CT (and even MO, despite your assertion that more students there go to publics) than in CA.
But, to be more comprehensive.</p>
<p>California
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 44.6%
pub uni pct 34.5%
percentage of 4-year attendees at pub unis: 77.3%</p>
<p>Institution Type Connecticut
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 71.6%
pub uni pct 32.5%
percentage of 4-year attendees at pub unis: 45.3%</p>
<p>Institution Type Georgia
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 56.7%
pub uni pct 44.7%
percentage of 4-year attendees at pub unis: 78.8%</p>
<p>Institution Type Illinois
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 58.8%
pub uni pct 34.3%
percentage of 4-year attendees at pub unis: 58.3%</p>
<p>Institution Type Indiana
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 77.6%
pub uni pct 59.0%
percentage of 4-year attendees at pub unis: 76.0%</p>
<p>Institution Type Maryland
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 61.2%
pub uni pct 33.6%
percentage of 4-year attendees at pub unis: 54.9%</p>
<p>Institution Type Massachusetts
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 68.5%
pub uni pct 30.1%
percentage of 4-year attendees at pub unis: 43.9%</p>
<p>Institution Type Missouri
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 60.0%
pub uni pct 41.5%
percentage of 4-year attendees at pub unis: 69.2%</p>
<p>Institution Type New Hampshire
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 78.2%
pub uni pct 38.3%
percentage of 4-year attendees at pub unis: 49.0%</p>
<p>Institution Type New Jersey
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 59.8%
pub uni pct 30.3%
percentage of 4-year attendees at pub unis: 50.1%</p>
<p>Institution Type New York
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 65.5%
pub uni pct 33.1%
percentage of 4-year attendees at pub unis: 50.5%</p>
<p>Institution Type North Carolina
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 51.3%
pub uni pct 39.4%
percentage of 4-year attendees at pub unis: 76.8%</p>
<p>Institution Type Pennsylvania
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 68.3%
pub uni pct 40.3%
percentage of 4-year attendees at pub unis: 59.0%</p>
<p>Institution Type Rhode Island
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 61.6%
pub uni pct 35.2%
percentage of 4-year attendees at pub unis: 57.1%</p>
<p>Institution Type Tennessee
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 62.6%
pub uni pct 46.7%
percentage of 4-year attendees at pub unis: 74.6%</p>
<p>Institution Type Texas
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 46.3%
pub uni pct 35.8%
percentage of 4-year attendees at pub unis: 77.3%</p>
<p>Institution Type Virginia
4-year uni percentage (sum of pub uni, pri uni, pri LAC): 59.6%
pub uni pct 44.1%
percentage of 4-year attendees at pub unis: 74.0%</p>
<p>So, when you only consider the appropriate set of students, those attending 4-year colleges, you find that CA and TX, not surprisingly are tied for second in percentage of students in public institutions, behind only Georgia, and far ahead of MA, CT, NH, NY, MD, RI, IL, PA, and NJ.</p>
<p>Well, no. The distance from Palo Alto (Stanford) to LA is farther than the distance from Boston to Philadelphia. Using state lines for sphere of influence will make universities sitting in tiny states look better.>></p>
<p>This has been pointed out to collegehelp numerous times, about how the algo disfavors large states such as CA and favors smaller states such as the New England ones. It was even pointed out when he wondered why Brown (RI) and Dartmouth (NH) had surprising strength.<br>
Frankly, c-help, if you just humored us and did your algo based to total college population of a state (which would be proportional to total population of the state for the most part, although I suppose it’s possible some states have younger or older populations than the nation as a whole) it would be more useful than arbitrarily cutting off the kids who wind up at public unis, which (once again) disfavors states with better public uni systems.</p>
<p>Where did you get the information about the number of students from each state? I cannot seem to find it anywhere. A link(s) would be much appreciated.</p>
<p>The information came from the IPEDS Data Center</p>
<p>[IPEDS</a> Data Center](<a href=“http://nces.ed.gov/ipeds/datacenter/InstitutionByName.aspx]IPEDS”>http://nces.ed.gov/ipeds/datacenter/InstitutionByName.aspx)</p>
<p>The system is pretty difficult to use but, if you can figure it out, you will find the information under “Fall Enrollment”, then under “Residence and Migration of First Time Freshmen”.</p>
<p>I downloaded a large file with data from many universities. I did not do this one college at a time. I imported (opened) the data file with Excel.</p>