« Caught my eye | Main | Taxation »

Should high schools compute class rank?

...something is missing from many applications: a class ranking, once a major component in admissions decisions...In the cat-and-mouse maneuvering over admission to prestigious colleges and universities, thousands of high schools have simply stopped providing that information, concluding it could harm the chances of their very good, but not best, students.

One college administrator notes:

"The less information a school gives you, the more whimsical our decisions will be," he said. "And I don't know why a school would do that."

Here is the full article.  As implied, colleges will still form rough expectations of what your rank would be.  Given this response, we can see at least one reason why parents, and thus high schools, might prefer a fuzzy or ambiguous class ranking to a definite one.

Let us say your kid is smart but has a small chance of making it into a top school.  At Yana's high school (Woodson, in Fairfax) I've seen folders of students with 4.0 and 1600 SAT scores who did not get into Harvard or Yale.  Getting into those places has elements of a crapshoot.  You are gambling, with the odds against you, and a payoff varying only at some threshold level of success (i.e., getting in is what matters; if your kid doesn't get in, it doesn't matter how close he came.)  Those are the classical conditions where the gambler prefers to take more risk.  On the upside, your chance of getting in goes up and on the downside, the longer left-hand tail doesn't hurt you.

Consider an analogy and assume I am trying to date Salma Hayek.  Should I tell her what car I own (GeoPrizm, basically the same as a Toyota Corolla), or should I be vague?  Now Salma is no dummy.  If I am vague, she will not infer that I own a Rolls-Royce.  But a GeoPrizm is clearly below her cut-off point, so with vagueness there is at least some chance she will not nix me right off the bat.

When I was a kid, a great resume meant you could go to Brown (and you only needed a Cadillac to date Raquel Welch).  There was less reason to be ambiguous about class rank, as vagaries could hurt your chances in a very real way.  You started off with something to lose, and being fifth in your class put you in pretty good stead.

Natasha tells me that many good law schools no longer provide class rank for their students.  Is the same mechanism at work here, with many people chasing after a few hard-to-get plum jobs?  Can you think of other examples where the principle of deliberately ambiguous rankings applies?  Can this rationalize grade inflation?

Posted by Tyler Cowen on March 7, 2006 at 06:36 AM in Education | Permalink

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c66b253ef00e550835bb38833

Listed below are links to weblogs that reference Should high schools compute class rank?:

» More on Schools Avoid Class Ranking, Vexing Collegs from School Information System
Tyler Cowen comments on this recent article:Let us say your kid is smart but has a small chance of making it into a top school. At Yana's high school (Woodson, in Fairfax) I've seen folders of students with 4.0 and... [Read More]

Tracked on Mar 7, 2006 10:27:55 AM

Comments

Is this perhaps a result of grade inflation? I think it used to be much rarer for a student to have a 4.0 and a 1600.

It's a curious kind of inflation, in which there is an absolute upper limit--kind of like price controls that set an absolute ceiling.

Posted by: Sandy Smith at Mar 7, 2006 8:20:55 AM

If kids with 4.0 GPA's and 1600 SAT are not getting into Harvard, then perhaps we should blame the re-scaling of the SAT that ETS did about 10 years ago. ETS has reasons of its own to be vague (to prevent the SAT from seeming too important or too racially imbalanced), and its vagueness increases the high school's opportunity for vagueness.

I would think that this would create a market opportunity for someone to design a new, harder test, capable of separating the Harvard-worthy from the people hoping to fudge their class rank. Or, perhaps the real opportunity is for a smaller college to play "Moneyball" and concentrate on taking the supersmart but non-athletic Asian and Jewish kids whom Ivy League admissions policies are designed to keep out.

Posted by: DK at Mar 7, 2006 8:23:54 AM

Rolls Royce?

Surely Hayek has better taste than that. Say, a Porsche or a Bugatti?

Posted by: Dave at Mar 7, 2006 8:27:25 AM

Hey, you wouldn't expect a guy with a GeoPrizm to know the difference, would you?

Posted by: Tyler Cowen at Mar 7, 2006 9:01:50 AM

It indeed is likely that a similar mechanism is why some law schools no longer issue class rankings. Recent years have seen the development of a caste system in the increasingly glutted legal profession. A comparatively small percentage of law school graduates get high-paying, prestigious jobs with top law firms, while most graduates have to scuffle around in the hopes of making a lower-middle-class living. Because law schools love to brag about the high salaries their graduates earn, they're willing to do whatever is necessary to boost graduate earnings, which may include being vague about class rankings for the reasons you have outlined.

Posted by: Peter at Mar 7, 2006 9:21:25 AM

Class ranking consists of an averaging of scores. If all the students take exactly the same courses, then perhaps some ranking can be created that has merit. But I'm not persuaded that a usable one can be.

First, the students are not taking exactly the same courses. In other words, a weighting scheme is required. So suppose there are four levels of Math available, does an "A" in the least difficult get downweighted to a "C+"? In practice, there is no way to generate appropriate weights.

Which leads to the problem of differential weighting within a particular level, but across classes. Shouldn't one also weight by instructor? Heck, even the time of day that the class was given (I've never been a morning person, so should I get greater credit for an 8 am class than I would if I were taking a 3 pm one in the same topic?).

In sum, there would need to be as many rankings as there are students, I would guess. So assign each of them a random ranking between 3 and 12 for reporting to the colleges.

(Note: when I was in high school we were told that our grades would be weighted. Naturally, we only learned that they were not when we went to apply to colleges .... So I have a rather dim view of the entire exercise.)

Posted by: Paul McMahon at Mar 7, 2006 9:31:18 AM

FYI, the SATs are now out of 2400, so a 1600 SAT is actually not so hot. They changed it two years ago, possibly just to make us old fogeys look dumb when we cite our out-of-1600 scores.

Posted by: Grant Gould at Mar 7, 2006 9:36:31 AM

If "many good law schools" don't rank any more, then what will the second-tier schools do?

I can imagine them ceasing to rank in hopes that this will make them look more prestigious.

OTOH, if one's stuck with a degree from (say) Ole Miss, then it's best to be near the top of one's class at Ole Miss.

Posted by: Anderson at Mar 7, 2006 9:41:40 AM

This is all part and parcel of the ridiculous tendency to devalue what objective universal measures there do exist, namely the SAT and achievement tests. Throw in grade compression in both high schools and colleges and you really hurt many of the best students.

Without class rank, getting straight A's becomes even more important, which penalizes kids who are really bright but who bomb a tough class or two in a very good school.

At Caltech and MIT, I think that for internationals (who are judged by a much higher standard than domestic applicants) success in International Math or Physics Olympiads is almost necessary since these are so tough and discriminating.

Of course, all of this is made worse by the fact that HYP reject 4.0 1600 students for some 3.5 1400 kid with "good ECs" and "leadership potential." Leaving aside whether this is even desirable, I have seen no study that corroborates their claim that they can in fact SPOT leadership or creativity in hs students. And let's not get into the disease of sports, legacies and AA.

Someday a super elite school will form which will only judge on academic, objective criteria but until then, we have the revenue max weirdness of Ivy League college admissions which are nothing like GRAD school admissions to those same schools. [Try telling Princeton Econ that you should get into their Phd program because you did community service and acted in a film even though your GRE Math is 650.]

Posted by: jn at Mar 7, 2006 10:01:42 AM

Elite schools seek future leaders, who will make a lot of money and donate some of it back to the school.

Nerdy smart people wind up with jobs that plateau around $100K and will never be big donors.

Posted by: Half Sigma at Mar 7, 2006 10:33:46 AM

There is a paper by W. Chan, H. Li, W. Suen that explains grade inflation in colleges along the lines of your post. Link here: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=336880
It is optimal for colleges to inflate the grades in order to introduce noise in the signal about the quality of their students when employers cannot tell whether a school has good students or gives easy grades. The top students are hurt, but many more mediocre ones are helped. And once a school inflates the grades, others have incentives to do so.

I do not recall exactly, but I believe their argument depends on employers behaving competitively. I am not sure the same argument would hold at the high school level with extremely picky colleges. To build on your example, just as Salma Hayek could demand any dating applicant to disclose the quality of their car without being worried she might end up without a date, Harvard could demand high school rankings and reject any applicant coming from high schools that do not provide them. I suspect high schools do not provide them precisely because Harvard does not want them.

Posted by: Doru Cojoc at Mar 7, 2006 10:55:54 AM

The North Carolina School of Science and Math has never given class rank, because they feel it would be misleading since it's a statewide public residential magnet school. The worst students at NCSSM were among the best at their home school. I wouldn't be totally surprised if Thomas Jefferson in Virginia didn't for the same reason.

Posted by: John Thacker at Mar 7, 2006 11:00:34 AM

Who's going to rank the high schools?

Posted by: Keith at Mar 7, 2006 11:02:57 AM

I went to NCSSM like John Thacker, and a class ranking there would definitely be counterproductive to what they were trying to do -- besides, we were there for only 2 years. I don't remember if they compute a GPA now, but when I was there (1990-1992) NCSSM would release transcripts, but gave no official GPA. The colleges were free to calculate as they wish, of course, but it's a pretty good bet that getting an A in Calculus at NCSSM was different from getting an A in Calculus at one's "home school".

In any case, most of us ended up at UNC, NC State, or Duke, which knew all about the pecularities of our high school.

Posted by: meep at Mar 7, 2006 11:14:48 AM

I went to a similar program (to NCSSM) in Texas. We didn't give ranks, but did GPA's. Besides discipline, our Deans were responsible for flying out to alternate coasts to explain the program to adcoms. So, it was a two-part strategy: lessen the GPA impact on the middle-class, and work the committees for the super-achievers. Very responsible, now that I think about it.

I think a lot of people agree that the SAT is not very representative. And the addition of another 800 points of writing didn't do much to help. From what I recall, I had plenty of time to check my work, so it might be acceptable to simply add more questions. On the other hand, I wouldn't want them to make it like the GRE, where speed trumps all reason.

Also, undergrad admissions at the elite is as much about passion and drive as it is about "EC's", these days. Those are the things that will get you through college.

Posted by: badidea at Mar 7, 2006 11:44:58 AM

Let's not forget that admissions for Ivy League schools have never been based on strictly quantitative criteria as Mr. Gladwell reminded us in his piece on the subject for The New Yorker ( http://www.gladwell.com/2005/2005_10_10_a_admissions.html ).
Also, I don't know many lawyers, regardless of the tier from which they came, that "...scuffle around in the hopes of making a lower-middle-class living." Unless it's by choice, of course.

Posted by: Josh at Mar 7, 2006 12:08:17 PM

That student with 1600 SATs and a 4.0 GPA who doesn't get into Harvard or Yale can count himself lucky. I feel sorry for those deluded students who think that they're going to get the best possible education at an Ivy League school. Far better to go to a smaller, less prestigious college where you're actually going to meet the professor. I once had the opportunity to read the freshman English reading list at Brown--practically all feminists and queer studies, with nothing written before about 1970. Compare this to my the reading list at my own alma mater (University of the South in Sewanee, TN), where we read a Shakespeare play every two weeks, with a generous serving of 15-19th century English poetry. Which school do you think better serves its students?

Posted by: Thelonious_Nick at Mar 7, 2006 12:14:01 PM

When my son's high school quit providing class rank a few years ago, their explanation was similar to that attributed to NCSSM by John Thacker. They said that as a (somewhat) selective private high school, they felt that their students were being penalized because class rank is determined relative to the strength of the entire class.

Interestingly, they seemed most concerned with less-selective colleges that try to raise their US News ranking by admitting only students in the top X% of the class than with reducing the Ivies' ability to rank students.

One additional thought: Which highly-qualified students gain admission to very selective schools appears quite random to those outside the process. Perhaps making it more truly random is in the high schools' interest. Each student can attend only one college. Providing a class rank allows Harvard, Yale and Princeton to all admit the top students in the class, while without it each college might pick different students to admit, leading to more students being admitted from the HS overall.

Jim

Posted by: scarhill at Mar 7, 2006 12:58:29 PM

"Far better to go to a smaller, less prestigious college"

The less prestigious college does crap for you when you are trying to get a high paying job.

Posted by: Half Sigma at Mar 7, 2006 2:14:47 PM

Here is an empirical paper by Don Moore & Samuel Swift arguing that grade inflation at a school helps students in that school get admitted at the next level, since admissions committees do not adjust sufficiently for the easier grading scale: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=728627 . (Their article is on the transition from college to graduate school, but I imagine that the effect extends to other transitions.)

Withholding class rank information could have an analogous effect, or it could interact with the grade inflation effect by making admissions committees adjust even less for the easy grading that allowed the students to get such high GPAs.

Posted by: Blar at Mar 7, 2006 2:27:05 PM

"This is all part and parcel of the ridiculous tendency to devalue what objective universal measures there do exist, namely the SAT and achievement tests."

Oh, JN ... I am an achievement-test expert. I scored 34 on the ACT back when 35 was the top score, 1510 on the SAT, 2240 on the GRE, 169 on the LSAT, and except for some night-before on the LSAT, I never cracked a test-prep book, let alone a Kaplan course. Not perfect scores, but enough to pay my tuition wherever I went, and to win fellowships from grad schools.

But I'm also lazy and procrastinating and too inclined to think I can do everything at the last minute, traits which impaired my GPA, as they should have.

Achievement tests are "objective," perhaps, but they leave out a lot.

Posted by: Anderson at Mar 7, 2006 2:30:46 PM

Just thought I'd mention another example of information degradation I have written about: Deleting adverse information from credit reports because of regulatory restrictions (actual and proposed). In as much as this is happening, the likely results are less confidence and less accuracy in credit/employment/insurance/housing decisions, which means good consumers get opportunity less often than they would have, and bad consumers more often. Upshot: less opportunity, and on less advantageous terms. Like in grade inflation etc., the results are not merely distributional. The pie shrinks.

Posted by: Dan Klein at Mar 7, 2006 3:03:59 PM

Don't forget Montgomery Blair in Maryland among the magnets.
IIRC they sent a letter with your college applications describing the program, but this was 10 years ago, I don't know what they do now. My sub 3.0 would have kept me out of an Ivy, but Carnegie Mellon didn't seem to mind. Of course Carnegie Mellon also didn't inflate grades like the Ivys so maybe they had some sympathy.

Posted by: BillWallace at Mar 7, 2006 3:13:55 PM

Half sigma,

You seem confused about who becomes rich in America. Look at the Forbes 400: there plenty people lot of people at the top like Larry Ellison who were "nerds" that would not have excited an Ivy admissions officer. The county where I live has four publicly disclosed billionaires, two of whom are engineering professors, one who started as a contractor for Hollywood, and one who worked his way through Chapman college . All four have given heavily to higher education.

Posted by: tylerh at Mar 7, 2006 3:46:02 PM

At my Top Law School That Does Not Rank (Penn), there was never any indication that it was to "fool" employers. Almost everyone got a good enough job.

What chosing not to rank did, though, was to do away with the need to compete. We were all students learning together, not competing for a slot. Everyone was happy to share notes or outlines. Lots of inclusive study groups. It just made Law School a much better experience.

People who want to rank want school to be a competition rather than a learning experience. But it doesn't have to be.

Posted by: Richard Bellamy at Mar 7, 2006 3:59:37 PM

Post a comment