Graduation/Retention Rank-Is this Overrated by USNWR?

<p>A school’s Graduation/Retention Rate makes up 20% of its total score in the USNWR rankings. A closer look at the number reveals that 80% of the G/R rank is based on a school’s 6-year graduation rate and the other 20% relates to the freshman retention rate. </p>

<p>After Peer Assessment (25% of the score), the Graduation/Retention number (20%) is the 2nd-most heavily weighted factor in the USNWR calculations, equal to Faculty Resources and more than Selectivity (15%) and Financial Resources (10%)? This weighting is not intuitive to me as I would think that the quality of and resources dedicated to faculty and students would both be more important than G/R rank in measuring the excellence of a school. Can anyone help me better understand just what the G/R number really measures and why it should carry such a strong weight?</p>

<p>I agree. I never thought about it and never knew the percentages, but the G/R rate shouldn't influence something that much. </p>

<p>These are my reasons:
They don't take into account all the football players who graduate a year early for the NFL or play for four years but don't qualify for a degree. (grad rate)
They don't take into account freshmen basketball players who go pro after a year. (retention rate)
They don't divide into sections as to WHY a student transferred/withdrew after their freshman year. Some don't like the school, which I could see being used as a legitimate factor. But what about those students who want to change majors but their current school doesn't offer what they want? What if a student is having family problems at home and chooses to return home for a year or more to take care of things? They still affect retention rate even though it's not something that the school did wrong that made the student leave.</p>

<p>It also penalizes schools that have larger numbers of working and independent students. Not everyone has mom and dad write big checks for the easy 4 year ride.</p>

<p>I agree that athletes probably have some influence on this score's calculation although I suspect it is more than just a few players going to the NFL or NBA. Division I athletes (outside of the Ivy League) routinely redshirt for a year to get a physical and maturation advantage. I don't know what % of athletes do this, but it is not small. However, I don't think that the overall effect of this practice is so large as to have a major impact on a school's 6-year grad rate. </p>

<p>There are a number of top schools with 6-year grad rates at 90% or below. At some of the top privates, eg, Caltech has a 90% 6-year graduation rate, Emory 89%, Vanderbilt 88%, Carnegie Mellon 86%, USC 83%, etc. What could be the factors to cause these levels? I might expect such levels at the larger state schools as perhaps some % of their student population came there for financial reasons or have outside working responsibilities and these factors extend their attendance time at college. But what about some of these "elite" privates where this is likely not the case. Is it something that the schools are doing or what? Who is responsible for these comparatively low rates of graduation? Is there an explanation for this and is this correlated to the "excellence" of a college?</p>

<p>Yes, it's overrated and too heavily weighted.</p>

<p>It's even worse for schools like Caltech that are very challenging academically and enroll a highly qualified student body.</p>

<p>USNEWS uses a "predicted graduation rate" which essentially assumes that you should have a very high grad rate based on high SAT scores. So a school like CalTech that has incredibly high SAT scores has a ridiculously high predicted grad rate. They get docked signficantly for underperforming against their predicted rate -- a double whammy, docked once for the lower grad rate, docked again for missing their predicted grad rate. Guess what? CalTech is hard. Like nobody would have guessed, right?</p>

<p>Of course, the absurdity of it all is that CalTech could improve in the USNEWS rankings by replacing all those pesky physics and engineering seminars with basketweaving for dummies classes, giving everyone gentleman's As, and having a 98% grad rate.</p>

<p>Swarthmore gets docked on this too, despite the fact that their six year grad rate is above 90%. They get docked because their SATs are so high. They could improve their ranking by letting slackers graduate with D's instead of requiring at least a C in a course to get credit towards graduation. Would that make it a "better" school? No. It would actually make it a worse school, if you believe that students should actually be expected to study a little bit at an elite college.</p>

<p>Cream-puff curriculum, as Gerhard Casper put it, I believe.</p>

<p>I'm going to take a minority position and provide a partial defense of the graduation rate. Note, that doesn't mean that the use of it by the rankings is perfectly justified. Let's keep in mind that no ranking is perfect and no ranking category is perfect. They all have their flaws. The question is, does a particular ranking category provide more benefits than drawbacks? I would argue that in this case, it probably does. </p>

<p>
[quote]
I agree. I never thought about it and never knew the percentages, but the G/R rate shouldn't influence something that much. </p>

<p>These are my reasons:
They don't take into account all the football players who graduate a year early for the NFL or play for four years but don't qualify for a degree. (grad rate)
They don't take into account freshmen basketball players who go pro after a year. (retention rate)
They don't divide into sections as to WHY a student transferred/withdrew after their freshman year. Some don't like the school, which I could see being used as a legitimate factor. But what about those students who want to change majors but their current school doesn't offer what they want? What if a student is having family problems at home and chooses to return home for a year or more to take care of things? They still affect retention rate even though it's not something that the school did wrong that made the student leave.

[/quote]
</p>

<p>I agree. But you have to keep in mind that it's all relative. Schools are being ranked relative to each other. All schools with division 1-A football teams will have players leaving early for the NFL. All schools have some students who have family problems and therefore have to withdraw. The key is then, which schools seem to have more problems with retention/graduation relative to others? </p>

<p>
[quote]
It also penalizes schools that have larger numbers of working and independent students. Not everyone has mom and dad write big checks for the easy 4 year ride.

[/quote]
</p>

<p>I agree with this partially. But see below. </p>

<p>
[quote]
I agree that athletes probably have some influence on this score's calculation although I suspect it is more than just a few players going to the NFL or NBA. Division I athletes (outside of the Ivy League) routinely redshirt for a year to get a physical and maturation advantage. I don't know what % of athletes do this, but it is not small. However, I don't think that the overall effect of this practice is so large as to have a major impact on a school's 6-year grad rate.

[/quote]
</p>

<p>Let's also keep in mind that college athletics consists of far more than just football or basketball. Plenty of college sports basically have no real professional option and thus don't really have much of a problem at all with students leaving early to go pro. For example, there is no established and lucrative professional league for crew. Or field hockey. Or wrestling (no, the WWE doesn't count). Or gymnastics. Or fencing. Or cross-country. Or water polo. Yes, some of these sports have nascent professional leagues. But come on, college students aren't exactly champing at the bit to drop out of school to become instant millionaires as professional cross-country runners. </p>

<p>I seem to recall having read somewhere that the overall retention and graduation rate for all college athletes (for all sports) is actually HIGHER than it is for the college student body as a whole. Obviously, football and men's basketball (and to some extent baseball) tend to exhibit lower graduation rates. But most of the sports are not money sports. And athletes in these sports actually tend to do better than the average non-athlete. Hence, having a large sports contingent may actually make a school BETTER in terms of graduation/retention rate. </p>

<p>
[quote]
It's even worse for schools like Caltech that are very challenging academically and enroll a highly qualified student body.</p>

<p>USNEWS uses a "predicted graduation rate" which essentially assumes that you should have a very high grad rate based on high SAT scores. So a school like CalTech that has incredibly high SAT scores has a ridiculously high predicted grad rate. They get docked signficantly for underperforming against their predicted rate -- a double whammy, docked once for the lower grad rate, docked again for missing their predicted grad rate. Guess what? CalTech is hard. Like nobody would have guessed, right?</p>

<p>Of course, the absurdity of it all is that CalTech could improve in the USNEWS rankings by replacing all those pesky physics and engineering seminars with basketweaving for dummies classes, giving everyone gentleman's As, and having a 98% grad rate.</p>

<p>Swarthmore gets docked on this too, despite the fact that their six year grad rate is above 90%. They get docked because their SATs are so high. They could improve their ranking by letting slackers graduate with D's instead of requiring at least a C in a course to get credit towards graduation. Would that make it a "better" school? No. It would actually make it a worse school, if you believe that students should actually be expected to study a little bit at an elite college.

[/quote]
</p>

<p>This actually gets into a long debate that I have discussed with numerous other people on CC. And don't get me wrong, interesteddad. I sympathasize with your position. What you are describing is how the world should be. </p>

<p>But, it's not the way the world is. There's a big difference between the way the world should be and the way the world actually is. We have to deal with the world the way it actually is, not how we would like it to be. Sad but true.</p>

<p>Here's what I mean by that. I believe there is actually a good reason for Caltech to get docked for a graduation rate that is relatively low (when compared to peer schools like HYPSM). Basically, it serves to signal that Caltech needs to be more careful about who it admits. Why admit students who aren't going to graduate anyway? You're just wasting the school's time and the students' time. If the school insists on admitting him regardless, then, frankly, I think the school deserves to get punished with a lower ranking. The signal to me is fairly clear - if some students are unable to graduate because they can't handle the work, then you shouldn't have ever admitted them. </p>

<p>On a related note, this holds for students who can't graduate on time or don't get retained because they run into financial problems and need to work. The way I see it is, the school is ultimately responsible for providing sufficient financial support to every student it brings in. Why admit somebody who isn't going to be able to make it because he runs into financial trouble? Either fully support that student with a proper financial aid package. Or just not admit the guy at all. The worst choice is to admit him, only to have him drop out in the middle because of financial problems.</p>

<p>Now, I know what some of you are thinking. You're thinking that I am just being cold by advocating that schools not admit these students. To the contrary, I would argue that I am actually being compassionate. Like I said, the worst scenario is to bring in a student only to have him leave halfway without a degree. If you admit a student, you should be fully prepared to back him all the way to graduation. Or, don't admit him at all. All or nothing. The worst choice is that "intermediate" choice where you admit him but don't fully support him. Think about it. If he's not going to graduate, then, frankly, he's just wasting his time and wasting the school's resources. </p>

<p>Not only that, but he's just wasting his money. Again, let's talk about the guy who has to drop out because he runs out of money. That guy had presumably spent some of his money for those years during which he was in school. When he drops out, he's not going to get any of that money back. So think about what we have here. We have a guy who didn't have much money to begin with, and now he has ** even less ** money, and he ** still ** has no degree. Hence, I would argue that this guy is even worse off than if he had never gone to college at all. Keep in mind that you don't just go to college just for the heck of it. You go to college to get a degree. If you're not going to get a degree, you might as well have not even bothered to go at all. </p>

<p>Hence, I would actually argue that it is the compassionate thing for a school who can deduce that an applicant isn't going to make it to graduation (whether for academic or financial reasons) to simply not admit the applicant in the first place. Everybody would then be better off. I agree that it's not pleasant to be rejected from a school. But it's even worse to go to that school and then later have to drop out. </p>

<p>Let's look at it from the student's perspective. I think that most students, like myself, are highly risk-averse when it comes to their degree. I remember going through the trauma of a harsh weeder undergrad program in which I and others in my program weren't even assured that we were going to graduate at all, and in which the pressure to avoid flunking out was ever-present. Trust me, it is a most unpleasant experience that I wouldn't wish on anybody. I remember thinking to myself what I would say to my family if I flunked out. That was pure hell, because you can't plan for anything if you don't even know if you are going to graduate. That pressure to avoid flunking out is always hanging over your head, clouding your entire experience and making it impossible to enjoy anything. Later on in life, I went to a graduate program where finishing the program was basically assured. You might not get top grades, but you are basically assured that if you do the work, you would get the degree. Contrast that with undergrad in which you can do all the work, and flunk out anyway (because it was a curved weeder program). Judging from that experience, I would argue that there is a tremendous amount of value in knowing that if you get admitted to a particular program, that you are basically assured of graduating (as long as you put in a good-faith effort to do the work). Hence, I think that schools probably are well-advised to only admit those students who they honestly believe will make it, and not admit those students who are doubtful. After all, if you're going to weed your students anyway, I would argue that it's more compassionate to do it through the admissions process than through the curricula. That way, you have saved people time and money. Not only that, but if the student is simply not admitted, then the student can just go to some other school that did admit him (and presumably does believe that he will make it). In contrast, if a student comes to your school and then flunks out, he is screwed, because no other reputable school wants to admit a transfer student who flunked out of his previous school. Such a student is far far worse off than if he had simply never been admitted at all.</p>

<p>{Now, of course, one might say that the student should not have chosen to enter a program that was just too hard for him. I agree that the student bears some responsibility. But, again, let's not be too harsh. Let's have some compassion here. These guys are just 17-18 year old kids. People of that age do foolish things all the time. Hence while these kids bear responsibility for making bad choices, schools also bear responsibility for allowing them to make bad choices. A better solution for all parties involved is to simply not provide that bad choice.} </p>

<p>Besides, look at it this way. Like it or not, the market for labor is far from perfect. In particular, it rewards the fact that you have a degree far more than how difficult it was to get that degree. Somebody who graduated from Fresno State is valued more highly in the market than somebody who went to Caltech and flunked out. That's because, like it or not, most decent jobs out there require that you have a degree to even get an interview. In many cases, the employers don't care very much about how difficult your school was, or (sometimes) even what you majored in. What matters is that you have a degree. A guy who flunked out of Caltech may well be better than a guy who barely graduated from Fresno State, but employers don't know that and don't care. All they care about is whether you have a degree. No degree, no interview. </p>

<p>{Now, of course, I exaggerate. It's not that strict of a dichotomy. But I think you get my point. In this labor market, there is a big difference between having a degree, even if it's in a creampuff major at a no-name school, and not having a degree at all.} I think schools like Caltech ought to be more cognizant of just how damaging it is to somebody's marketability to flunk him out. Right now, as it stands, those who went to a school and flunked out would probably have been better off if they had never gone to that school at all. </p>

<p>Bottom line - schools should either do whatever they can to help as many of their students graduate as possible. Or, just don't admit those students who aren't going to make it anyway. A school that insists on bringing in students who aren't going to graduate is probably rightfully dinged by the rankings. What they're doing is bad for the students, bad for the school, bad for everybody.</p>

<p>But back to the issue of rankings. Keep in mind that rankings like USNews are basically 'consumer information goods' that provide information to prospective students regarding what schools they should be preferring. Take the notion that if I get admitted to a particular school and choose to go there, then I am highly assured of graduating. That, to me, is a highly piece of information and is a legitimate reason to prefer one school over another. Again, this has to do with man's natural risk-aversion. It's the same reason for which psychologists have discovered that most people will take an choose a guaranteed $400,000 vs. a 50% chance of getting $1 million (and a 50% chance of getting nothing), despite the fact that the expected value of the latter is higher. People place value on certainty and want to reduce their risk. By going to a school like Caltech, as opposed to a peer school like HYPSM, you are basically taking a higher risk of not graduating. Given man's inherent risk aversion, I think it is reasonable to place less value on the riskier choice.</p>

<p>Just for my own knowledge: If a student transfers from a university, is he counted against the graduation rate?</p>

<p>By the way, interesting post sakky. I hadn't considered some of your arguments.</p>

<p>sakky,
I think we are confusing some aspects of the USNWR calculation. Your post referred mostly to the difference between the actual graduation rank and the expected graduation rank (which is based on standardized test scores). But this is actually a different component of the methodology and carries only a 5% weighting. I think you make a good case that this calculation has some value and I would agree with USNWR's 5% weighting. It's important, but not incredibly so.</p>

<p>My question-which gets ever more interesting to me because no one has much of an idea why-remains why is the Graduate/Retention rank 20% of the calculation for USNWR? Using the subfactors, we know that 16% of a school's score dedicated to its 6-year graduation rate and another 4% is dedicated to its freshman retention rank. These numbers likely merit some weighting, but to this degree? I am really puzzled by this as I have a hard time understanding why this number is considered by USNWR such a huge part of what constitutes a top national university.</p>

<p>I'm not sayinig that I buy it, but I can at least see the argument for using this measure. First, you've got to assuume that the percentage of students leaving school for unavoidable reasons like illness, to take care of a family members, etc. is pretty uniform and turns out to be a wash.</p>

<p>So graduation/retention rates reflect, however imperfectly: a) student satisfaction; and b) the quality of a support network that would include things like: advising, mentoring, access to tutoring, access to medical and psychological help, willingness to reconsider FA in light of changing family circumstances, availability of required classes, and more.</p>

<p>Graduation rate doesn't quite capture everything, but graduation rate captures a great deal about the quality of a college or university.</p>

<p>Graduation rate primarily reflects student quality. There is a strong relationship between average SATs and graduation rate. Student quality, in turn, reflects selectivity, the level of teaching, culture and climate, faculty quality (good faculty and good students are mutually attractive to each other), reputation, and so on. </p>

<p>Graduation rate is also a reflection of the quality of the school. The higher the SATs, the higher the expected graduation rate. If a school's graduation rate exceeds the expected graduation rate, it suggests that the school is doing a good job with the students who enroll there. This is indicated by US News overperformance/underperformance index.</p>

<p>The US News predicted graduation rate is based on SATs, high school rank, and expenditures per student. The actual formula changes somewhat every year. There are different formulas for publics and privates and for liberal arts colleges. US News uses whatever works best that year.</p>

<p>Expenditures per student functions in a counter-intuitive way: The higher the expenditures the lower the graduation rate. Technical majors require the highest expenditures. They are also the most demanding majors. That is my interpretation, anyway.</p>

<p>Retention from freshman to sophomore year and graduation rates after 150% of normal time to complete a degree are the standard measures of "retention" for universities. They reflect somewhat different factors about the school. Retention after one year is affected most by student quality. Graduation rate after 150% of normal time is affected more by student quality, school quality, and program quality.</p>

<p>Relatively few students transfer to different colleges and universities because they want to change their major and their current school does not offer it. This happens more often at schools with a narrow focus. Most transfer for other reasons. They vote about the quality of the school with their feet. They transfer because their school is not worth the money, they are disappointed in their experience, they want a different location, and so on.</p>

<p>This is what I hear from people I know who work at universities.</p>

<p>graduation/retention rank is pretty important. It basically reflects the quality and performance of the attending/admitted students.</p>

<p>
[quote]
Here's what I mean by that. I believe there is actually a good reason for Caltech to get docked for a graduation rate that is relatively low (when compared to peer schools like HYPSM). Basically, it serves to signal that Caltech needs to be more careful about who it admits. Why admit students who aren't going to graduate anyway? You're just wasting the school's time and the students' time. If the school insists on admitting him regardless, then, frankly, I think the school deserves to get punished with a lower ranking. The signal to me is fairly clear - if some students are unable to graduate because they can't handle the work, then you shouldn't have ever admitted them.

[/quote]
</p>

<p>But, it's not an admissions issue. In CalTech's (or Swarthmore's) case, we are only talking about a difference of maybe ten students from each freshman class. And, the difference is that those ten students can't squeek by on a gentleman's C- gut track. There's nowhere to hide. By the time sophmore year rolls around and it's time to declare a major (requires B's in the intended major at Swat, I don't know about CalTech), a decision point has been reached.</p>

<p>I think that a equally strong argument could be made: An elite college that isn't too demanding for at least a few students is too easy and should stiffen up its academic requirements. We know that "x" number of students at all of these schools are slackers. It's not the students who make the effort that are flunking out. I can't speak for CalTech, but I know that, at Swarthmore, if you go to class, make an effort, use the resources, you are going to graduate. The weeding process only culls the students who aren't even making a minimum effort, like not even going to class.</p>

<p>The other flaw with the USNEWS approach is that it docks schools for students who transfer out, but does not count the students who transfer in. If you factored the two-way impact of transfers, all of these elite colleges would be at essentally 100% six-year grad rates.</p>

<p>I'm not saying that looking at graduation rates isn't worthwhile. It's most instructive in the middle of the bell curve where you see some real variation. At the top of the curve, is there really any difference between 92% and 96%? I don't think so.</p>

<p>HYP do not like to flunk student out, you do not want to flunk out a future president. A gentlemen's C is acceptable. So their graduation rate is higher than they should be and give them an artifical boost in the USNWR ranking.</p>

<p>I completely agree with Sakky, as far as he goes, but would add another point. </p>

<p>As collegehelp notes, some students transfer because they want a major the current school does not offer- comparative literature for a student at Caltech. Every year some students at elite colleges realize that they are not as interested in science, technology, engineering, and math and as they thought. For a student at Stanford or Princeton, one simply starts taking other courses and moves on toward a degree. For students at Caltech, a move away from STEM means a transfer application. So it would be nearly impossible for a college that offers only a narrow range of majors to have a graduation rate similar to Harvard's. </p>

<p>This is a major reason the graduation rate information is useful to students. It says to them "think about what you will do if you decide you don't want to major in the areas offered by your narrowly focussed college". As Sakky says, one expects only a limited amount of foresight from high school seniors. A warning in this regard is helpful to them.</p>

<p>Interesteddad, I would argue that Swarthmore is a generalization of the same pheonmenon. Some people get there and discover that they do not really want what the college offers: an intense pre-graduate school liberal arts education, with no alternatives. Just as Caltech-style science, science, and more science is not for everyone, neither is starting graduate school right out of high school (an overstated view of the Swarthmore experience, for the sake of this argument). As for Caltech, high school students should consider what they will do if they decide, as many, many college students do, that college will be the end of their formal education, that their extracurriculars are more important than their coursework, that there is more to life than studying?</p>

<p>The vast majority of elite colleges offer a wider range of college experience than do Caltech or Swarthmore. So they accomodate a wider range of student academic and personal goals. An outstanding high school student who may want to get a research doctoral degree someday, but also has enough insight to know that at age 17 she does not really know what she wants to do may be much better off at Yale than at Caltech or Swarthmore.</p>

<p>
[quote]
An elite college that isn't too demanding for at least a few students is too easy and should stiffen up its academic requirements.

[/quote]
</p>

<p>Why? I don't see any more Harvard grads than Swarthmore grads struggling to make a living. How are the Swarthmore grads any better off? Everyone who gets into these places can do the work, it is then the responsibility of the college to provide the support to help them do so.</p>

<p>Ultimately, places with an intense and narrow focus, like Swarthmore and Caltech, are going to have some leakage of students. That's just the way it is. USNews, for all its numerous flaws, is doing prospective students a service by pointing this out.</p>

<p>Collegehelp,
I have read and reread your thought provoking post and I am having difficulty swallowing several of the comments and conclusions:</p>

<ol>
<li>“There is a strong relationship between average SATs and graduation rate. The higher the SATs, the higher the expected graduation rate.”</li>
</ol>

<p>I understand the conclusion and, on its face, I might accept it, but there are dozens of examples that contradict this conclusion. Caltech is the most obvious contradiction, but there are many others throughout the rankings where their SAT scores and grad rates are in opposition. Selected examples include Wash U (1360-1520/91%), Carnegie Mellon (1290-1470/86%), Brandeis (1270-1440/88%), U Rochester (1250-1420/80%), Georgia Tech (1250-1440/76%), etc. In fact, the schools with an 80% 6-year graduation rate have a huge variance for SAT scores-U Rochester, Pepperdine (1120-1320), UC-Irvine (1110-1310), UC-Davis (1060-1300), and Miami of Ohio (25-29 ACT). I could go on and cite many other examples contrary to the thesis that high SAT scores mean higher quality students and better 6-year graduation rates. </p>

<ol>
<li>“Expenditures per student functions in a counter-intuitive way: The higher the expenditures the lower the graduation rate.”</li>
</ol>

<p>I have always assumed two things about expenditures per student. First, the technical schools must spend more due to technology and medical research needs. Second, the state schools are at a severe disadvantage in this measurement as they generally have more students and they normally get much of their money from the state and this funding has been under pressure for some time now. </p>

<p>Your statement is that a high Financial Resource rank (expenditures per student measure) would correlate with a lower graduation rate. But the actual data don’t lead me to that conclusion. For example, Yale is ranked 2nd in Financial Resources and 3rd in graduation rate. And there are many examples of schools with very similar Financial Resources ranks, but very different 6-year graduation rates (Notre Dame at 39th in FR and 95% grad rate vs. Pitt at 37th in FR and 70% grad rate. Such inconsistencies are found in abundance when you look at the raw data.</p>

<ol>
<li>“They transfer because their school is not worth the money, they are disappointed in their experience, they want a different location, and so on.”</li>
</ol>

<p>I probably agree with your statement, but I am not sure that I agree with the accompanying idea that this relates to the quality of an institution. People transfer for a variety of reasons, not least of which is that the student misjudged his or her “fit” at a school. Some at a more prestigious school might gut it out for four years because the students hopes that the brand value of the diploma might make the struggle worthwhile, but does this really mean that the prestigious school really is better? For example, a student from North Carolina is accepted at Yale and U North Carolina. She goes to Yale and hates it (for whatever reason). Will she transfer back to North Carolina where she would likely be much happier and would still get a very good education? Probably not because the brand value of Yale is appreciably stronger than UNC. But should such a decision lead one to conclude that Yale does a better job of educating and graduating its students vs what UNC does? </p>

<ol>
<li>“US News uses whatever works best that year.”</li>
</ol>

<p>I’m not sure what you mean by this. One could interpret this to mean that the data is massaged and weighted in such a way to generate an acceptable result. In other words, USNWR has preordained ideas of what the rankings should be and thus adjust their methodology annually in an attempt to perpetuate the status quo. As a longtime skeptic of USNWR and also one who believes that there is a large amount of underrated talent outside of the NE and CA, I could certainly believe that this is true. But I don’t want to get into this argument at this point as I don’t want to distract from the discussion about Graduation/Retention scores, rankings and weights.</p>

<p>
[quote]
Everyone who gets into these places can do the work, it is then the responsibility of the college to provide the support to help them do so.

[/quote]
</p>

<p>The support to help them do the work? Or the support to let them graduate without lifting a finger or attending a class? If the purpose of an elite college is to simply sell degrees, then they could dispense with classes altogether. Just cash the check and mail 'em the diploma.</p>

<p>I have no beef with USNEWS including grad rates and even their arbitrarty "expected grad rates" in the extensive pages of data for each college. Heck, USNEWS provides much more useful data than grad rates that isn't included in the rankings at all. Why not included "diversity" and double-dip by comparing to an "expected diversity" based on financial aid budgets? Why not include financial aid and double-dip by comparing against an "expected financial aid" budget based on endowment size?</p>

<p>The answer? USNEWS needs grad rate to make sure that state universities don't beat out old-money privates at the top of the rankings, since it is only at the old-money privates where paying for college is not a significant determining factor in graduation rates. </p>

<p>But, by weighting graduation rates heavily in a ranking system, USNEWS is actually rewarding colleges with less rigorous academic programs and docking colleges with notably challenging academics. That doesn't make any sense. Do we want to encourage CalTech to replace physics with basketweaving? [note: I would actually not rank tech schools and all-purpose schools together.]</p>

<p>hawkette-</p>

<ol>
<li>“There is a strong relationship between average SATs and graduation rate. The higher the SATs, the higher the expected graduation rate.”</li>
</ol>

<p>You said: "the schools with an 80% 6-year graduation rate have a huge variance for SAT scores-U Rochester, Pepperdine (1120-1320), UC-Irvine (1110-1310), UC-Davis (1060-1300), and Miami of Ohio (25-29 ACT)"</p>

<p>My reply: these don't seem like such large differences in SAT scores so I am not surprised their graduation rates are about the same</p>

<p>You said: "Caltech is the most obvious contradiction, but there are many others throughout the rankings where their SAT scores and grad rates are in opposition. Selected examples include Wash U (1360-1520/91%), Carnegie Mellon (1290-1470/86%), Brandeis (1270-1440/88%), U Rochester (1250-1420/80%), Georgia Tech (1250-1440/76%),"</p>

<p>My reply: The correlation between average SAT and graduation rate is very high, about +.85 to +.90. This means that, if you placed schools on a graph of SAT by graduation rate, they would form a very narrow diagonal oval kind of scatter plot....not quite a straight line. If you drew a line down the middle of the oval, some schools would be above the line (overperformers) and some schools would be below the line (underperformers). </p>

<p>There would still be some variation at each SAT level but not much. The variation in graduation rates among schools with similar SAT scores would be due to things like:
(1) the mix of programs-schools with a high proportion of math, science, engineering, comp sci, & econ students tend to have somewhat lower graduation rates. e.g. Caltech, U Rochester, and Carnegie Mellon
(2) upward mobility-is there a more desireable competitor school to which top students might transfer?
(3) debt burden-how much financial support does the school offer?
(4) academic quality, social experience, faculty support, quality of instruction, school spirit, and so on-these quality factors can lead students to the conclusion that they are disappointed in their choice</p>

<ol>
<li>“Expenditures per student functions in a counter-intuitive way: The higher the expenditures the lower the graduation rate.”</li>
</ol>

<p>You said: "Your statement is that a high Financial Resource rank (expenditures per student measure) would correlate with a lower graduation rate. But the actual data don’t lead me to that conclusion."</p>

<p>My reply: Actually, I was referring to a figure that the school's accounting office reports to IPEDS each year. US News uses a two-year average of this figure taken from the first two school years of the current graduating class. It refers to the actual amount of money spent on academics or academic support in those first two critical years of the current graduating class. It is different from Financial Resources Rank. </p>

<ol>
<li>“They transfer because their school is not worth the money, they are disappointed in their experience, they want a different location, and so on.”</li>
</ol>

<p>You said: "I probably agree with your statement, but I am not sure that I agree with the accompanying idea that this relates to the quality of an institution."</p>

<p>My reply: I think a high percentage of students transferring out would itself detract from the quality.</p>

<ol>
<li>“US News uses whatever works best that year.”</li>
</ol>

<p>You said: "I probably agree with your statement, but I am not sure that I agree with the accompanying idea that this relates to the quality of an institution."</p>

<p>My reply: What I meant was that US News uses whatever graduation rate prediction formula does the best job of accurately predicting the actual graduation rates that year. This kind of tweaking is considered acceptable procedure in the world of predicting things. If you use the data from 1000 schools to come up with a prediction equation and then apply it back again to the 1000 schools, you can see how accurately the prediction equation works compared to the known graduation rates. The formula is tweaked to minimize the actual prediction errors.</p>

<p>Thanks for your interesting observations....</p>

<p>Collegehelp,
I am still trying to process all of your comments, but a few clarifications and follow-ups:</p>

<p>1.Please reread my prior posts. For schools with an 80% graduation rate, the SAT scores are as follows:
University of Rochester (1250-1420)
Pepperdine (1120-1320)
UC-Irvine (1110-1310)
UC-Davis (1060-1300)</p>

<p>Do you really believe that these scores are not very different? I would disagree as Rochester looks a lot higher than UC Davis (or the others) to me.</p>

<p>2.With regard to your correlation analysis for SAT and graduation rates, these are directionally correlated but not scale. I would expect the directional correlation to be very high. But your qualitative comments on why the results would vary at different SAT scores make intrinsic sense to me. </p>

<ol>
<li><p>I was comparing Financial Resources with Graduation Rates in response to your prior comment that these are inversely correlated. I think the data say something else. </p></li>
<li><p>With regard to the number of transfers, is this number reported anywhere? It would be interesting to see just what these numbers are. I agree that a large transfer percentage is not desirable, but I hesitate to accept your conclusion as I don’t know the absolute or relative number of transfers. Furthermore, while transfers aren’t beneficial to a school, the students probably bear some of this responsibility and I still struggle to see why this action plays such a major role in evaluating the quality of a school. Some role? Yes, but not 16% (which is the amount of weight accorded to 6-year graduation rates).</p></li>
<li><p>I think you mixed my comments about USNWR using whatever works best. No harm, but it sounds like creating a hypothesis and then testing it using data and weights that prove it. What am I missing? </p></li>
</ol>

<p>The more that I think about, the more I believe that the brand effect is underappreciated as it relates to Graduation/Retention rates. When I began this thread, I didn’t expect to consider a college’s brand as such a large factor in determining a school’s Graduation/Retention rate. With the exception of the technical schools (which have their own unique challenges in retaining very high percentages of students) and a couple of the top publics, the G/R ranks look very much like a survey on which private schools have the strongest brands.</p>