Diplomas and Dropouts - Which Colleges Actually Graduate Their Students (and Which Do

[vossron]

"A far more efficient method for all parties concerned is for the colleges to simply not have admitted those students in the first place. Why admit somebody who isn't going to graduate anyway? You're just wasting everybody's time - the school's and the student's."

If only those students could have been identified in advance! Every school would like this crystal ball!

[MSmom&dad]

If only those students could have been identified in advance! Every school would like this crystal ball!

I don't think that's true - there are lots of colleges that woud have to close their doors were it not for the tuition they receive because of the revolving door admission of unqualified freshmen

[psych_]

Actually, some stats DO track the outcomes of transfers in and out. I think it's called the VAC or VAS survey...

[vossron]

^^ Which ones practice this revolving door? Can you post the list?

[psych_]

It's the VSA (Voluntary System of Accountability). Here's the participating schools:

Voluntary System of Accountability

[newmassdad]

As to how a school would know who is unlikely to graduate, one can invoke the same statistical regression analysis that insurance companies use to underwrite premiums. For example, if smoking is statistically correlated with high medical expenditures, then health insurance firms respond by charging higher premiums to smokers, or not even offering insurance at all.

This is a pretty naive approach. One needs to consider such issues as cross-correlation of variables, the degree of variance a variable explains and so forth.

In the case of deciding whether a kid will graduate, I suspect it involves variables such as "motivation" and "discipline" (whatever those truly mean) that are impossible to measure with accuracy. Worse, things change over time.

After all, how many of these students that don't graduate start out planning to not graduate? Not many, I suspect. Then consider how many that don't graduate fall victims to things outside their control? Quite a few, I suspect.

One thing today's economic environment is bringing out is how fragile the economic status of many families is - all it takes is one illness or one lay off for a family to move from middle class, where college is affordable, to a class where college is not affordable. How do you quantify that in advance? Just not admit kids whose family incomes are below average, since that is where the risk is greatest?

[northeastmom]

Interestedad, I agree. I am repeating part of the explanation that I was given when I called up a state public U. and asked them why their 4 year graduation rate was so low. I just decided to ask CC particants if they thought 3/2 programs are a part of the reason. Part of the answer I was given was because of the 3/2 programs. The other reason given was that many students are just not in any rush to graduate and they choose to take a part time course load.

[Erin's Dad]

My wife worked for the Inst Research dept at a major U for a couple of years collecting this data (among other info). The 6 year timeframe is used to avoid penalizing universities for students who change majors. Changing majors would affect LACs less since the requirements for specific majors are usually more flexible. What it does in actuality is allow some U's to get away with not scheduling classes often enough so students CAN complete in 4 years.

Other reasons that students take longer than 4 years are varied - working part time, having enough funding to finish school (haven't we recommended people stop school to earn money and return later?), the rigor of studies (in my mind one of the reasons Reed has a low rate), lack of institutional support to assist students in passing. I note that one of the underperforming schools in the Competitive category is Central State U, a historically black university. Unfortunately, most HBUs are well behind in graduation rate.

[epiphany]

A far more efficient method for all parties concerned is for the colleges to simply not have admitted those students in the first place. Why admit somebody who isn't going to graduate anyway?

....which is why, if you're a Pell grantee admitted to a U with a single-digit-admit rate and generous FA, the school has already made a calculated investment in you based on your past outstanding performance and the extreme likelihood of your consciousness of this unusual and monetarily valuable opportunity. You are a low-risk admit with a high probability of graduating within the ceiling of academic-unit cutoff for FA, because that right there is your motivator.

It's amazing how much mythology there still is on cc about FA recipients. (They're losers; they'll drop out; they should not have been admitted. Wrong.) In fact I'll bet it's just the opposite: the non-graduates in my experience (anecdotal only) tend to be from the class of people with little to lose. Non-anecdotal (statistical) tendencies in this regard might be borne out by the UC's ELC study of academic performance after admission (side by side comparison of rich ELC admits with poor ELC admits). The study did not follow the 2 contrasting groups all the way to graduation, but it would not surprise me if rich students starting out with C's as freshmen ended up with some F's as upperclassmen -- or simply lost their interest/motivation by that time, since partying was their preferred activity.

So we need more controls than just transferring rates (I agree, also important). Broad "graduation rates" or even 4-year rates are not meaningful without attention to the rigor of the institution, the academic material going in, and the degree of generosity when it comes to FA.

[jbusc]

Putting so much emphasis on 4-year graduation rate is harmful, because it encourages schools to put arbitrary restrictions on students just to inflate some statistic.

For example, the opportunity to take a lot of classes outside your major, or to change your major freely, or take one fewer class than typical occasionally, or any number of legitimate things that might make a student graduate in 5 years instead of 4. Modulo money/tuition/fin aid issues, there is not any reason to restrict any of those - except that they can harm the 4 year graduation rate.

[sakky]

While in general I agree that colleges should admit only those likely to graduate, I do think some institutions have as part of their mission giving students with imperfect records the opportunity to succeed. They recognize that many of these non-sure-thing admits will fail to graduate, but if a portion of them do the school will be fulfilling its mission and the successful students will have profited from the lenient admissions policy.

Sure, but at the same time, students should then rightfully view those schools as being less desirable places to study. Why take a chance on going to a school of which the strong possibility exists that you won't graduate (unless it's the only school you can get into)? Let's face it - taking that chance is not costless. Not only will you still have to pay tuition to the school that you won't actually graduate from, but you are also likely to damage your permanent academic record, hence impinging your chances of getting into any other school. No respectable school wants to admit somebody as a transfer who flunked out of his previous school.

As a case in point, I know a guy who went to Berkeley and flunked out, and then couldn't even go to UCDavis - a school that he got into as a high school, for UCD didn't want to take in a transfer applicant who had flunked out of Berkeley. In other words, the guy would have been better off if he had not even gone to Berkeley at all and had just gone straight to the workforce, for then he would still be able to apply and surely be admitted to Davis as a freshman.

[sakky]

This is a pretty naive approach. One needs to consider such issues as cross-correlation of variables, the degree of variance a variable explains and so forth.

In the case of deciding whether a kid will graduate, I suspect it involves variables such as "motivation" and "discipline" (whatever those truly mean) that are impossible to measure with accuracy. Worse, things change over time.

After all, how many of these students that don't graduate start out planning to not graduate? Not many, I suspect. Then consider how many that don't graduate fall victims to things outside their control? Quite a few, I suspect.

So control for all those variables. Make the model as complex as you wish. What I think is naive is thinking that it can't be done. After all, colleges are sitting on mountains of data of past students, just waiting to be mined.

Besides, I do not pretend that the analysis has to be perfect, nor does it need to be. To demand that it be so is naive. All it needs is to be better than the current system, and let's face it, the current system ain't that great. You complain that my proposal doesn't control for a host of variables that are impossible to measure with accuracy, but let's be perfectly honest - neither does the current system. Does the current admissions process measure motivation and discipline with any accuracy? Does the current process correct for time variance? The current system is remarkably unsophisticated - usually heavily weighted on high school GPA and standardized test scores, with some essay questions, interviews and teacher rec's thrown in. It uses very little past regression data from former students. Why not? It can't only make the present system better. Let's be honest - it's not that hard to do better than the current system.

Sure, my proposed system would still admit some students who probably shouldn't be admitted and reject some who should be admitted. But that happens right now under the current system.

So, to those who would continue to object to my proposals, I leave you with what I think is an impossible task - namely, defending the current system. After all, that is in effect the choice we have at hand - to take my system, or to stay with the current (badly flawed) system. Why do you guys insist that the current system is so great? I think that's an indefensible position, but I would dearly love to see a brilliant defense of the current system, if somebody can indeed provide one. Otherwise, you have no choice but to admit that my proposal is better than what is available now.

One thing today's economic environment is bringing out is how fragile the economic status of many families is - all it takes is one illness or one lay off for a family to move from middle class, where college is affordable, to a class where college is not affordable. How do you quantify that in advance? Just not admit kids whose family incomes are below average, since that is where the risk is greatest?

Like I said, the optimal solution would be for schools to pledge to provide full financial support for every student that needs or may need that support until graduation.

But given that many schools will not do that, and a statistical analysis does indeed indicate that certain students are not going to be able to afford to pay their entire way through, then why not reject them? After all, if a family is under financial stress, or predictably will be in the near future, then what good does it do to take that family's money and not even grant the student a degree? You're then making a bad situation even worse.

One possible proposal is for schools to simply not charge those students who don't graduate. This might be implemented through a refund system, or by simply not charging until graduation (so if the student doesn't make it to graduation, he pays nothing). Of course, some adjustment would need to be made to allow students to be able to transfer their credits to another school should they choose to do that, but nevertheless, the general principle stands.

However, I'm not going to hold my breath for schools to do that. Given that schools are going to charge students whether they graduate or not, then the humane thing to do is simply not admit those students who aren't going to be able to pay their whole way through. After all, if somebody is poor, you shouldn't take their money, hence rendering them even more poor, without even granting them something valuable in return.

[sakky]

If only those students could have been identified in advance! Every school would like this crystal ball!

Actually, I think MSmom&dad nailed it - many schools, especially lower-ranked ones, don't really care because they want to bring in as many students as they can, including many who won't graduate, as that means more revenue for them.

^^ Which ones practice this revolving door? Can you post the list?

An obvious one: University of Phoenix. Heck, many (perhaps most) of the for-profit universities would fall under this category. After all, by definition, these schools are out to make a profit.

[newmassdad]

Sakky,

Many colleges have an activity called "institutional research" that does the data mining you suggest. They do use their historical data in ways like you suggest. And those schools that are freer of political influence (none, even Harvard, is totally free. State U, the biggest "offenders" here are often totally bound by political considerations. ) do a remarkably good job of using such data to form a class.

However, few Unis focus totally on graduation rates as the predicted variable. To do so misses many of the objectives of the institution.

This is hardly the place to debate statistical methodology or admissions decision making, and I have a long weekend road trip starting shortly. I will bow out of this discussion.

[sakky]

Many colleges have an activity called "institutional research" that does the data mining you suggest. They do use their historical data in ways like you suggest. And those schools that are freer of political influence (none, even Harvard, is totally free. State U, the biggest "offenders" here are often totally bound by political considerations. ) do a remarkably good job of using such data to form a class.

However, few Unis focus totally on graduation rates as the predicted variable. To do so misses many of the objectives of the institution.

And that's the real question - what exactly should be the objective of the institution? I would argue that the point should be to actually graduate students. You don't just bring in students for the sake of bringing in students. The point is to actually finish the degree.

After all, you said it yourself, some students are not able to complete their studies due to financial difficulties. My first response to that is why can't schools provide better financial support to those students? My second response would then be to not admit those students (or at least, not charge them). After all, if a student is suffering from financial problems, then charging them without even granting them a degree only makes them worse off. You took whatever little money they actually had - does that really make them better off? The point is, or ought to be, to improve the station of these students, but that goal is clearly not accomplished by sticking them with a bill sans degree.

This is hardly the place to debate statistical methodology or admissions decision making, and I have a long weekend road trip starting shortly. I will bow out of this discussion.

Uh, if you wanted to bow out of the discussion, why did you even respond at all?