Blizzard employees work at Princetown University

[Ashkael]

As I was reading this piece of news, I was oddly reminded of the way Blizzard likes to fix some loopholes and problems in their games.

In short: starting this Fall, Princetown University will put a cap on the number of A's that can be attained per course. The cap will be at 35% for undergraduate courses, and at 55% for junior and senior independent work. At least that's what the news piece says. They don't mention how they are actually going to implement that, but the obvious guess would be that only the top students in the A bracket would get the A's, leaving some students on the lower end of the A bracket with B's. The reason behind this, according to Princetown, is that they currently average 44% to 55% A's per class and they want to go back to the early 90's average. Obviously, a University with the high standards of Princetown doesn't want people saying that all their students are getting easy A's, so the quick, band-aid solution is to put a "cap". <_<

The way I see this, this cap can only help make the competition among top-tier students even more fierce than it already is. Survival of the fittest. I'm a top-tier student, and I'm always engaged in a neverending battle to get the highest grade of my class. Putting a cap on the number of A's that can be awarded per class is not going to change the way I study; I'll just be more stressed than I already am. But what about the B student who occassionally gets to enjoy a hard-earned A? They get ripped-off.

Regarding professors, I don't know what's going on at Princetown but professors are either making their classes really easy or they are just too good at teaching that students get good grades. My experience as a student has shown me that no matter how good the professor, there's always going to be a good portion of the class that will fail miserably, so I would lean more towards the first option. In any case, the problem here are the professors and/or their teaching standards. Princetown is basically going for a big band-aid, slap-in-the-face solution that completely bypasses professors and places the burden of the problem on students. They are the ones that have to deal with it.

I wonder if Princetown actually took the time to conduct an extensive study of how their own teaching faculty is working.

[jahcs]

That strategy sounds flawed to me. If you do extrordinary work (read: near perfect) your grade should reflect that. If many people in the class do outstanding work then maybe it was an outstanding class. If they feel that more A's means the course is too easy then increase the difficulty of the material taught or lower the level of the course, as in changing a math 300 to a math 290. (I don't know all the nuance of the course numbering system so that may not work.)

Students should recieve the grade they earn, not the grade they are given.

[--Pete]

Hi,

Students should recieve the grade they earn, not the grade they are given.

And just what grade do students earn? How do you determine it?

For simplicity, let us suppose that the grade is based on one test consisting of 100 true-false questions. Further, let us assume that the questions are all based on statements from the required textbook, there is no room for ambiguity, and the book's "opinion" is used to determine "truth". Also, let us assume that the class consists of a large enough number to be statistically significant (say 100 or more) and that the people in the class all have met some minimum prerequisite.

Even with all those assumptions (many of which are unrealistic or idealized), a test could be designed so that the statistics are high or low and tight or loose. So, how is the person grading class to use those numbers? On a simplistic <60 = F, 60 to 70 = D, etc. scale, the whole class could fail (low and tight or very low), the whole class could get A's (high and tight) or there could be any number of distributions.

So, now, look at the case of the same course taught by two (or more) teachers. Assume all the same assumptions as above, but that each teacher makes up his own test. Some are "easy" the others "hard". Now, what is fair? To penalize those that got the hard tests?

You might say "Make all the tests the same difficulty", but if you do, then I'll say "You've obviously never made up a test". The only true way to know the difficulty of a test is to use the same questions over multiple classes and keep statistics on them. That is the ideal way of doing it, but fails miserably in an environment where old tests are available to the students. So, that technique works well in, for instance, the military where there are usually no overlapping classes. It fails in the typical college, university, or even high school. Thus, a teacher is in the position of using judgment, experience and gut feel to make up a "fair" test. And they don't always succeed.

You might suggest a "standard" test be given to all sections of that course. Frankly, that idea is often used and it sucks. First, there is the problem that some teachers for reasons good or bad will "teach to the test". But even if all the teachers had the integrity to put education before grades, there is the problem that in almost any subject there are a number of equally valid topics which can be stressed, skimmed, or skipped. Each teacher has his or her own opinions on what is "important". Thus, even if the teachers were completely ignorant of what the test covered, there is still the "luck of the draw". Any given section of the class might have emphasized more or skipped more of what was on the test to the student's advantage or detriment. This is even more likely since the test *is* created by a human, probably a teacher, who also has his or her opinions as to what should have been covered and how much it should have been emphasized.

So, for the above reasons and many others, it becomes necessary to grade on a curve. Now, to many (especially those who have had to pass through the propagation of ignorance that is the typically "Masters of Education" program), the "curve" means the "Bell Shaped Curve", which most of them use, knowing neither its correct name (it's a "Gaussian" or "normal distribution", folks) nor its properties. But for many cases, the "curve" is adequate.

Aside: Rant on "grading on the curve". A straight line is, mathematically, a curve (one with an infinite radius of curvature). Thus, any method used to convert one measurement (grade on a test, perception of "how hard he's trying", "I don't like his attitude", etc.) to another measurement (F through A, possibly decorated with pluses, minuses, or whatever nonsense the "educators" come up with next) is "grading on the curve". Our society's attitude towards brutal frankness doesn't allow anything else. The teacher who would have the guts to say, "Mrs. Jones, I flunked your Billy because he's a spoiled brat that you've failed to teach discipline." would be my hero and the administrations scapegoat.

Back to the topic.

Grading on the curve implies deciding (in advance or after the fact) how many of each grade will be issued. One commonly used method is that a C is one deviation from the mean (~68%), two deviations are a D or B (~18% each) and three are your A's and F's (~2.5% each). Many of the concepts based on grades are founded on these values. So, the C which is now looked upon as almost a failing grade is still defined in most grade manuals as "average". And the A is usually called "exceptional". And we now have the ridiculous situation of the bulk of the population being "above average". You know, if every *applicant* to every national team for the Olympics was given a gold medal upon application, the gold medal would be worth less than an honest dog's droppings. And that's the way of grades.

And the prof teaching a class of ten? They might each and every one of them think they are brilliant, and he's thinking "How did these morons get here?" Should they get the F he thinks they're worth or the A they feel they "deserved"?

So, pissed at the grades your getting? Profs are unfair? Then get your degree, get behind that podium and do better.

--Pete

[Chaerophon]

three are your A's and F's (~2.5% each)

Just a minor quibble, but most curved schemes that I have seen employed in large classrooms don't go all of the way to a full three standard deviations but assign A's to those in the 90th percentile and above. I could be way off; I don't do the marking. ;) You'd probably know better than me, but such an exacting standard would seem to me (and I do acknowledge your broader point) to be a little bit too harsh. Particularly in the more competitive faculties, I would like to think that more than three students in a class of one hundred could produce work worthy of an 'A'.

[--Pete]

Hi,

First, looking over what I said, I expressed it wrong. That should read:

Up to 1 sigma is a C.
From 1 to 2 sigma is a D or B.
Over 2 sigma is an F or A.

That, I think, is in line with the numbers I gave.

Personally, I think that grades should be given in terms of how an individual is doing in the class and how the class compares with previous classes. And I think it is unfair of any one person to try to reverse grad inflation all by himself -- all that does is punish students for their misfortune of drawing a fool for a prof.

Other than that, I don't see any problem with using a "standard" curve.

I would like to think that more than three students in a class of one hundred could produce work worthy of an 'A'.

That all depends on what an "A" means. If it means a little better than average, then you are right. If it means "one in a million", then you're probably wrong (although any given class might produces such wonders, it's unlikely). That's the whole problem with grades and grade inflation. There is no standard to objectively compare grades. Not to mention there's no "difficulty rating" to balance the whole mess out. In college, a four oh in a three hour class gives the same contribution to one's GPA whether it was obtained in Music Appreciation 101 or Genetic Engineering 499.

Frankly, the system sucks, but there's no drive, no funding, and not even any good ideas of how to replace it. Instead there's a bunch of half-assed programs (No Child Left Behind being a great example) that substitutes platitudes for performance and memorization for education.

OK, gotta get my BP back in the three digit range -- enough of that rant ;)

--Pete

[Selby]

This is what a certain professor I had to take classes did. 1 A, 2 B's, everyone else gets a C regardless of grades. So my 54 out of 60 points overall for the class is identical on the transcript to someone who only got 23 out of 60 points. Regardless of how well I understood the material or the concepts. I have no idea why he did this other than it was easy for him to grade. I prefer how most professors did with the 100-88% for an A, 87-75% range for a B, and 74-60% range for a C (having a professor who didn't round up and getting an 87.6% for the class and ending up with a B+ made me laugh more than be upset like my other "set # of grades" professor). What I REALLY dislike is a professor that gives you his grading scheme and then finds out he made it too easy and then changes everyone's grades to match a normal distribution AFTER all the tests and grades are in (I would have had a solid A by his original standards and the standard that I was working towards and ended up with a B- at the end since too many A's were being given out to non-PhD students). Princeton just sounds like they are trying to do what other professors are already doing. I see nothing wrong with it, but I do wonder why they have to set it in stone. Just tell the professors what they are expected to do.

As far as getting my PhD and going back and doing a better job, if I cared enough I might. As it is I just complain semi-seriously about how artbitrary grades are. I realized within my first year that college is just another game like high school and grade school. Just play the game and jump through the hoops while learning your material and you'll be fine. Companies who hire solely based on GPA always make me wonder why after having seen how arbitrarily they are assigned. Ah well, I'm out of college and school for the first time in 21 years in 2.5 weeks, so I'll just chalk it up as a learning experience and be glad it's behind me.

[Bob]

Personally, I think that grades should be given in terms of how an individual is doing in the class and how the class compares with previous classes. And I think it is unfair of any one person to try to reverse grad inflation all by himself -- all that does is punish students for their misfortune of drawing a fool for a prof.

so, by that logic:
I should get some concessionary marks on my A-level maths?

My 'lead' maths teacher, and head of maths in the school left at easter... and my other maths teacher has been on long term sickness leave since 2 months before that...
That's right, In the 6 weeks leading up to my final exams, I'm being taught by NEITHER of my maths teachers.
The school has shuffled timetables to put a member of senior staff who used to be head of maths to teach us for 1/2 our lessons (pure maths, and to be fair, I think she's doing better than my original teacher), but the other 1/2 (mechanics) are being covered by a supply teacher who hasn't done mechanics before (she's a statstician). She has to try and teach us after teaching herself from the same textbook that we do. This textbook has poor explainations, weak examples, no structure, and ~ 25% of the answers in the back are wrong. We're presently in the situaution of the blind leading the blind.

Anyway, think that'll earn me some free marks?

Frankly, the system sucks, but there's no drive, no funding, and not even any good ideas of how to replace it. Instead there's a bunch of half-assed programs (No Child Left Behind being a great example) that substitutes platitudes for performance and memorization for education.

A similar situation is developing in Britain.
Mr Bliar would like 75% (or some other impossible number) of people to go to university. So exams have to get easier to get more people to pass them, which means that those of use who could pass them without them being easier still get qualifications that are now discredited by industry.

And, we have thse fantastic AS-Levels. Which means that we now get to do important external exams for 3 years in a row, and at least 2 months of the lower-sixth is spent revising for exams instead of just doing the course. So you have less time to do the subject properly, and now get to 'learn to the exam'... yay! If you're like me, and actually want to know WHY something works (and find it hard to remember if you don't) you're at a serious disadvantage, since there now isn't really time to go over it.

But, this has another downside. With the massive influx of people applying for courses, there's more competition than ever for places.
Oh, and of course they can't afford to teach all these new students. So Mr. Bliar is introducing top-up fees to help pay for it... one of his manifesto pledges was that he WOULDN'T do that.

Ok, that's enough of my rants about the UK education system... I ought to work and actually try to pass these exams.

-Bob :P

edit: typos

[Selby]

Well Bob, it's good to see the Brits are taking a page from America's messed up system and trying to implement it themselves. Wonder if they'll make the same mistakes America did like implementing such a messed up system ;-)

I remember being in "English" at 14 and having to learn how to take the TAAS test (state test that you MUST pass if you want to go on to the next grade). The test was so easy it hurt, yet we had to sit there for 3 months having to learn how to take the test so that the school would get it's funding for next year (more money if more students in the school pass). When we asked why we had to do this since it was obvious, our teacher told us that it was important to the school and we needed to take it seriously too. Talk about misguided...

[Griselda]

I despise grading students, although I enjoy other aspects of my job. Grade inflation is a problem, and it makes sense for universities to address it at the university level rather than at the individual level. That particular system sounds too arbitrary for my taste, but that could be due to the nature of the class I am teaching (small numbers of students, upper division required course, no tests or quantitative data collected during the course). I have found that there are large variations in the students' performance from term to term. Also, these variations tend to be multiplicative- when the class happens to have several strong students it seems like all of the students do better as a result. We have richer discussions and can explore some topics in a little more depth. On the other hand, if the overall ability of the students is lower, there is nobody left to drive those in-depth discussions (I talk to myself often enough but we can only get so far that way), and the entire group ends up skimming the surface rather than doing much meaningful learning. Or, worse yet, we spend a lot of time covering material that they should have learned in prerequisite courses. Some terms, I honestly believe that everyone deserves an A. Some terms, almost nobody does. I'm sure that Princeton would have similar variations in their smaller courses.

The other thing is that students pretty much universally feel that they deserve an A, regardless of their effort or abilities. They respond to that belief in several different ways-

1. Some apply themselves and work hard to make sure they get the A that they are seeking.
2. Some complain and quibble any time they are not given full credit for any assignment.
3. Some let you know the exact ways in which their life will be ruined if they do not get an "A".

When students use #2 or #3, they are often prepared to fight for the "A" after the fact. They will either contact the professor's superior to complain or lodge a formal complaint with the university. At that point, the teacher needs to be prepared to jusitify the decision to give the student a lower grade. Even if the teacher is justified, this process can be mentally draining and time consuming.

So, in many cases, teachers may feel that it is easier to give an A than to face the consequences of not doing so. As long as that's the case, grade inflation will probably continue to be a problem.

It's not inevitable, though. I got my BA at a college that has little or no grade inflation. I believe that, even today, "C" is the average grade. However, this can make it difficult for their students to get into graduate schools, because the students' GPA will be lower than those of students from other schools. They send a note along with every transcript explaining their "no grade inflation" policy, but I'm not sure that's enough to cut through the red tape in many graduate admissions departments.

What's the difference between that college and all of the other ones experiencing grade inflation? Well, grades were de-emphasized. Students got verbal and written feedback most of the time. We were graded, but did not even see our grades unless we went to the registrar personally and asked to see them. That's kind of a goofy system, and it does have some flaws, but it was effective in creating a school culture where grades were not often discussed, in class or even just between students.

The one drawback is that it's very expensive to have small classes so that it's realistic for a professor to give extensive verbal and written feedback to the students. So, I'm not sure it could work on a larger scale. But, I would prefer to use that model than a more traditional one.

-Griselda

[kandrathe]

Bleh, I agree. Focusing on grades and grading is a horrible way to run education. It would almost be fairer to do everything on a pass/fail basis. You either get it enough to move on, or not.

I really think that people should get a fair grade, and not limited by some quota. If a particular class has an unusually large number of bright people in it, then it might be the case that 40 or 50 percent of them get higher than average. But, statistically over time it should be a wash. What infuriates me is that peoples dreams are at stake, and universities are so damn cavalier about messing them up. It's not like attending is unneccesary, easy or cheap.

For example, at the university my wife attended, after her sophmore year she was to transfer to the school of her major having completed the prerequisites. The year she was ready to transfer to the school of Business (which she must do to be able to register for any further classes in her major) due to a freak bulge in the number of applicants, the GPA requirement was 3.85 in the core of 12 prerequisite courses which included 3 calculus courses, a stats course, 3 accounting courses, 3 computer, micro/macro econ. She had a 3.83. Two of her accounting courses were the kind where the prof only awarded 1 A, and a handful of B's in a class of 100 students. Those B's meant she could not pursue her major, and instead of getting a degree in business, she chose the shortest path out of the University. She was accepted into the school of Mathematics and got a degree in Statistics. She went on to get 2 Masters degree's at a different, more highly accredited school.

Either extreme in grading mucks up the works. I hope the engineer who designed the safety devices on my car went to one of those harder, intense schools. :)

[Griselda]

Actually, the course I teach was pass/fail only at some point before I taught it. What happened was that most students did just enough to pass, but no more. :(

-Griselda

[jahcs]

kandrathe Posted on Apr 29 2004, 01:21 AM

Two of her accounting courses were the kind where the prof only awarded 1 A, and a handful of B's in a class of 100 students

This is what I was getting at in my earlier statement that students should recieve the grade they earn. I was in a course a few years ago where 6 students were within 3 points of each other at the top of the class. Another 20 or so students were doing average work. The rest of the class, about 15 students, were doing just enough to get by. The students struggling with (and that cared about) the material were in the second category because they applied themselves, asked for help, turned in all assignments, and attended class, even on non-test days. Those top students earned about 97% of the points available that quarter. This instructor did not believe in extra credit. Two of the six recieved A's because of his grading scale.

I was one of those two so please don't think I was writting this because I felt spurned by the instructor. Looking at the overall rank chart for the class it was fairly easy to see what the grades could be. Top six: A's, 25 B's/mostly C's, and about 10 D's. Instead it was 2 A'S, 4 B's, about 30 C's, and 4 D's. No one failed but a couple of Incompletes were given out.

[kandrathe]

What happened was that most students did just enough to pass, but no more.

I can see that happening.

So, then the question is should everyone be expected to exceed the bar, or far exceed the bar. You being a teacher I can empathize with you wanting your students to get as much as they can out of the curriculum. I guess you could always set the pass bar higher to the "B" level. I guess from a purely mechanical POV, if one has grasped the fundamental concepts needed to progress to the next course work, then it is really their loss or gain. At some level they might find that their lack of enthusiasm in learning more of the fundamentals has limited their ability to progress further. Luckily when I was younger I had a Socratic mentor who inflamed my passion for learning -- anything.

I think many times in life we are too busy to "exceed the bar" in much other than a few selected passions. Sometimes in the world of work,speed of delivery of the report/project on time is as important as the quality of the work. Over engineering, or spending extra time in what I do is viewed as a waste. It might be that at times "merely adequate" is what is called for. In my line of work I am constantly fighting "creeping elegance" or the "wouldn't it be better if..." After many thousands of dollars have been spent for little gain, the answer becomes obviously "no".

Putting it back into a University context; Is it beneficial or realistic to expect a person to be "A" level dedicated, have energy for and be enthusiastic about 3-5 courses simultaneously for 4-5 grueling years? When I see someone who graduated University with a 4.0 I think they are either A] workaholic or B] fricking brilliant. Most often they are A].

[Ashkael]

I do believe in GPA. A high GPA is more often than not a good indicator of how much that student has busted his ass to learn the material. Right now I'm on my junior year of electrical engineering, and I'm still amazed at the number of people who can't do common factorization or simple derivatives. I'm still amazed at the amount of people who, at this level, don't know how to solve systems of equations with two unknowns and have to ask the professor to dedicate a whole Circuits I lecture to that topic.

Just today I had my Circuits I final. After two really easy partial examinations, the professor decided to be a "jerk" and surprise us all with a really hard final exam. I'm confident I did pretty good on it, and I'm also confident in the fact that all the people who studied under my tutelage for that final are going to fail miserably, because they studied only enough to pass. The professor knows that even though the first two tests were easy, those who actually learned the subject are going to be able to solve the final exam, as hard as it was. That's the way a professor should solve the problem of A's without resorting to silly "caps".

[Chaerophon]

Unless they are an absolute genius, I don't see any way to achieve a perfect 4.0 in a stringent, competitive humanities field. There is just too damn much difference in the subjective marking criteria of different profs. What is an 'A' for one will be an 'A-' for another, and yet, the essay that the second prof would have assigned an 'A' would be assigned a 'B+' in the first course. I know that Canadian GPA's receiving full-ride philosophy scholarships to graduate school at Oxford / London School of Economics sit around 3.9, or even slightly below that level. I tend to believe that I am capable of producing 'A' level work in any course that I take. I'm also a workoholic. However, there's really nothing that I can do to avoid the occasional 'A-'. Every professor has a different idea about what constitutes a flat A, and it's hard to convince them otherwise once they've made up their mind. All that it takes is a slight misinterpretation on their part or even a difference of opinion, and an A can easily be downgraded. Accounting for such occasional subjective discrepancies has been, in my experience, all but impossible. Those who are capable of graduating with a perfect 4.0 in philosophy/political science/history, etc. must have, to my mind, some serious talent or insight, on top of an ample work ethic. I've yet to come across anyone that is managing to do it at my school. I know of a couple of people who are carrying 3.9's, but none with 4.0's. Of course, I can imagine it being a little bit easier in a less subjective field....

[Selby]

As someone who just completed his Master's in Electrical Engineering I say "welcome to the engineering world" ;-)

We had problems with students not knowing how to do basic math as well. Our department was structured to give out 3 stumbling blocks to students with those problems. The Linear Systems\Electronics I semester would weed out the really stupid\slacker people, the Communications\Electronics II would get rid of the ones that slipped by the cracks the first time and the final one of Electromagnetic Field Theory II would separate the men from the boys when it came to the EE students. The rest of the classes were fine, but those particular ones were some of my hardest and most challenging times in engineering school. Graduate school was more fun though, probably because we had a network of the same 5-6 students in my research area that all took the same classes and all worked on the material together so that we would cover all of the bases and do well in the classes.

And we did have 1 4.0 student in EE. He was one smart guy too. Not a workaholic, just very intelligent. And all of his circuits and systems would WORK the first time when he built them. He couldn't do PSpice to save his life though ;-)

[kandrathe]

Unless they are an absolute genius, I don't see any way to achieve a perfect 4.0 in a stringent, competitive humanities field.

True enough. Although, I saw alot of grade focused people choosing courses based on the difficulty of getting an "A" and how the professor graded.

I was being biased to the non-subjective world of the so-called "hard" sciences and mathematics where essays on the uses of Eigen values are not required. Although, after awhile it became very easy for me to get good grades in almost any course. Partly because I was mature enough to set aside my ego and recognize the game of stroking the professors and just giving them the answer they wanted to hear. Through practice I think a persons short term memory improves such that it is easier to regurgitate all the facts and theories and their implications -- according to what the professor thinks those implications are. At least for undergraduates, I think the Bachelors degree from what I saw was an exercise in teaching humility and conformity. I had the advantage of being slightly older when I started, with 5 years of professional experience before I started at the University. I was used to a daily hard work ethic and being told what to do by then. I had a 4.0 in all my computer science courses, but not quite that in my other course work. I think if I recall my GPA was like 3.6. I started working professionally on writing software when I was 16, and I've never stopped. So, my schooling was a neccesary check box for me to remain competitive in my field but was never a factor in any job application. Also, I guess since I was paying all my own bills, I wanted to get the most out of every dollar I spent.

[whyBish]

In college, a four oh in a three hour class gives the same contribution to one's GPA whether it was obtained in Music Appreciation 101 or Genetic Engineering 499.

Well, over here (at my Uni at least) they have a system where there is a guideline depending on which year the study is being done at.
For 1st year it is about 20% A grades, for 3rd year I believe it is 50% A grades.

This guideline is not the target number of such grades but is the trigger point for an external audit on the grading for each course. So if you do a third year math paper that 5 of 9 students get an A grade for, then that will be reviewed to see if that particular paper had a 'good' set of students, in which case those A grades would stand. This is all done prior to marks being sent out to students.

We also have a couple of not so good practices.

1) Exam marks are not sent to students (You never have any idea how you score on an exam unless you pay a fee to see your paper)
2) Some lecturers still fit to a gaussian, but mainly it is for standard papers such as "Intro Calculus" etc. where the standards are known and the papers are low level.

Also furtunate, is that for the four year engineering course I was doing, honours was awarded for particular GPA results. These GPAs were weighted for which year which as I remember was:
(Year, Weighting):
1,0
2,20%
3,30%
4,50%

Very fair (even though I would have done better with a flat weighting ;) )