GPA: Can you be too perfect?

twofish-quant

Because the B student spent time outside the classroom writing poetry.

I'll let others talk about graduate admissions committees, but I do know first hand that employers are a little worried about people with GPA's that are too high, because it suggests that they might focus too much on classes and not on things that aren't graded.

One other difference is that most managers are people that don't have perfect GPA's so that having perfect GPA's is not something that gets you much respect in industry.

bcrowell

I was a grad student at Yale, which I think has average grades about the same as Harvard. I taught undergraduate labs (to premeds), graded the work, and recommended final grades. Although the grades I recommended were very high, the thing was that most of these students really were very good students. In most cases, there was essentially nothing being taught in the course that they didn't master, and they basically never did anything major that was incorrect in their written work. The Ivy League isn't like it was back when Bush went to Yale. Admissions standards are extremely high.

lisab

I'm not so sure there's a real difference between a 4.0 student and a 3.9 student.

twofish-quant

It would be interesting to do a statistical study of people's undergraduate grades and outcomes after getting their Ph.D., but personally, I can tell you that I've had a much easier time in industry because of things that lowered my GPA in college.

I'd expect zero correlation, but it wouldn't surprise me if the correlation was negative.

twofish-quant

I've had experience at both Harvard and UTexas Austin. The thing about UTexas Austin (at least when I was a TA there, I hope it has changed) was that you had very good students also, but there weren't enough upper class places for all of the people taking physics, so the lower division courses graded extremely harshly in order to weed out students, and the attrition rate was extremely high. Part of the way you did this was to set up the tests so that silly minor mistakes could kill you on the tests, and there were students that got "weeded out" at UTexas physics that I thought would have done just fine at MIT or Harvard.

One thing that the professors at MIT did which I thought was a great thing but which would get you screamed at in some places is that they generally put problems on the final exam that were not covered in class. The philosophy was that "life gives you problems that we didn't cover in lecture, and so will we." That sort of thinking (which I think is great) would get you in trouble at UTAustin and at least with the courses that I took at Harvard as well.

This insures that no one got anywhere near 100% on the tests, but then the final grades were scaled so that you ended up with reasonable GPA's. Also the way that tests at MIT were graded was pretty good. You got lots of points off if you missed the concept, but you got few points off if you "got it" but just did something stupid. The consequence of this is that you cannot machine grade tests, because you need someone that is pretty skilled to figure out what the student was doing. Hand grading is extremely time consuming, but people at MIT thought this was vital enough so that you had to hand grade the tests. Once you start machine grading tests, then what happens it that you end up playing a game of "gotcha."

Curiously machine grading is something that the University of Phoenix does not do, for the same reasons.

Something that I find interesting is that how you grade is part of the "hidden curriculum". There is a very deep and (I think wonderful) philosophical message in how MIT grades that makes it different from how Harvard grades. One thing that makes it really interesting is that it's "tacit knowledge." People are used to a given grading system and they assume that how the world works and they don't think very deeply into how that system works, and the "deep philosophy" that is embedded in the system.

twofish-quant

Also part of the reason I happen to believe that "grades aren't that important" is that I got that philosophy from some of my teachers both in high school and at MIT. The problem with that philosophy is that it got me into the meat grinder. Grading policy at MIT is something that people have screamed about since 1861, and the people that I was strongly influenced by were what I called "new school" people in the Office of Undergraduate Education (the names there are Benke, Paul Gray, and Margaret Macvicar) that had a very different philosophy than the "old school" people that ran the departments.

I was very strongly influenced by "new school" people but "old school" run the graduate admissions so the fact that I had relatively low GPA meant that I wasn't able to get into graduate schools that I wanted to. On the other hand because of that background, I think I've done better after I got my Ph.D. so if I had to talk to a younger me (which is what I'm doing now), then I'd give "new school" advice.

HeLiXe

Thanks for starting this thread MIH :)

R.P.F.

Hmmm...I'd say 14 hours a day during school years. It is a lot but that's how I function.

twofish-quant

One other thing that causes problems is *defining* achievement. In a lot of places GPA *defines* achievement, so there is a 100% correlation between GPA and achievement.

The problem is that if you define achievement another way, then the correlation is different. Even the act of defining achievement in a way that is mathematically quantifiable restricts you.

For example, one thing that is important for my definition of achievement is "not being a jerk." How do you put a number to that?

twofish-quant

One other thing. I said that Harvard inflates grades. I didn't say that it was a bad thing. Personally, I think it's good that Harvard does that because it helps makes grades a bogus measurement.

Something that I would like to do one day is to teach a class, and tell everyone on the first day that they all get A's. They can leave the class, do nothing, and they will get an A+. Heck if they want, I'll give them an A++++++

Of course, everyone knows that they will all get A's which makes that A totally meaningless. Anyone that stays around and tries to earn an A that they get automatically are the people that I want to teach.

clope023

The A student could just be a good test taker...

Also your question seems silly itself btw, I am more articulate and write better than most of my fellow students in my engineering classes but quite a few score higher than me in exams.

MathematicalPhysicist

It's a slippery slope you slip, I can argue that your'e not so sure there's a big diff between 3.9 student and a 3.8 student... and we all know how this all ends in.

Leveret

Or went out binge-drinking. Or was in bed with mono. Or simply got unlucky with exam schedules, and had to take three big ones on the same day. One B can make the difference between a 3.9 and a 4.0, and is well within the bounds of the "**** just happens" factor. By extension, the suggestion that the 4.0 student is more likely to have better/worse communication skills than the 3.9 student is only slightly less absurd than the notion of 4.0 students being better/worse accordion players.

Ryker

See, now I think you're just being contrarian. Sure, maybe some B student did spend time outside the classroom writing poetry, but that doesn't imply that B students are better prepared for life (or whatever standard you're trying to measure them up to) than A students. It's great that it worked out so well for you, but now you're making it seems like by definition A students are worse than those with B's.You haven't answered his question, though.

And really, some of you are now making it seem as if it's the admission committee's task to try and come up with as many "excuses" for those A students performing well as they can, and then when they do, experience that "gotcha!" moment and adamantly refuse to let a "good test taker" into their school. This is getting ridiculous.

What boggles my mind most, though, is the fact that all of you are or striving to be scientists. If these inferences and conclusions are based on logic employed in science, then slap me silly and call me Sandy.

turbo

I happened to have mono and bronchichitis back-to-back and missed more that a months worth of class-work, but did my best to catch up and ended up with better than a B. Not bad for a challenging engineering school with a 5-year pulp and paper scholarship in the works.

flyingpig

Let's see I got 100% in Calc II from last term even though I've made mistakes on the exams, but my professor handed out 100% to a few others who score well on it so that worries me because I feel that "100%s" tels them that it was a breeze course. So i think a 99% would've probably looked nicer

cjl

But the percentage doesn't even show up on the transcript, does it? I haven't checked mine recently, but I could've sworn it only had letters, not percents...