Low curves and understanding of the material

[Arsenic&Lace]

I was very shocked by the class averages and curve for my quantum mechanics course. The exam was curved to (brace yourselves) 55/100 as an A. Yeah. Kind of shocking, frankly. Of course, the class average was 47.5 or something on that order. My own score was an 80, which I'm not terribly happy with, but apparently that's an A++++ now.

This is at Arizona State University. How would this compare with curves other people have experienced? I would assume the curve at Stanford or MIT would be much less generous...

That's just the first part though. I'm curious, how much comprehension of the material would you say you had as a student? There are all sorts of results in my quantum course I probably couldn't derive if you asked me in the street, although there are certainly a few which I could (e.g., what are the eigenfunctions of an infinite square well, derivative relations for delta function potentials etc). How many of the details would you say you had mastered as students? With curves this low, can't students slide past without really understanding anything and still get good grades?

 

[ModusPwnd]

Thats the problem with having non-standardized exams. The teacher just makes up some arbitrary level of passing and you have no idea how that is correlated to your peers across the country that you are in competition with.

Yes, students can slide past without really understanding.

 

[TMFKAN64]

The worst curve I ever saw was in an undergrad EE semiconductor course I took. It was the only time I had ever seen a test with an exponential distribution.

The average was 12. With 36, I had a rather high A.

The thing that still stuck with me is that there were quite a few zeros and ones on that test, and I always wondered what the difference between them was.

 

[Arsenic&Lace]

The exam wasn't difficult. The first "problem", 25% of the grade was (brace yourselves again) merely to write down the time dependent Schrodinger equation. Most people got that.

The next was to solve the square well for a function with a constant ground state. Trivial, methinks.

The 3rd was somewhat weird, it was difficult to acertain what he wanted. Upon figuring out what it was after the fact though, the question was also very easy.

The last question was "hard"; proving that the eigenvalues were real and the eigenfunctions orthogonal. Only "hard" if you don't understand hermitian operators, really, but a lot of people screwed this one up.

 

[Travis_King]

Thats the problem with having non-standardized exams. The teacher just makes up some arbitrary level of passing and you have no idea how that is correlated to your peers across the country that you are in competition with.

Yes, students can slide past without really understanding.

Well, first off, standardization of exams would be ridiculous. I wouldn't have wanted my fluid dynamics exams from my top-ranked alma mater to be standardized to the level of a course at community college. Many people choose their schools based on the curriculum and differences in subject matter and method of delivery necessarily result in non-standardization.

Second, curves are not always arbitrary. Sure, sometimes they are, and that's kind of wrong, but many times there is a good reason for curving an exam. It is difficult to make an exam for a large college class with people of all levels of aptitude. So sometimes tests are created so that a student with an acceptable understanding of the subject matter will get a B or C, students with a pretty good understanding will get an A, and the people who really understand the stuff will shine (not grade-wise, but to the professor).

Sometimes it is a method for determining who really are the best or most dedicated students in the class without being unfair to the students who are putting in average effort.

 

[Arsenic&Lace]

Why would exams at top ranked alma mater university be any more difficult than those at U of A?

 

[Travis_King]

It was more of a jab at UofA, given that you said you're at ASU, but my point was that standardization for schools with vastly different curriculum and delivery methodologies would be ridiculous.

I've changed it to "community college", as I figured I didn't want to insult any engineers from UofA. Not to mention the fact that it is a pretty decent school anyway and the point wouldn't be as poignent.

 

[Arsenic&Lace]

The most difficult maths professor I've ever had taught a community college...

I mean, there was no curve, and out of 30 students only 3 passed the class, myself included.

(of course, a considerable number dropped before the final so who knows how many would have passed before then)

 

[Travis_King]

The point isn't the level of knowledge or difficulty of the professor, it is the level of knowledge, effort, dedication, etc of the students on a general basis.

One (ideally) chooses a specific program because he/she believes that the students, the professor, and the curriculum are best suited for what they want out of their education. You want students who are at or above your level so that you are challenged, you want a professor who will teach but not spoon feed and give you exams that will challenge you to learn the material.

There's no "right" way of teaching any subject. Each school, each professor, has their own ideas of what is most important.

If tests were standardized, the only criteria for choosing a program would be "How well does this professor prepare me for the exams?" That is obviously a ridiculous way to approach higher education.

Not only that, but the curve on such a standardized test would destroy all but the most intelligent or dedicated of the class year.

 

[ModusPwnd]

If tests were standardized, the only criteria for choosing a program would be "How well does this professor prepare me for the exams?" That is obviously a ridiculous way to approach higher education.

I disagree completely. Its far from obvious, its highly debatable.

 

[Travis_King]

I do not see how standardization of testing would in any way benefit higher education. I'd really like to see an argument for standardized testing in universities, because I can honestly not see any true benefit.

In the classic sciences it is ridiculous because research direction and methodology have little to do with testing.
In engineering it's redundant because you have to pass a standardized test to enter into the professional world anyway.

University should not be pushed more toward grading and curves than it already is.

 

[ModusPwnd]

So you disagree with nearly every University's requirement of a physics GRE for grad school admission? They need that standardized test because grades are completely subjective to the whims of the individual instructor and ultimately only as comparable to the extent that you believe physics teachers are uniform...

Standardized testing benefits all education. Its the only way to compare apples to apples. Its the only way to see how you measure up against the competition, which is necessary because its very competitive. If you don't standardize the test then your grades are meaningless indicators of performance. Imagine if all data gathering were done that way... Word the questionare differently for everybody, but then compare their answers as if they were the same... lol, yea right. That is no way to gather data.

I would also say that it is a conflict of interest to have the same person who is doing the educating also judging the efficacy of that education. There is no way around that - its not right and its not fair.


I worked at a university and community colleges that had standardized testing. One was gen chem classes that had to pass a type of qualifying exam that was authored by the american chemical society. This was great for the students. They had clear goals and expectations and they got useful data regarding their performance. Another was a county wide calculus qualifying exam for the community colleges. This was also a great thing. Before the variation in the teachers methods and grading was just too wild and not comparable. Some would rake the kids over the coals, but most would be far to easy and the students would think they were fine for moving on to university but they were not.

The arguments I see against standardized testing boil down to protecting the teacher's conflict of interest and protesting against a particular standardized test (rather than the idea of them).

 

[Arsenic&Lace]

No, I concur, standardized testing is an absurd notion for universities, mainly because it achieves nothing particularly useful.

Mostly I was just questioning the implicit assertion that difficulty of exams depends upon the ranking of an institution, since I don't think anybody has a definitive answer to this.

 

[ModusPwnd]

The difficulty of exams depends on the teacher who authors the exam. They are usually given free reign and do what they think is acceptable.

I don't think that implicit assertion holds and I am not even sure it is implicitly asserted...

 

[Travis_King]

So you disagree with nearly every University's requirement of a physics GRE for grad school admission? They need that standardized test because grades are completely subjective to the whims of the individual instructor and ultimately only as comparable to the extent that you believe physics teachers are uniform...

A standardized test for determining one's level of education and retention is a good tool for grad school entry when people are coming from different schools. There is a difference, however, between benchmarking between milestones (i.e. undergrad -> grad), and standardizing course exams and grading curves.

School A can have a overal program curve of say, 65=B, while School B has a curve of 75=B and yet both schools can have students with the same grade distribution on a standard test like the GRE.

ONE standardized test is enough. If the kid can do well on the GRE, then who cares what grades he and his peers got in his university classes?

You are putting too much emphasis on grades.

If a college wants to see how Bill's 3.5 at School A matches up with Jack's 3.9 from school B, then they look at your GRE scores.
If a company wants to do the same, they interview you.

It's not in the best interest of universities to churn out crappy students.

Standardized testing benefits all education. Its the only way to compare apples to apples. Its the only way to see how you measure up against the competition, which is necessary because its very competitive. If you don't standardize the test then your grades are meaningless indicators of performance.

No. The grades indicate how well you are performing as per the professor's judgement. There is an implicit agreement that the school is trusting the professor to ensure he passes only quality students. Indeed, the reputation of the school rides on this notion.

Again, school isn't about grades, it's about content.

Imagine if all data gathering were done that way... Word the questionare differently for everybody, but then compare their answers as if they were the same... lol, yea right. That is no way to gather data.

That's why there are standardized exams for milestones (i.e GRE, FE, etc). And also why institutions and companies don't accept/hire based on grades alone.

I would also say that it is a conflict of interest to have the same person who is doing the educating also judging the efficacy of that education. There is no way around that - its not right and its not fair.

What? Professors aren't judged based on passing rates, they don't get bonuses if all their students get A's. Heck, many schools tout their dropout rates precisely because they want to be perceived as difficult and competitive. Most professors don't just pass people for the heck of it. If you aren't learning the curriculum, you are wasting their time, and you are potentially taking the seat of another student who would be more dedicated to learning what the professor is teaching.
I don't see that there is anyone more appropriate for grading in a university setting. (despite the fact that, in actuality, it's generally the TA who grades everything)

 

[Arsenic&Lace]

All this hoo haw about standardized testing aside, how much of the knowledge presented in your undergraduate courses would you say you ordinarily obtained and preserved over the semester? Various details escape me until I need them, usually.

 

[chill_factor]

someday you will be the guy with the below average grade.

will you beg the professor: "I think my grade is too high because you curved it. I demand a lowered grade."

 

[Lavabug]

I was very shocked by the class averages and curve for my quantum mechanics course. The exam was curved to (brace yourselves) 55/100 as an A. Yeah. Kind of shocking, frankly. Of course, the class average was 47.5 or something on that order. My own score was an 80, which I'm not terribly happy with, but apparently that's an A++++ now.

This is at Arizona State University. How would this compare with curves other people have experienced? I would assume the curve at Stanford or MIT would be much less generous...

That's just the first part though. I'm curious, how much comprehension of the material would you say you had as a student? There are all sorts of results in my quantum course I probably couldn't derive if you asked me in the street, although there are certainly a few which I could (e.g., what are the eigenfunctions of an infinite square well, derivative relations for delta function potentials etc). How many of the details would you say you had mastered as students? With curves this low, can't students slide past without really understanding anything and still get good grades?

The exam wasn't difficult. The first "problem", 25% of the grade was (brace yourselves again) merely to write down the time dependent Schrodinger equation. Most people got that.

The next was to solve the square well for a function with a constant ground state. Trivial, methinks.

The 3rd was somewhat weird, it was difficult to acertain what he wanted. Upon figuring out what it was after the fact though, the question was also very easy.

The last question was "hard"; proving that the eigenvalues were real and the eigenfunctions orthogonal. Only "hard" if you don't understand hermitian operators, really, but a lot of people screwed this one up.

Wow, seriously? This is considered a "hard exam"(judging by the class average)? It does seem worrying. It's very frustrating to see some people walk away with good grades from well known undergrad institutions but with little understanding or preparation. These are the people getting preferred entry into graduate programs, while people from more rigorous but lesser known schools end up with little options for continuing their studies, despite having better preparation for grad school (afaik, from looking at grad homework and exams on the internet(and using them for help!)).

I actually feel better about my chances at admission in grad school now, provided the grad committees (ASU being one of them) take a good look at my undergrad curriculum and not put so much weight on the PGRE. My university does no curving. Final exams are typically 4-5 hours and almost always account for 100% of the grade. The problems are always original and quite detailed. For reference, first round of QM1 exams = 100% failure rate. 2nd and 3rd round 40% and 20% of the people retaking the exam failed , respectively. I took the 2nd exam. It involved 2 original theory questions (20% of the grade) and 2 problems that accounted for the bulk of the grade: two particles with spin=3/2 in a magnetic field at a given angle, where the initial state was given as a tensor product of J eigenstates in the x-basis. Lots of things to calculate in different scenarios: time evolution, state collapse, expectation values of observables, predicted time for finding a given probability of finding a given eigenvalue of Jy... Basis change galore. That was the easiest problem.

The other was a 3-level potential well problem where you were given the functions outside and inside the well for each of the levels, "encoded" in the form of eigenvalues of some other observable. The problem involved about 4-5 pages of integrals (we were given a table of useful integrals though), but nobody finished it.

I've had Lagrangian dynamics problems on exams that make some of the problems in Landau's mechanics look easy. 1 rigid body dynamics problem (spinning cone in a helicopter like motion attached with a spring, total 4 degrees of freedom) that weighed 80%, and 2 short proofs.

Of course I would say this, but I don't think the PGRE is a good indicator of having had a good education. The 2 most recent sample exams I've done are more of a formula memorization contest than a test of physical concepts. I am likely having a crack at it in a week but I'm not hopeful I'll do well, among personal problems I've got a lot of other things on my hand like an ambitious research project and a bunch of courses to worry about. I would hope that grad committees put more weight on the content (course outlines, textbooks) of one's undergrad education than performance on a PGRE.

 

[Arsenic&Lace]

I've already had below average grades by my standards, and I've felt several things:

1. Sometimes relief if it's a course I can barely keep up with.
2. Sometimes irritation on the grounds that I feel like I was lazy, and the whole point of grades is to encourage me not to be.

Would I "beg the professor to enforce a stricter curve"? No, on the grounds that this whole bizarre charrade of exams and points and grades and numbers is utterly heinous, and if somebody makes my journey through it easier without good reason, who am I to care; there's hardly a good reason to have the system as it is in the first place.

Also I am intrigued by Lavabug's post, it sounds extreme frankly. What's the purpose of asking students to produce answers to extremely difficult questions in very short periods of time? What's difference does it make for their education if you gave them 2, 3, 4 fold as much time? Now I would presume that you'd be a better exam taker than me after that torturous experience... but will you be a better candidate for becoming a scientist? I don't have an answer to that question, actually.

 

[Lavabug]

I've already had below average grades by my standards, and I've felt several things:

1. Sometimes relief if it's a course I can barely keep up with.
2. Sometimes irritation on the grounds that I feel like I was lazy, and the whole point of grades is to encourage me not to be.

Would I "beg the professor to enforce a stricter curve"? No, on the grounds that this whole bizarre charrade of exams and points and grades and numbers is utterly heinous, and if somebody makes my journey through it easier without good reason, who am I to care; there's hardly a good reason to have the system as it is in the first place.

Also I am intrigued by Lavabug's post, it sounds extreme frankly. What's the purpose of asking students to produce answers to extremely difficult questions in very short periods of time? What's difference does it make for their education if you gave them 2, 3, 4 fold as much time? Now I would presume that you'd be a better exam taker than me after that torturous experience... but will you be a better candidate for becoming a scientist? I don't have an answer to that question, actually.

I wouldn't consider myself a "good exam taker" at all, the scores I've been getting on PGRE practice exams show that(plus I've rarely had more than 3-4 detailed problems on an exam, nothing like the PGRE). I'm just someone who studied as hard as he could to prepare for largely unpredictable exam questions, my preparation did not consist in memorizing formulae which is what the most recent pgre's seem to boil down to.

I don't think I'm a better candidate for becoming a scientist than the next guy just because I studied a ton of theory ... I don't have actual research experience. I do have tons of experience looking through dozens of obscure textbooks trying to find examples and hints, since often I found exams significantly harder than problems given in class or on sheets. The process of trying to understand papers and consulting advanced texts for my senior project (actually a MSc thesis) feels a lot like how I prepared for exams, but it's a "skill" that is probably useless if I'm not able to continue in academia, which makes me slightly sad.

 

[ZombieFeynman]

Wow, seriously? This is considered a "hard exam"(judging by the class average)? It does seem worrying. It's very frustrating to see some people walk away with good grades from well known undergrad institutions but with little understanding or preparation. These are the people getting preferred entry into graduate programs, while people from more rigorous but lesser known schools end up with little options for continuing their studies, despite having better preparation for grad school (afaik, from looking at grad homework and exams on the internet(and using them for help!)).

I actually feel better about my chances at admission in grad school now, provided the grad committees (ASU being one of them) take a good look at my undergrad curriculum and not put so much weight on the PGRE. My university does no curving. Final exams are typically 4-5 hours and almost always account for 100% of the grade. The problems are always original and quite detailed. For reference, first round of QM1 exams = 100% failure rate. 2nd and 3rd round 40% and 20% of the people retaking the exam failed , respectively. I took the 2nd exam. It involved 2 original theory questions (20% of the grade) and 2 problems that accounted for the bulk of the grade: two particles with spin=3/2 in a magnetic field at a given angle, where the initial state was given as a tensor product of J eigenstates in the x-basis. Lots of things to calculate in different scenarios: time evolution, state collapse, expectation values of observables, predicted time for finding a given probability of finding a given eigenvalue of Jy... Basis change galore. That was the easiest problem.

The other was a 3-level potential well problem where you were given the functions outside and inside the well for each of the levels, "encoded" in the form of eigenvalues of some other observable. The problem involved about 4-5 pages of integrals (we were given a table of useful integrals though), but nobody finished it.

I've had Lagrangian dynamics problems on exams that make some of the problems in Landau's mechanics look easy. 1 rigid body dynamics problem (spinning cone in a helicopter like motion attached with a spring, total 4 degrees of freedom) that weighed 80%, and 2 short proofs.

Of course I would say this, but I don't think the PGRE is a good indicator of having had a good education. The 2 most recent sample exams I've done are more of a formula memorization contest than a test of physical concepts. I am likely having a crack at it in a week but I'm not hopeful I'll do well, among personal problems I've got a lot of other things on my hand like an ambitious research project and a bunch of courses to worry about. I would hope that grad committees put more weight on the content (course outlines, textbooks) of one's undergrad education than performance on a PGRE.

I have a tough time believing that you could survive that sort course and and not perform well on the gre Perhaps it is a bit of evidence that super difficult exams do make students any better than those that take easier exams. What university do you attend?

Grad schools do not look at your course outlines from what i know.

For what its worth the exams at my undergrad university were easier than lavabugs but harder than arsenic and laces. I survived grad classes just fine.

 

[Lavabug]

Like I said, my undergrad exams were nothing like the pgre, I don't think it's a case of "train 1km to run 100m". Sure I can respond to most of the QM questions that don't involve knowing the Bohr radius expression by heart, but overall it hasn't prepared me for memorizing tons of formulae. I had a really tough geometric optics course where we were given all the formulae on the exam and it still wasn't of much help, so I can and have done fairly challenging optics, but I cannot remember the general lens formula (the one with the inner and outer radii) off the top of my head, which is what a few pgre problems consisted in.

I've seen some grad schools request you provide a list of the textbooks used in your courses.

I attend a little-known (outside of the EU) university in Spain and it isn't one of the toughest institutions in the country, there are others with even tougher curricula and harsher grading. Also until last year (before the Bologne reform that shortened bachelors from 5 years to 4), all undergraduates were required to take Jackson's electrodynamics(on top of a year of sophomore EM), atomic physics (Brandsen) on top of a year's worth of Cohen-Tannoudji's QM, among other requirements that are generally always taught at a high level(2-3 digital and analog electronics courses, for example).