Does college quality result from money or minds?

<p>Many criteria are used to judge college quality. I have often heard the following used for comparing colleges:</p>

<p>SAT scores
instructional expenditures per student FTE
admit percent

<p>I started with the assumption that college quality is reflected by graduation rates. I then calculated the correlations between graduation rates and each of the above criteria using about 1000 LACs and universities.</p>

<p>Admit percent and yield had very weak relationships with grad rate. Endowment and instructional expenditures per student had moderate relationships with grad rate. SAT scores had a very strong relationship with grad rate.</p>

<p>.83 SAT median
.53 instructional expenditures per student FTE
.41 endowment (private schools only)
-.28 yield
-.26 admit rate</p>

<p>I was really surprised by the correlation with yield. I double-checked all these correlations.</p>

<p>Minds matter more than money.</p>

<p>data source: IPEDS most recent year available</p>

<p>Faculty minds too...</p>

<p>The reason for the yield correlations: schools like ASU have some 80% or 90% yield (heck, they may have even counted me as going to their school, because I paid my deposit and went to orientation, despite choosing somewhere else once I got in). State schools tend to have high yield, because they're often the only school people apply to. I bet many state schools have a higher yield than Harvard, because if you expect to be a commuter student, and you're guarunteed admission (with ASU, given X test scores and Y grades, you can't be rejected), why would you apply anywhere else?</p>

<p>^ That explanation for the yield correlation sounds reasonable. It is in the opposite direction one would expect.</p>

<p>Student minds affect student success (graduation). Faculty minds affect faculty success (research). Faculty minds also affect student minds through their teaching but how much?</p>



<p>Why start there? I'm not saying you're wrong, but there are quite a few other outcomes one could start from. Examples: average starting salaries, professional school placements, PhD completions. Or, one could start from a perspective other than graduation outcomes. Examples: alumni giving rates, federal grant receipts, peer assessments.</p>

<p>Graduation rate is mostly correlated with family income.</p>

<p>^ only because smart parents make more money on average and smart parents have smart children</p>

<p>^and also because parents making more money don't typically require their child to work while in college, and they can send them to smaller private colleges where they'll have a better support network.</p>

<p>The 4 year graduation rate even at some very selective public universities (e.g. Berkeley, 64%) is much lower than the rate at many private colleges that are not only much more expensive, but also are not necessarily among the very most selective (e.g. Bates, 85%). I can only speculate why this is the case. Money surely is a factor in many cases. Selective private schools tend to have a high concentration of affluent students, and at the same time are more likely to meet 100% (or nearly 100%) of financial need. </p>

<p>I think another factor is differences in community life. Selective private schools usually are residential communities that pay attention to an applicant's potential to contribute to campus life. Public universities tend to be much more numbers-driven, and enroll more commuters or "suitcase" students.</p>

<p>I've observed among relatives, co-workers and neighbors that kids who attend directional state schools sometimes don't commit themselves fully to breaking away from the family nest and neighborhood. In contrast, kids at private residential schools often come to regard college as their new home. They quickly plug into a national social and economic network. At graduation, they are more likely to look well beyond the opportunities in their home states; they may be more motivated to graduate on time to take advantage of those opportunities. So I think that some differences of psychology and socialization, in addition to financial or academic quality factors, may account for some variation in the graduation rates.</p>

<p>One way to test what I'm suggesting would be to look at student body geographic distributions. Does the percentage of out of state students correlate strongly with the graduation rates?</p>

<p>"Graduation rate is mostly correlated with family income." "only because smart parents make more money on average and smart parents have smart children". "and also because parents making more money don't typically require their child to work while in college, and they can send them to smaller private colleges where they'll have a better support network." ---</p>

<p>All good points</p>

<p>^ also because as tuitions and parents plus loans grow year after year at the same time that real estate and limited investments go down in value (or a job loss) = possible transfer from private to state school. And thus lower graduation rates.</p>

<p>But interesting work...</p>

<p>Of the items collegehelp lists:</p>


SAT scores
instructional expenditures per student FTE
admit percent


<p>The SAT scores are heavily correlated to family income and parent education level; hence SAT scores are playing the role of a * de facto* measurement of family economic status, and as others have commented, family economic status is very tightly correlated to graduating on time.</p>

<p>Also, it's important to understand that "graduation rate" is typically measured as:</p>

<p><a href="number%20of%20first-time%20freshmen%20at%20college%20Y%20in%20cohort%20for%20year%20X%20who%20graduate%20from%20College%20Y%20within%206%20years">i</a>
(number of first-time freshmen at college Y in cohort for year x)

<p>So freshmen students who transfer * out * and graduate elsewhere hurt a college's graduation rate. And students who transfer * in * and graduate on time do NOT improve the graduation rate; they simply don't affect it in either direction.</p>

<p>Since the private schools with really, really high SAT scores typically have very few students who transfer out, this formula "helps" their graduation rate. And it hurts many public colleges for a variety of reasons, the three most important of which are "lots of students transfer out and graduate elsewhere", "lots of students transfer in and graduate on time", and "lots of students take more than 6 years to graduate."</p>

<p>For the publics where "lots of students take more than 6 years to graduate" the reasons can be related to students not being adequately prepared (hence likely to have lower SATs). Or it could be that some students need more than 6 years to graduate because of FA issues that cause them to "stop out" and work full-time for more than two years during their undergraduate career. Or it could be budget cuts at the college leading to fewer classes leading to harder to get into courses the student needs to graduate .... </p>

<p>Others have said such things as:

"only because smart parents make more money on average and smart parents have smart children".


There are plenty of people of average intelligence who make mucho bucks. The kinds of intelligence needed to be a sucessful CEO or entrepreneur are not the same as the traditional "smart" in terms of academic talent. The typical college professor is very smart, but hardly rich.</p>

<p>And while there seems to be some anecdotal evidence that "smart parents have smart childern", it's not as tight of a correlation as you might expect.</p>

"and also because parents making more money don't typically require their child to work while in college, and they can send them to smaller private colleges where they'll have a better support network." ---


<p>Concerning the first part of this statement: There is solid evidence that undergraduates who work 5--10 hours per week typically do better in school than their classmates who do not work at all. Particularly if that part-time job is campus-related. There is also solid evidence that undergraduates who work 20+ hours per week (typically at an off-campus job) do less well academically than their classmates who do not work. Students who try to work full time (35+ hours per week) and go to school full time (12+ credit hours per semester) are at high risk of not graduating on time or not graduating at all.</p>

<p>Concerning the second part of this statement: While the truly wealthy can afford to send their kids anywhere the kid can get into, the merely "well off" upper middle income folks (those making, say roughly $90K to $200K per year) often find themselves caught a real bind at expensive private schools and OOS publics: The student is accepted at Expensive College, but the family cannot afford their FAFSA EFC, and that's the minimum amount the college expects them to contribute each year. There are lots and lots of threads here at CC about this phenomenon.</p>

<p>My own take is that the correlation of graduation rates with SATs is doing nothing more than measuring the correlation between graduation rates and family economic status for the typical student. And students from families with annual incomes of above $90K a year are more likely to graduate and "graduate on time without transfering" because of the * family support system * both in the financial and emotional sense of "support system."</p>

<p>The relationship between money and minds (mental prowess) is probably a "chicken-and-egg" kind of phenomenon. Stronger minds generally have greater and more stable income. Higher family income enhances the environment for growing minds.</p>

<p>Aside from the money/mind issue among individual families, there seems to be a relationship between money and minds at colleges.</p>

<p>The correlation between endowment and SAT median among private schools is moderate (.55). The correlation between instructional expenditures per student and SAT median (public and private) is a bit higher (.64). Perhaps schools that spend more on instruction have better reputations and attract better students.</p>

<p>Again, this is based on about 1000 colleges and universities.</p>

I would actually say that family income is a de facto measure of SAT scores, not vice versa. The causal relationship is between academic ability (e.g. SAT) and graduation. Family income and graduation are both byproducts of being smart people. Even college professors make pretty good money in the grand scheme of things.</p>

<p>Family income issues probably help explain why the correlation between SAT and graduation is less than perfect but the fact seems to be that two-thirds of the explanation for graduation is academic ability (SAT) and one-third is due to other factors such as family income issues.</p>

I would actually say that family income is a de facto measure of SAT scores, not vice versa.


<p>The debate about * causation * vs. *simple correlation * between family income and SAT scores is open. Correlation has been clearly established, but whether coming from a family with high parental incomes * cause * students to score higher on the SATs is open to debate. But it's clear that scoring high on the SATs * does not cause * a student's parents to have high income.</p>

<p>But for the issue at hand: You have no direct data concerning family income at any of the universities you studied. Hence the SAT score data is serving as your * de facto * measure of family income in your study. In other words, that's the only variable you've studied that clearly correlates to the students' family incomes at a college. So if, by some chance, * average family income* is a more significant factor affecting a college's gradation rate than any of the other variables you studied, then it's no surprise that * mean SAT score* would be much more tightly correlated to * college graduation rate.*</p>

<p>You might want to argue that endowment somehow correlates to students' family income as well, but endowment figures can be drastically improved with a very small number of very generous donations. Percentage of alumni giving (which also improves endowment) is a measure of how well the students do * after * they graduate as well as a measure of how connected the alumni feel to the college. Neither is a measure of the income level of the students currently enrolled at the college.</p>

<p>Of the following factors, the factor that seems to be the most powerful predictor of graduation rate (college quality) is SAT math 25th percentile. </p>

<p>SAT scores
instructional expenditures per student FTE
admit percent

<p>SAT math 25th percentile alone correlates with grad rate .80. SAT median correlates .81. When you add instructional expenditures per student the joint correlation does not increase significantly (still .81). Student mental prowess alone seems to account for grad rates. When you also add admit rate and yield to the equation, the correlation only increases to .83. Admit rate and yield don't tell you much additional about grad rate (quality).</p>

<p>The above info is from multiple regression analysis using Excel.</p>

<p>Among universities with very high research activity (like MIT, Harvard, Berkeley), the correlation between SAT median and grad rate is .87 which means that the mental abilities of students accounts for about 76% of grad rate, a very high percent.</p>

<p>At private LACs, the correlation between SAT median and grad rate is also .87.</p>

<p>I have heard people say on CC that SATs don't matter, that SATs don't predict success in college. Some schools have even gone SAT optional. This is clearly wrong.</p>

<p>I have always found it irritating that people downplay student mind/academic ability in what makes a school great. I still think the strength of the student body is the most important part of the undergraduate experience in a college and is a good marker for the academics of the college as opposed to faculty research or faculty minds.</p>

<p>I agree with you, sefago. Student ability is the most powerful and valid index of college quality. I agree with UCBChemEGrad that faculty minds matter too but even a genius can't make a silk purse out of a sows ear. Great faculty minds can do their research but they can't significantly control student learning. Only the student can learn.</p>

<p>I also agree with tk21769 that there are other measures of college quality besides graduation rate but graduation rate is most relevant to the central mission of a university. What's more, factors such as starting salaries and job placement depend on so many other factors unrelated to college quality.</p>

<p>Endowments at private LACs are only correlated .34 with graduation rates. At universities with very high research activity, endowments have a moderate correlation with grad rate (.55) but that is probably because schools with higher SATs also have higher endowments. I don't think endowments have a direct affect on grad rates.</p>