financial factors and college quality

College Search & Selection
1

<p>repeating this post as a separate thread from the "30 Wealthiest Universities" thread (with a correction)</p>

<p>I did some statistical analysis to study the relationship between financial factors, enrollment factors, and quality factors. The numbers are for the list of 30 wealthiest universities from the thread by that name. The correlations do not imply causal relationships. (A typo in the original post said they did imply causal relationships.)</p>

<p>correlations between .2 and .4 are low
between .4 and .6 are moderate
between .6 and .8 are very high
between .8 and 1 are near perfect</p>

<p>financial factors:
total endowment
endowment per student (total endow divided by total enrollment)
endowment per undergrad (total endowment/ug enrollment)
instructional expenses (IPEDS)
instructional support expenses (IPEDS)
percent of undergrads getting university grants (IPEDS)</p>

<p>quality factors:
reputation score (US News)
graduation rate (US News)
SAT 75th percentile (US News)</p>

<p>enrollment numbers:
total enrollment
undergrad enrollment
percent undergrads</p>

<p>all the correlations were statistically significant
Here is what I found:
(1) As endowment per undergrad increases, SAT increases. very high corr=.75 The corrs were lower between SAT and endow/total (.70) and absolute endowment (.55)
(2) As endowment per undergrad increases, grad rate increases (.52)
with raw endowment dollars the correlation was lower (.40)
(3) As endowment per undergrad increases, US News reputation scores increase (.59) with raw endowment dollars the corr is lower (.51)
(4) As undergrad enrollment increases, instructional expenses per student decreases (-.57). This is the "economy of scale" factor.
(5) Instructional support expenses are more closely related to endowment than instructional expenses. .56
(6) The correlation between SAT and graduation rate was very high .85
SAT and reputation correlation was also high .72
(7) As the percent of students receiving institutional grants increased, the graduation rate increased .48 and SAT increased .60
(8) As instructional expenses increased, grad rate increased .41 and SAT increased .50 </p>

<p>I know this is heavy reading, but maybe some of you will find it interesting.
(1)college (2)endowment (last 6 zeros dropped) (3)instructional expenditures (4)percent of students receiving institutional grant(5)government subsidy per student for public universities only</p>

<h1>1 Harvard University, $22,600 $37150 $21654 94%</h1>

<h1>2 Yale University, $12,700 $60907 $12985 83%</h1>

<h1>3 Princeton University, $9,920 $30772 $9441 84%</h1>

<h1>3 Stanford University, $9,920 $41327 $13500 88%</h1>

<h1>5 Massachusetts Institute of Technology, $5,870 $7954 $3936 97%</h1>

<h1>6 Columbia University, $4,490 $42223 $3888 91%</h1>

<h1>7 University of Michigan-Ann Arbor, $4,240 $14759 $3569 81% $8911</h1>

<h1>8 Emory University, $4,090 $25707 $4923 84%</h1>

<h1>9 Washington University, $4,080 $63493 $8636 95%</h1>

<h1>10 University of Pennsylvania, $4,020 $26664 $21920 88%</h1>

<h1>11 Northwestern University, $3,880 $23453 $7126 91%</h1>

<h1>12 University of Chicago, $3,620 $35805 $0?? 84%</h1>

<h1>13 Cornell University, $3,310 $18505 $6867 79%</h1>

<h1>14 Rice University, $3,300 $29121 $4701 94%</h1>

<h1>15 Texas A&M University, $3,240 $10735 $1457 26% $9875</h1>

<h1>16 University of Notre Dame, $3,120 $19036 $3570 91%</h1>

<h1>17 Duke University, $2,830 $30061 $1951 92%</h1>

<h1>18 Dartmouth College, $2,450 $38951 $14228 100%</h1>

<h1>19 University of Southern California, $2,400 $18523 $1538 100%</h1>

<h1>20 Vanderbilt University, $2,260 $40997 $9466 94%</h1>

<h1>21 University of Texas-Austin, $2,040 $8488 $1858 61%</h1>

<h1>21 University of California-Berkeley, $2,040 $12566 $2348 61% $15179</h1>

<h1>23 University of Virginia, $1,980 $9787 $3546 90% $7302</h1>

<h1>24 Johns Hopkins University, $1,910 $57526 $1620 86%</h1>

<h1>25 University of Minn-Twin Cities, $1,770 $11849 $5923 65% $13056</h1>

<h1>26 Brown University, $1,670 $20082 $4957 90%</h1>

<h1>27 University of Cal-Los Angeles, $1,530 $20537 $6752 55% $17353</h1>

<h1>28 New York University, $1,470 $23429 $1262 100%</h1>

<h1>29 Case Western Reserve University, $1,470 $23261 $2148 93%</h1>

<h1>30 Williams College, $1,390 $81153 $18156 94%</h1>

2

<p>Since, in the COFHE survey, attending students ranked Harvard 26th in academic quality and quality of campus life (out of 31 schools - the majority of which are not on your list at all), your work is a nice exercise. ;)</p>

<p><a href="http://www.boston.com/news/education/higher/articles/2005/03/29/student_life_at_harvard_lags_peer_schools_poll_finds/%5B/url%5D"&gt;http://www.boston.com/news/education/higher/articles/2005/03/29/student_life_at_harvard_lags_peer_schools_poll_finds/&lt;/a&gt;&lt;/p>

3

<p>mini-
Thanks for your reply. True, the conclusions from my analysis apply only to the group of 30 wealthiest colleges (which includes Williams although Williams is an LAC).</p>

<p>Are the COFHE results available online? I could add that to my data and see what turns up. </p>

<p>I wish I had the time to add the top colleges from the COFHE survey to my data. It would be interesting to see if the same conclusions apply. It might show that universities and LACs operate according to different principles.</p>

<p>Graduation rate is a good quality indicator. It captures many attributes, both student qualities and college qualities. SAT captures student quality. Reputation scores from US News seem valid for the most part. But, student opinions (COFHE) certainly mean a lot. The problem with surveys is getting honest responses from a representative sample. (I am not familiar with COFHE procedures).</p>

<p>Can you give me the COFHE rank for the 30 in the list?</p>

4

<p>Sorry. COFHE is kept strictly private by the participating colleges. Many are afraid to make them public. Harvard's only hit the public screen because of a leaked memo.</p>

<p>Graduation rate is heavily a reflection of family income (and race). If you look at the publics, you quickly find that those with the highest graduation rates are those with the richest and least diverse student bodies (William and Mary is an excellent example.) If you want to use graduation rate as an indicator of quality, you need to correct for family income. A good, but very rough way to do that (for private colleges and unis) is to take the actual graduation rate, and subtract 28 (the highest Pell Grant percentage) minus the actual percentage on Pell Grants. (Example: Harvard at 97% - minus 21 (28 - 7) = 76. All off a sudden, the rankings look very different. (There are probably more sophisticated ways to do this, but I don't have the time.)</p>

<p>SAT captures student quality, but at points over 1200 very poorly. The CollegeBoard says that, in aggregate, a 1400 score is simply a 1200 plus $100,000 in family income. So you can better capture quality (among the privates) by taking the actual avg score and subtracting a factor based on the percentage of students receiving no needbased financial aid.</p>

<p>Reputation should count for something. But not those in USNEWS, where deans at the University of New Hampshire are evaluating the "reputation" of a school in Walla Walla, which they might not be able to find on a map. It will overly reward east coast schools, and those with graduate programs. The COHFE surveys sample more than 50% of each student body. Regardless, they are the consumers - when I buy a car, I look for consumer reviews, not comments from rival manufacturers.</p>

<p>The general point is that, to my way of thinking, the only really sound way to rank an undergraduate education is to look for the "value-added" - the "output" (minus what would have been expected of the entering student body generally speaking), rather than inputs. But to each her own.</p>

5

<p>mini-
It is ludicrous to say that low family income is the reason for lower graduation rates and lower SAT scores and therefore we have to adjust grad rates and SAT scores for family income. The primary reason for lower grad rates is lower SAT scores. The primary reason for lower SAT scores is lower student ability and motivation. </p>

<p>Adjusting SAT scores and grad rates for income and for pell grants is like saying the baseball team at Williams would beat the Yankees if you adjusted for the size of their muscles.</p>

<p>Student achievement (ability + motivation) comes partly from good upbringing and partly from familial traits (traits that run in families). To an extent, family income also results from these same factors.</p>

<p>By the way, I think talented children from low income families should be given as much assistance and support as possible to help them realize their potential.</p>

<p>I disagree with you on this issue, mini, but, as always, with much respect for your thoughts.</p>

6

<p>Wow. That's a lot of work, Collegehelp. Interesting reading.</p>

<p>Actually, if you wanted to test an hypothesis, just looking at financial underpinnings of LAC undergrad only colleges would serve you better because the endowments, revenues, and spending can all be isolated such that you are looking only at undergrads. That is impossible to do at most universities.</p>

<p>For example, here's the breakdown of Harvard University's endowment funds at the end of 2003 (when the total endowment was $19.2 billion) from the Harvard website:</p>

<p>
[quote]
Faculty of Arts and Sciences $8,743,246,000
Business School 1,330,368,000
Dental Medicine School 120,992,000
Design School 244,803,000
Divinity School 346,614,000
Education School 294,038,000
Kennedy School 582,141,000
Law School 927,308,000
Medical School 2,094,252,000
Radcliffe Institute for Advanced Study 289,939,000
Public Health School 633,753,000
Other Academic Departments 742,188,000
Central Administration 2,210,885,000
Service Departments 73,647,000
Pledge balances 445,244,000
Interests in perpetual trusts held by others 215,317,00 </p>

<p>Total $19,294,735,000

[/quote]
</p>

<p>The $8.7 million "Faculty of Arts and Sciences" endowment covers the business unit of the university that includes Harvard College (the undergrads). Even that is not a pure picture, because that endowment covers all grad school faculty, student spending, and reasearch in the arts, humanities, social sciences, and hard sciences.</p>

<p>It is still, of course, a very large endowment. But, given the large size of Harvard's undergrad enrollment, it is virtually certain that some of its competitors have larger endowments (and spending) as it pertains to each undergrad.</p>

<p>For another example. Let's assume that Emory nets $20,000 tuition revenue per undergrad (on the high side for elite colleges). That means that, as a corporation, Emory received $120 million in revenues from their undergrad business unit last year. Among the corporation's other revenue streams was $534 million from the sale of a single drug patent. Is undergrad enducation their main business? Or is it the relatively small consumer products division of a much larger corporation? When the corporation makes its spending decisions for the next budget year, do they invest more in the high revenue business units or the low revenue business units? See how tricky it is to look at University endowments and spending?</p>

7

<p>"The primary reason for lower SAT scores is lower student ability and motivation."</p>

<p>Well, the CollegeBoard would say you are correct for families in the same condition, and would disagree with you for families with differing financial conditions. If you have an argument, you should take it up with them. ;) Actually, though, the CollegeBoard doesn't even go that far. It makes no claim that the SAT is a reflection of student ability, only that it is a sound indicator of first-year college performance. That is the ONLY claim they make for it, so I don't know why you'd want to go further than they would. The California university study found that this was true, but just among white students - minorities signficantly outperformed what was expected in the first year, causing the Cal system to re-evaluate the use of SATs.</p>

<p>"It is ludicrous to say that low family income is the reason for lower graduation rates and lower SAT scores and therefore we have to adjust grad rates."</p>

<p>If you were poor, you would know that difficult family conditions are far and away the single most important factor in students leaving college early, or not finishing in four years. The data on that are unequivocal. But you can run the data yourself if you like.</p>

8

<p>interesteddad-
Someday I may try to repeat this analysis with LACs to minimize the "noise" in the endowment numbers. Refining the endowment numbers would tend to strengthen the relationships that I found. The "noise" in the endowment numbers probably adds randomness which would tend to lower the correlations. I thought it was interesting, too. Endowment per student seems to be more meaningful than total endowment. Endowment per student is closely associated with quality. Expenditures per student decrease as enrollment increases (economy of scale). Graduation rates are associated mostly with student SAT scores.</p>

<p>mini-
I am quite interested in helping people fulfill their potential. I agree that poverty is diabolically oppressive. I greatly admire resilient individuals who overcome obstacles to achieve success.</p>

<p>I am not familiar with the Collegeboard's positions on their own SAT test. The Collegeboard doesn't really have to say it; SAT scores reflect intellectual achievement and such achievement stems from the joint effects of ability, motivation, and opportunity. Poverty is detrimental to ability, motivation, and opportunity. </p>

<p>Children reared in poverty are more likely to be handicapped by the impoverished environment and high stress. The sad fact is they may never fully overcome their deficit and they may be rendered less able to do well on SATs and in college. </p>

<p>Familial traits are perhaps the other half of the reason for low achievement. Familial traits set limits on a child's potential. This is something about which little can be done except to create a more favorable environment so that a child's potential can be realized, whatever that potential might be.</p>

<p>The fact is, college is about learning. Whatever the reason, if you don't learn, you don't graduate. SATs indicate a student's liklihood of succeeding and graduating (note the very high correlation between SAT scores and graduation rate=.85).</p>