Grade Inflation in the US

[Yakk]

Actually they probably can, if there's not too much overlap in the material covered in successive grades.

Even if different material is covered, it often requires the understanding of previous material, or at least the same skill set that was required to understand previous material.

For example, my experience is that people who struggle to understand stoichiometry will go on to struggle to understand chemical equilibria, because they both require the same skill (the ability to apply math to a real-world situation).

Or, in an English class, if you get a poor grade on the first essay, it might be because you just didn't grok that first book, but it's more likely that it's because you weren't a great writer, in which case you still won't be a great writer on the second essay.

Even if the variables are correlated. If you add enough together, you will get something that's approximately normal.

Proof or it didn't happen. (or are you talking "observationally"?)

...

Note that belling the grade curve in a particular class being justified by the CLT requires that each individual student grade be an average of a collection of independent random variables. It has less to do with the "collection of 100 students in the class" in some sense (the difference in the class distribution from the "ideal normal curve" has to do with the number of students in the class, but the fact that the ideal curve is the normal one is what I'm talking about).

[Dason]

CLT may well apply, but it rather requires large numbers doesn't it? Even in a class of 100 students, that's not 'that' large, and many classes are smaller ..snip..)

CLT is pretty good for reasonably behaved random variables if your sample size is around 30 or above. But really we would be talking about the sampling distribution of the mean so I don't really see how we got CLT involved.

Even if the variables are correlated. If you add enough together, you will get something that's approximately normal.

Unless the correlation is too strong.

[achan1058]

Proof or it didn't happen. (or are you talking "observationally"?)

I think I can prove it under certain (limited) assumptions. But otherwise, it's mainly due to observations. It obviously does not work for certain cases. I am all against forcing grades to become a bell curve, but for the most part it normally happens.

Even if the variables are correlated. If you add enough together, you will get something that's approximately normal.

Unless the correlation is too strong.

You also need the fact that the original variables aren't Gaussian (or close to) already.

[Dason]

Even if the variables are correlated. If you add enough together, you will get something that's approximately normal.

Unless the correlation is too strong.

You also need the fact that the original variables aren't Gaussian (or close to) already.

I don't quite get what you're getting at here. CLT works fine for Gaussian random variables. It even has an amazing convergence rate!

[achan1058]

Even if the variables are correlated. If you add enough together, you will get something that's approximately normal.

Unless the correlation is too strong.

You also need the fact that the original variables aren't Gaussian (or close to) already.

I don't quite get what you're getting at here. CLT works fine for Gaussian random variables. It even has an amazing convergence rate!

The comment was on "unless the correlation is too strong", as in, if your variables are already Gaussian, it doesn't matter what the correlation is.

[Dason]

Well if the correlation is -1 between successive pairs then there might not be convergence at all even if they're all Gaussian.

[achan1058]

That's a good point. My bad.

[katethegreat]

I just finished writing a research paper on grade inflation last week, oddly enough. Because I'm too lazy to rewrite my whole paper, here are some points I want to touch on:
1. The bell curve as a grade distribution, with C as the average, is fundamentally flawed. When it was made, it took into account only difficult graders. While it has some merit, it makes it seem like C ought to be the average, but it should really be somewhere in the Bs. I don't think averages should be as high as they are now, but if teachers treat the bell curve as science and think that "This many students have to fail or get Cs, because the curve says so!", they aren't right either. Also, logically, if a teacher teaches well, few to no students should fail.
(See Harry R. Lewis's "Excellence without a Soul")

2. Grade inflation makes students today have inflated egos and overestimate their skills. In the 1970s, about 50% of students felt they were "above average or in the top 10%", today it's around 70%. This starts early on, particularly in high school. In the 70s, about 30% had high school grades in the As and Bs, and now over 80% do.

3. Since good grades can take little effort, students study less and don't improve their skills (almost half didn't improve skills like critical reading or writing their first 2 years of college).

4. Tenure is a major factor behind grade inflation- if professors give out high grades, students will rate them as better teachers, helping them get tenure.

5. Since grades don't mean much now, employers and grad school rely less on grades (since everyone has a 3.something, they can't tell the difference) and more on standardized testing (also, the kids with lower GPAs because of stricter professors are screwed).

6. Students are more likely to avoid challenging courses and math and science courses to get better grades.

7. Students complain to professors more, and they'll give in and give them the higher grade. After all, higher education is a business, and you need to keep customers happy. Administration actually encourages this, since they want to keep graduation and retention rates high.

I think that's mostly everything. If you need citations, just ask.

[GenericAnimeBoy]

Near the end of last semester, my engineering stats professor showed the class test grade curves from the past 3 or 4 years of his class, including ours. In that class, it was a two-bump distribution with a fairly narrow spike around 90 and a wider lump around 72. He says that's been pretty consistently true as long as he's been calculating it.

[ImTestingSleeping]

Also, logically, if a teacher teaches well, few to no students should fail.

I thought your entire post put grade inflation and its consequences into perspective, but I don't agree with what one sentence.

How do we define teaching well?

Say we have a calculus 1 course in an average undergraduate college and the professor is finding that the kids are struggling with their algebra so much that she/he can't even get to the calculus concepts. Does that professor teach down or continue as the course would be taught if the class was full of kids who were fundamentally sound in the prerequisites of the class (high school algebra, trig, geometry, etc). What is teaching well in that situation? I would say that the former would have the lesser failing rate, but the latter is the better teacher.

I don't think we can give even the best of teachers students who have been undereducated for years and years and expect them to catch them all up.

One reason why it isn't fair to expect that from teachers is because a lot of these students wouldn't be used to the amount of work required to catch up at that point. A recent study at my college asked students "How many hours on average do you work for ALL of your courses outside of class?" This was asked of all majors and the answer they got was less than 3 hours of work per week. I hope that isn't representative of all liberal arts colleges, but all I know was that I was embarrassed to hear that. Consider that earlier I said the average GPA at my school was around a 3.3 and it was done so with an average of 3 hours of work per week. This is a small college whose reputation is growing every year and accepts the top 20% from high school classes.

[katethegreat]

Also, logically, if a teacher teaches well, few to no students should fail.

I thought your entire post put grade inflation and its consequences into perspective, but I don't agree with what one sentence. 

How do we define teaching well?

Say we have a calculus 1 course in an average undergraduate college and the professor is finding that the kids are struggling with their algebra so much that she/he can't even get to the calculus concepts. Does that professor teach down or continue as the course would be taught if the class was full of kids who were fundamentally sound in the prerequisites of the class (high school algebra, trig, geometry, etc).  What is teaching well in that situation? I would say that the former would have the lesser failing rate, but the latter is the better teacher.

I don't think we can give even the best of teachers students who have been undereducated for years and years and expect them to catch them all up.

One reason why it isn't fair to expect that from teachers is because a lot of these students wouldn't be used to the amount of work required to catch up at that point. A recent study at my college asked students "How many hours on average do you work for ALL of your courses outside of class?" This was asked of all majors and the answer they got was less than 3 hours of work per week. I hope that isn't representative of all liberal arts colleges, but all I know was that I was embarrassed to hear that. Consider that earlier I said the average GPA at my school was around a 3.3 and it was done so with an average of 3 hours of work per week. This is a small college whose reputation is growing every year and accepts the top 20% from high school classes.

You're right; I was looking at the situation without considering other factors.  If a class is adequately prepared and puts in an appropriate amount of effort, the majority should pass.  But, as we both know, neither happen.  
I think teachers have to strike a fine balance between helping the students and keeping the standards of their course.  It seems like an utter waste to pretend that students are prepared and teach at them, and have them leave the class without understanding the material and setting them up for failure.  While it is more work for the professor, maybe there needs to be more remedial resources offered, perhaps as an after class session.  The professor or TA could review basic skills while incorporating concepts from the course.

I'm only in high school, but the lack of preparation is already pretty evident.  I don't think we've ever been properly taught how to write, but teachers expect us to write with a sophisticated style and make profound inferences.  Somehow I've managed to pick up some skill over the years, but a lot of my friends struggle.  They want to improve. Teachers make the speeches about how "you will leave this class a better writer", but they never teach us the necessary techniques.

[Yakk]

Really, the lack of preparation is indicating that the failure should have happened earlier.

People will probably learn at different rates. Teaching at a rate so that the slowest person can pick up and master the skill, even presuming adequate preparation, seems somewhat wasteful -- because anyone who is significantly faster than the slowest person is now forced to sit around and burn time and is getting next to no real positive feedback from extra effort.

There are mastery based systems that do something like that -- each student takes as long as it takes for the student to master the problem domain (not 50% accuracy pass, but mastery) then moves onto another problem domain. Student velocity is a function of how fast they master material. "Failure" maps to "it takes longer". I don't know what this kind of system does for drop out rates (as someone who can't/won't master a problem domain gets frustrated and quits).

[ImTestingSleeping]

Really, the lack of preparation is indicating that the failure should have happened earlier.

People will probably learn at different rates. Teaching at a rate so that the slowest person can pick up and master the skill, even presuming adequate preparation, seems somewhat wasteful -- because anyone who is significantly faster than the slowest person is now forced to sit around and burn time and is getting next to no real positive feedback from extra effort.

There are mastery based systems that do something like that -- each student takes as long as it takes for the student to master the problem domain (not 50% accuracy pass, but mastery) then moves onto another problem domain. Student velocity is a function of how fast they master material. "Failure" maps to "it takes longer". I don't know what this kind of system does for drop out rates (as someone who can't/won't master a problem domain gets frustrated and quits).

I see mastery based education as the future, but it needs to be done correctly. I've heard stories where it failed, but in schools that tried it with only a single teacher and some aids helping out. Basically what happened was they used worksheets for each topic, but when one kid finished a worksheet, they had to wait for a teacher to give them the next one, which is not a great system.

What is missing is use of technology. Check out this video from Salman Khan from the now Google funded Khan Academy:

http://www.youtube.com/watch?v=Tksh1zHEkIM (2:41)

If you haven't seen this guy, he basically started out putting little tutor videos on YouTube intended for a nephew of his to see. He got so many positive responses, he started doing more of them and now there's thousands of videos that span all types of disciplines online for free use. He's starting to get a ton of media attention and attention from big names like Bill Gates.

Here's the new site (courtesy of Google's money really ) in case you're interested:

http://www.khanacademy.org/

[GenericAnimeBoy]

How do we define teaching well?

Say we have a calculus 1 course in an average undergraduate college and the professor is finding that the kids are struggling with their algebra so much that she/he can't even get to the calculus concepts. Does that professor teach down or continue as the course would be taught if the class was full of kids who were fundamentally sound in the prerequisites of the class (high school algebra, trig, geometry, etc). What is teaching well in that situation? I would say that the former would have the lesser failing rate, but the latter is the better teacher.

I don't think we can give even the best of teachers students who have been undereducated for years and years and expect them to catch them all up.

If I'm understanding you correctly, you're effectively claiming that if a professor is not failing at least (however many %) of his students he is probably compromising instruction? I'm desperately trying not to set up a strawman here, but are you seriously claiming that "quality = difficulty"? It is the students' responsibility to catch up, but it is the instructor's job to do whatever is necessary to ensure that it is possible for his students to learn, over the course of the class, the skills and knowledge that class is supposed to teach. That includes dropping or referring to tutoring those students who are clearly deficient in the prerequisites, but it also includes adopting teaching techniques which illuminate the information best for their students.

In my experience as a student, the better professor is whichever one goes beyond simply presenting book knowledge by teaching the skills needed to apply that knowledge. Spoiler'd: I elaborate on my experiences in a verbose manner--

Spoiler:

I had a Chem 2 professor last year who gave very detailed theory lectures that comprehensively covered the subject matter, but did not do a very good job of teaching the skill of applying that theoretical knowledge to solve problems. He only ever demonstrated the simplest general case problems, and not very many of those. In short, you could learn just as much by just reading the textbook. When confronted with student complaints about his teaching style, he angrily implied that doing things differently meant "dumbing down" the curriculum.

By contrast, my Chem 1 professor simply covered an overview of the theory (we were expected to learn the details by reading outside of class) and spent the remainder of the class time interacting, having students work problems on the board and if necessary talking them (and the rest of the class, who was following along) through the process of recognizing the applicable theoretical information and applying it correctly to solve the problem. If anything, the Chem 1 class was actually more demanding in terms of both the breadth of knowledge and depth of understanding required to pass than Chem 2 was, but the pass rate and the percentage of students who stayed for the whole class were both a helluva lot higher.

Parting shot: I took Chem 2 at a big research university; I took Chem 1 at a community college with 1/4th the tuition rate. What made the difference for the community college? It was also 1/3 the class size (40 students rather than ~120).

Optimally, nobody would fail classes. Some kids will fail anyway because they aren't very good at dealing with abstraction, and some will fail because of sheer laziness, but to whatever extent we are discarding bright, capable minds we are wasting potential productivity and hurting our society.

[ImTestingSleeping]

If I'm understanding you correctly, you're effectively claiming that if a professor is not failing at least (however many %) of his students he is probably compromising instruction? I'm desperately trying not to set up a strawman here, but are you seriously claiming that "quality = difficulty"? It is the students' responsibility to catch up, but it is the instructor's job to do whatever is necessary to ensure that it is possible for his students to learn, over the course of the class, the skills and knowledge that class is supposed to teach. That includes dropping or referring to tutoring those students who are clearly deficient in the prerequisites, but it also includes adopting teaching techniques which illuminate the information best for their students.

No! I'm not saying that a professor should be predetermining how many students should fail in their class. What I'm saying is that a teacher should NEVER spend a significant amount of class time going over material which should've been learned in a prerequisite of the course. To do so is a disservice to everyone, even to the students who need to go over that material before moving on to the intended material.

For example, I TA for a calculus class at my college. Coming in, over half the class doesn't know their unit circle. They don't know what sin(0) equals. They don't know that a^2-b^2=(a-b)(a+b). These are just a few of the basics high school concepts they do not know coming in. The professor refuses to go over those things during class and I applaud him for it. If he did, the pace of the class would slow and then they wouldn't get to integration by the end. Then they'd get to calc 2 without knowing the basics of integration and they'd be behind again. It is a disservice to EVERYONE to slow down the class for prerequisites. Either work your ass off and catch up or fail.

[GenericAnimeBoy]

No! I'm not saying that a professor should be predetermining how many students should fail in their class. What I'm saying is that a teacher should NEVER spend a significant amount of class time going over material which should've been learned in a prerequisite of the course. To do so is a disservice to everyone, even to the students who need to go over that material before moving on to the intended material.

For example, I TA for a calculus class at my college. Coming in, over half the class doesn't know their unit circle. They don't know what sin(0) equals. They don't know that a^2-b^2=(a-b)(a+b). These are just a few of the basics high school concepts they do not know coming in. The professor refuses to go over those things during class and I applaud him for it. If he did, the pace of the class would slow and then they wouldn't get to integration by the end. Then they'd get to calc 2 without knowing the basics of integration and they'd be behind again. It is a disservice to EVERYONE to slow down the class for prerequisites. Either work your ass off and catch up or fail.

Ah, okay. Yes, I can agree that classes shouldn't slow down for prerequisites, for sure.

[Izawwlgood]

That's well and good in math courses, which generally have a very linear progression, but in other coursework it's not always fair to assume that people who have taken the intro will have had the same experience and material as anyone else.

In terms of failure rates of a class, I think it's better when a professor just sets a standard and keeps it. I've been in courses with only 10% pass rates, and I've been in courses where a 40% average results in a B+, and the lowest grade was a C+.

The advantage of setting a curve is obvious; you actively select for the best students, while weeding out the worst. It isn't a leap to assume that someone who fails Intro Bio isn't going to succeed in Genetics, and if your program prides itself on excellence, only allowing the best of the best to go on is a good way to ensure that only the best do.

[ImTestingSleeping]

That's well and good in math courses, which generally have a very linear progression, but in other coursework it's not always fair to assume that people who have taken the intro will have had the same experience and material as anyone else.

That's a very good point. I honestly couldn't comment on much besides the hard sciences. I have no idea what makes a humanities course good or bad. A little off-topic with this next bit so I'll spoiler it:

Spoiler:

I'm biased as hell, but I just see humanities as fun. I would take a philosophy course for fun. I would take a history or literature course for fun. I think you become a better person after taking a good humanities course and it is in your best interest to do so, but those are the type of courses in which a grading system is a disservice. You get out of them what you put in.

I'm not saying that people should only major in vocational careers and hard sciences, but I think it is much harder to differentiate humanities majors and the traditional system is a mistake for those courses.

In terms of failure rates of a class, I think it's better when a professor just sets a standard and keeps it. I've been in courses with only 10% pass rates, and I've been in courses where a 40% average results in a B+, and the lowest grade was a C+.

The advantage of setting a curve is obvious; you actively select for the best students, while weeding out the worst. It isn't a leap to assume that someone who fails Intro Bio isn't going to succeed in Genetics, and if your program prides itself on excellence, only allowing the best of the best to go on is a good way to ensure that only the best do.

I think setting a standard of where the grades should be at is not a great idea because you punish great classes and help poor classes. What I would hope to see more often is a set standard of material to be gone over each semester. Meet as a department, create curriculum which covers all the things you think a student should learn in their time at university and stick to it. Pick material and a pace that meets the talent of your students. Don't waver from it. Be transparent with what the syllabus of the course will cover. Be upfront about what knowledge is required coming in to the course.

[Bakemaster]

(OOPS: Apologies if any of this post rehashes things already stated; several posts were made between when I started and finished writing it.)

In my opinion, grades should be used primarily to indicate the level to which a student has achieved the goals stated in the course syllabus. This is in contrast to the primary purpose being a comparison of the student's level of success to any other student or group of students, real or theoretical; where top grades imply a measure of exclusivity and/or correspond to a strictly-defined percentile of achievement. My ideal system of passing grades (using the standard A-F) is as follows:

• "A" indicates the student has demonstrated mastery of those tasks explicitly set forth by the instructor and is fully prepared for further study; repeating the course would not further the student's understanding of the subject matter.
• "B" student demonstrates mastery of a majority of the tasks explicitly set forth by the instructor, is fully prepared for further study, and can catch up on anything they missed without additional instruction.
• "C" student demonstrates basic competence and the mastery of enough of the course material that continued success (at a "C" level or higher) is reasonably likely in further study.

It's crucial that a passing grade isn't given to any students who require further instruction in the material of the course in question before they can handle the material of any follow-up courses (for which the course in question is a prerequisite). This is so that the professors teaching the follow-up courses can follow their syllabus reasonably closely, and not have to devote any class time to "catching up" students on material that's not part of the course. "D" and "F" simply provide a way to differentiate between two categories of students who are unprepared for further study. The "D" student comes away with some skills and understanding that could be applied outside of school, but aren't sufficient to allow the student to continue to more advanced material. The "F" student comes away with no significant skills or understanding.

As long as the line that divides passing students from failing students is reasonably successful in separating those students who need to repeat the course from those who don't, I honestly don't think the distribution of the rest of the grades is that big a deal. Not all course material can be cleanly divided up to facilitate the sorting out of As from Bs from Cs; not every course is about teaching skills. Some are about introducing students to a field, or to a way of thinking; some are about giving students an opportunity for personal growth in an area. Not every course requires that all students start at one level and end at another.

GPA as admissions criteria is really only a shortcut for what you want to do, which is get to know a student well enough that you're confident they can succeed in your program. Numbers do very little to introduce individuals, but as long as there are numbers which serve as a decent approximation for the above, schools with limited resources will use them in that way, and most schools have limited resources. It's wonderful to find programs that actually interview their candidates, but that's very rare at the undergraduate level; even graduate programs usually only interview a small percentage of their qualified applicants (when they interview at all). Admissions offices are aware that standards shift, and if GPA becomes less reliable, they move to other criteria; they do what they can to achieve their goal, whether grades are being inflated or not, and their success is solely their own responsibility, not the responsibility of the teachers at preceding levels.

While it is more work for the professor, maybe there needs to be more remedial resources offered, perhaps as an after class session.  The professor or TA could review basic skills while incorporating concepts from the course.

Some universities assign "section leaders" who are responsible for collecting and grading homework from, organizing group study times for, and keeping track of the progress of a subgroup of students in a larger class. They may be peers who are taking the class for credit, or they may be students who have taken the class previously and have a position more like a TA (probably one or the other, depending on the school). In my opinion, peer study groups are one of the best ways to reinforce lectures and readings. Having someone who's already comfortable with the material serve as a "team leader" (something in between a supervisor and a tutor) is incredibly beneficial to this process, as is assigning some real value to the meetings—whether by awarding extra credit, incorporating attendance into overall grades, or simply designing the course in such a way that the group sessions are a necessity for full completion of regularly assigned tasks.

[Dopefish]

What I would hope to see more often is a set standard of material to be gone over each semester. Meet as a department, create curriculum which covers all the things you think a student should learn in their time at university and stick to it. Pick material and a pace that meets the talent of your students. Don't waver from it. Be transparent with what the syllabus of the course will cover. Be upfront about what knowledge is required coming in to the course.

My concern with that approach is that the inflexibility in what is covered can sometimes lead to some material being covered in a frantic flurry towards end, if the material up until that point didn't get as covered as quickly as expected (either due to downright poor teaching efficiency, or due to a class with lots of questions, or even something like snow days). If you get down to the last couple of weeks, but theres still material that the department is insisting must be covered, most profs (or at least the ones I've had) will fly through it at a rate that even the top students are unlikely to follow. (I'm pretty sure laplace transforms were [attempted to be] taught to me in the span of an hour in a single class, due to time running out in the semester.) The quality of learning that arrises from that would be so poor as to not be worth the time, and would be better spent doing something like review, to increase mastery of trouble areas that were covered to an adequate degree so people at least have some idea what's going on.

Classes that rely on that material that was covered inadequately or omitted entirely would still teach as if it was covered properly, or perhaps spend a class or two doing a lightning review of the material (which I'd probably recommend regardless, to refresh student's memories). Beyond that, it'd be on the student to figure things out (potentially making use of office hours). If the majority of the class has an issue, then potentially have a TA hold a tutorial session on the missing material, but it shouldn't take the place of the actual class time.

So basicly, yes, they should come up with a curriculum of what to cover in advance, based on what the students should hypothetically know based on pre-req's, but it should still be flexible enough that if it becomes impractical to cover some material, then it isn't covered, versus trying cover it super quickly and not being effective, since that just wastes everyones time.

[achan1058]

My concern with that approach is that the inflexibility in what is covered can sometimes lead to some material being covered in a frantic flurry towards end, if the material up until that point didn't get as covered as quickly as expected (either due to downright poor teaching efficiency, or due to a class with lots of questions, or even something like snow days). If you get down to the last couple of weeks, but theres still material that the department is insisting must be covered, most profs (or at least the ones I've had) will fly through it at a rate that even the top students are unlikely to follow. (I'm pretty sure laplace transforms were [attempted to be] taught to me in the span of an hour in a single class, due to time running out in the semester.) The quality of learning that arrises from that would be so poor as to not be worth the time, and would be better spent doing something like review, to increase mastery of trouble areas that were covered to an adequate degree so people at least have some idea what's going on.

You set a week by week schedule, with bonus topics which aren't important in the end. For example, I am reserving ~1 week of the class for numerical methods, which aren't in the syllabus. If there should be snow days, or drastically slow-down of the class, those topics will be cut. Of course, if your class ended up to be slowed down by 2 weeks or more, there's a much more fundamental problem.

[torgos]

For example, I TA for a calculus class at my college. Coming in, over half the class doesn't know their unit circle. They don't know what sin(0) equals.

I have a math degree, and I confuse my cosine/sine values all the time. Someone mentioned the unit circle a couple months ago, and I had to look it up("Oh, that thing"); when you're learning high school math, these are the sorts things you remember for a test and then forget(I don't even remember seeing the unit circle, although I probably did), because it's not clear what use they'll be in the future. Even though I've needed to know standard cosine/sine values in multiple courses over the years, I still have somehow never managed to retain them for longer than the test period.

Of course, these things should be learned on the student's time. It would generally be nice, though, if the professor would provide a handout of common things they "should' to remember from previous semesters but which they generally don't; after years of teaching, a professor should have some idea of what students don't retain from pre-reqs.

[ImTestingSleeping]

For example, I TA for a calculus class at my college. Coming in, over half the class doesn't know their unit circle. They don't know what sin(0) equals.

I have a math degree, and I confuse my cosine/sine values all the time. Someone mentioned the unit circle a couple months ago, and I had to look it up("Oh, that thing"); when you're learning high school math, these are the sorts things you remember for a test and then forget(I don't even remember seeing the unit circle, although I probably did), because it's not clear what use they'll be in the future. Even though I've needed to know standard cosine/sine values in multiple courses over the years, I still have somehow never managed to retain them for longer than the test period.

Of course, these things should be learned on the student's time. It would generally be nice, though, if the professor would provide a handout of common things they "should' to remember from previous semesters but which they generally don't; after years of teaching, a professor should have some idea of what students don't retain from pre-reqs.

Math is an enormous field and of course no one can remember everything. The difference between yourself and these students is that you have simply forgotten the material and they had never seen it. My idea of a successful undergraduate mathematician is someone who forgets mostly everything in 5 years (probably less) but when given a problem is able to reteach themselves relatively quickly. Problem solving is what makes the math man make the milk, man. (or woman, but the alliteration wouldn't allow it...)

[Griffin]

Grades are by their very nature a comparison to other students. I don't see how you could get around that.

I disagree with this at a fundamental level.

Grades should (and occasionally do), correlate solely with how much of a given set of material the student has mastered.

Contrary to the above statement, since high quality enthusiastic classmates are a huge determinant in the success of many students, grading on mastery generally means that you want as many other successfull students in your class as possible (to help you learn most effectively).

Contrary-wise, grading on comparison means you want the rest of your class to be as terrible as possible (so that you look better by comparison).

The students from the second class will be more inclined to see their classmates as competition rather than resources, and I don't think I need to go into the potential damage that can cause to the ability of a class to master material.

[ImTestingSleeping]

Grades are by their very nature a comparison to other students. I don't see how you could get around that.

I disagree with this at a fundamental level.

Grades should (and occasionally do), correlate solely with how much of a given set of material the student has mastered.

Contrary to the above statement, since high quality enthusiastic classmates are a huge determinant in the success of many students, grading on mastery generally means that you want as many other successful students in your class as possible (to help you learn most effectively).

Contrary-wise, grading on comparison means you want the rest of your class to be as terrible as possible (so that you look better by comparison).

The students from the second class will be more inclined to see their classmates as competition rather than resources, and I don't think I need to go into the potential damage that can cause to the ability of a class to master material.

I'm not sure you understood what I meant by grades being competitive by their very nature. I'm not saying we should give out grades by comparing the student to other kids in the class. I'm saying that a 4.0 would mean nothing when applying for a job if every one got a 4.0. I'm saying that in each of our heads we have some perception of what a "good" and "bad" grade is. It is in the very nature of the beast, if you will . You grade someone to have a record of where they are at. That person doesn't need to know their grade, because they know where they are actually at academically. It is to let other people know and that's where the competition lies.

I'm not sure if that is an appropriate response to what you said. If I misunderstood you, then please let me know.

[Edit]: I guess what I'm really saying is that I agree on a small scale such as a classroom that a high level of competition isn't a good thing. It is much better to get a group who wants to help each other out academically by positively contributing. However, on a large scale (such as job applicant population), the competition is very inherent and necessary.

[GenericAnimeBoy]

I'm saying that a 4.0 would mean nothing when applying for a job if every one got a 4.0.

How so? If grading were based on mastery of the material, having a good grade would indicate that the candidate has mastery of some set of subject matter, which frees the HR person doing the hiring to compare applicants with actually meaningful metrics, like "how well does this person fit into our corporate culture?"

What I don't get is how a comparison of one student's mastery and test-taking skills against other students who just happened to be in their university class is supposed to be at all relevant in evaluating an applicant's ability to do the job (Hint: It's not, and HR people only look at GPA because it provides a convenient filter for shrinking that stack of resumes as quickly as possible).

[Yakk]

The curse of statistics: If all you have are bad metrics, you still use them.

[achan1058]

I'm saying that a 4.0 would mean nothing when applying for a job if every one got a 4.0.

How so? If grading were based on mastery of the material, having a good grade would indicate that the candidate has mastery of some set of subject matter, which frees the HR person doing the hiring to compare applicants with actually meaningful metrics, like "how well does this person fit into our corporate culture?"

But when that happens, they would start asking for students with more mastery of a subject, like deeper understanding and etc. This means, they can ask for a harder course, like an honour's course, which not everyone will get 4.0 on, and now you are back to square 1. The nature of this world is about competition, and grades are sadly no different. At least, grades are more objective than the flimsy "how well does this person fit into our corporate culture?", which is IMO a even less useful metric, since you can't actually determine this without hiring the guy for some months.

[GenericAnimeBoy]

But when that happens, they would start asking for students with more mastery of a subject, like deeper understanding and etc. This means, they can ask for a harder course, like an honour's course, which not everyone will get 4.0 on,

Right. So universities would have an incentive to (heaven forbid) improve their fucking curriculum, as opposed to simply eliminating students who refuse to tolerate arbitrary and capricious obfuscation in classes they're goddamn well paying for (yes, I have an axe to grind ). If the material is presented in challenging honors courses where creative thinking is critical and not every student will get a 4.0 (I had assumed that "everyone" in the original quote was a figure of speech), GPAs would still represent a meaningful measure of comprehension. When universities simply make their curriculum arbitrarily more difficult or the grading curve arbitrarily steeper, they unwittingly favor brute force test-taking ability over creative thinking and innovation. It's very telling that many of the most successful and accomplished people in the world dropped out of college.

and now you are back to square 1.

No, now you have a bunch of people who, because of instructors taking the time to develop them, have acquired at a very young age the knowledge and skills and reasoning abilities to spend their lives creating wonderful things and making life better for all of us.

The nature of this world is about competition, and grades are sadly no different.

The competitive nature of the world actually implies that real life is not a zero-sum game.

At least, grades are more objective than the flimsy "how well does this person fit into our corporate culture?", which is IMO a even less useful metric, since you can't actually determine this without hiring the guy for some months.

People who succeed in getting a degree in fields wherein what degree you got actually matters (engineering, for instance) tend to be pretty consistently bright and capable in terms of book knowledge, thus the far more relevant criterion in choosing among applicants is whether they can work productively with the other people in the company. That is, in principle, the reason that companies in such fields pay pretty good money to undergraduate college students for "some months" (often in the summer) to come in and work on a project alongside experienced people in the company (it's called Internship, or in some cases Co-Op, and it's pretty difficult to get a job in one of these fields without it).

(Apologies for copious post-submission editing)

[achan1058]

But when that happens, they would start asking for students with more mastery of a subject, like deeper understanding and etc. This means, they can ask for a harder course, like an honour's course, which not everyone will get 4.0 on,

Right. So universities would have an incentive to (heaven forbid) improve their fucking curriculum, as opposed to simply eliminating students who refuse to tolerate arbitrary and capricious obfuscation in classes they're goddamn well paying for (yes, I have an axe to grind ). If the material is presented in challenging honors courses where creative thinking is critical and not every student will get a 4.0 (I had assumed that "everyone" in the original quote was a figure of speech), GPAs would still represent a meaningful measure of comprehension. When universities simply make their curriculum arbitrarily more difficult or the grading curve arbitrarily steeper, they unwittingly favor brute force test-taking ability over creative thinking and innovation. It's very telling that many of the most successful and accomplished people in the world dropped out of college.

A small list of people means nothing. Until you get to the point where the proportion of drop out being successful is higher than the proportion of graduates being successful, the metric is still somewhat useful. As for improving curriculum, you do know that in most cases in science/math/engineering courses, we ended up scaling grades up, right, even if we objectively do not think the test/assignment is hard? Many universities also added a bunch of "intro" courses as well. ie. We are dumbing down the curriculum so that the grades will not be horrible. And frankly, that's 1 of the reason for grade inflation, and I am certain that if we are to set tests the way we really want to (or at least certain profs do), you won't have a problem of too many people getting 4.0 at all. Indeed, you don't have the false dichotomy of mastery vs competition. Instead, you have the problem of "Am I going to pass".

and now you are back to square 1.

No, now you have a bunch of people who, because of instructors taking the time to develop them, have acquired at a very young age the knowledge and skills and reasoning abilities to spend their lives creating wonderful things and making life better for all of us.

The nature of this world is about competition, and grades are sadly no different.

The competitive nature of the world actually implies that real life is not a zero-sum game.

Zero sum or not, the total resource is finite, and you got to get the best bang for the buck.

At least, grades are more objective than the flimsy "how well does this person fit into our corporate culture?", which is IMO a even less useful metric, since you can't actually determine this without hiring the guy for some months.

People who succeed in getting a degree in fields wherein what degree you got actually matters (engineering, for instance) tend to be pretty consistently bright and capable in terms of book knowledge, thus the far more relevant criterion in choosing among applicants is whether they can work productively with the other people in the company. That is, in principle, the reason that companies in such fields pay pretty good money to undergraduate college students for "some months" (often in the summer) to come in and work on a project alongside experienced people in the company (it's called Internship, or in some cases Co-Op, and it's pretty difficult to get a job in one of these fields without it).

Indeed you have co-op, which is very popular nowadays. It is a useful thing to have, both for the employer and the student. Your comment on "People who succeed in getting a degree in fields wherein what degree you got actually matters (engineering, for instance) tend to be pretty consistently bright and capable in terms of book knowledge" I must say is dubious, at least for certain students. Maybe it's because I have been grading assignments for too long, or reading Ph.D. Comics too often.

[KestrelLowing]

Indeed you have co-op, which is very popular nowadays. It is a useful thing to have, both for the employer and the student. Your comment on "People who succeed in getting a degree in fields wherein what degree you got actually matters (engineering, for instance) tend to be pretty consistently bright and capable in terms of book knowledge" I must say is dubious, at least for certain students. Maybe it's because I have been grading assignments for too long, or reading Ph.D. Comics too often.

I think you haven't been in the 'real world' enough lately. It's amazing how quickly your views can be skewed. I went back to my high school after being in college to talk to some of the students there. Their general stupidity amazed me - I didn't remember it being that bad. However, I realize that's simply because I'm surrounded by people in a decent engineering program and have become accustomed to people being able to think logically.

Honestly, people are generally stupid - myself included! However, the majority of people in those type of majors would probably test at least above average on an IQ test. No, IQ tests don't mean a whole lot, but they mean something!

[Kag]

Maybe it's because I have been grading assignments for too long

I feel like the assumption that grades even vaguely represent what a student knows is highly suspect, even at the best of times.

[Bakemaster]

...even vaguely? Really?

[Game_boy]

The solution is the UK system: national standardised tests for all age 16-18 education (so everyone is graded on a curve for universitiy entrance purposes), and then universities are considered based on their reputation, not the degree class you got as long as it wasn't fail.

There are few enough good universities for everyone who needs to know having an idea of how good an institution is.

[KestrelLowing]

The solution is the UK system: national standardised tests for all age 16-18 education (so everyone is graded on a curve for universitiy entrance purposes), and then universities are considered based on their reputation, not the degree class you got as long as it wasn't fail.

There are few enough good universities for everyone who needs to know having an idea of how good an institution is.

I think you'll find many, many people disagree with standardized testing. It's done here in the US too - the ACT and SAT being the two major ones for college admission. Tests just cannot showcase everything a person is capable of, and frankly are really not representative of the potential someone has. (I've always done well on standardized tests, but I still think there's a huge problem)

[LilyN]

The solution is the UK system: national standardised tests for all age 16-18 education (so everyone is graded on a curve for universitiy entrance purposes), and then universities are considered based on their reputation, not the degree class you got as long as it wasn't fail.

There are few enough good universities for everyone who needs to know having an idea of how good an institution is.

I think you'll find many, many people disagree with standardized testing. It's done here in the US too - the ACT and SAT being the two major ones for college admission. Tests just cannot showcase everything a person is capable of, and frankly are really not representative of the potential someone has. (I've always done well on standardized tests, but I still think there's a huge problem)

Standardized tests are such bullshit. Remembering a ton of vocabulary, analogies and SAT 'trickses" do not show intelligence.

Another thing mentioned above is that professors are pressured into giving good grades. This is true until the professor has established his/herself. From my experience, older professors are typically much harder, expect more, and give worse grades. But it is always those classes where I learn the most. Coming from my major of mechanical engineering, the average GPA is much lower around 2.6-2.7. The classes where I STRUGGLED to get a B in ended being the most worthwhile.

[modularblues]

Whootles, list of easy classes!!!!! Why am I not surprised?

[Game_boy]

I think you'll find many, many people disagree with standardized testing. It's done here in the US too - the ACT and SAT being the two major ones for college admission. Tests just cannot showcase everything a person is capable of, and frankly are really not representative of the potential someone has. (I've always done well on standardized tests, but I still think there's a huge problem)

Standardized tests are such bullshit. Remembering a ton of vocabulary, analogies and SAT 'trickses" do not show intelligence.

I mean academic tests, not IQ. A national set exam?

like this:

http://store.aqa.org.uk/qual/gce/pdf/AQ ... SQP-07.PDF

Done by essentially every 18yo physics university applicant in the country.

I think you have AP, but that's not up to the level of A-level (evidence: isn't accepted by our universities when many other international qualifications are) and a large section of students don't have access to it.

Here your university admission is 90% dependent on your A-level grades. Not extra-curricular stuff or arbitrary GPAs from your school.

[KestrelLowing]

I mean academic tests, not IQ. A national set exam?

like this:

http://store.aqa.org.uk/qual/gce/pdf/AQ ... SQP-07.PDF

Done by essentially every 18yo physics university applicant in the country.

I think you have AP, but that's not up to the level of A-level (evidence: isn't accepted by our universities when many other international qualifications are) and a large section of students don't have access to it.

Here your university admission is 90% dependent on your A-level grades. Not extra-curricular stuff or arbitrary GPAs from your school.

Ah, we have different definitions of standardized tests. What you're referring to I would probably call a subject test, for lack of a better term. In the US, when something is a standardized test it's a test you can't really study for as you're not really sure what will be on it. Of course things tend to be the same over the years so you get an idea, but it's not a subject test. I certainly wouldn't call our ACTs or SATs IQ tests, but I think they may be closer to that than a subject test.

Looking through that test, it looked very similar to the AP Physics B test I took (algebra based physics - our school didn't offer C - calc based physics). I doubt the AP tests are not up to par with A levels, but are simply different. AP's are supposed to take the place of college level classes - not to be used for admission requirements. The point of AP is to take a college level class in college and the way to prove you've actually learned it is to take the AP test. It's possible to just take the test, but it's generally accepted that you should probably take an AP class as well. Out of curiosity, do you know if IB tests are accepted?

Basically, APs are for getting ahead so you don't have to pay so much money in college for entry level classes like Calculus, Physics, and Literature.

And yeah, it sucks that APs are not available everywhere, but that's America and capitalism for you.

I would be much more open to a system like that - tests that you can actually study for and know what will be on them. However, I do think that GPA and some extra curriculars and such are needed. Grades are typically a good indication of someone's work ethic and when paired with the results from the subject tests could probably be a very good indicator. Extra curriculars are used to show that people can handle extra work and sometimes allows for extra skills to be gained that may actually be useful for college. For example, I know many people are very impressed if someone is an Eagle scout - it shows multiple years of dedication and hard work - both are needed for college.

Still, I do like the idea of subject tests.

[Game_boy]

Ah, we have different definitions of standardized tests. What you're referring to I would probably call a subject test, for lack of a better term. In the US, when something is a standardized test it's a test you can't really study for as you're not really sure what will be on it. Of course things tend to be the same over the years so you get an idea, but it's not a subject test. I certainly wouldn't call our ACTs or SATs IQ tests, but I think they may be closer to that than a subject test.

OK. Thought it was something like that.

Looking through that test, it looked very similar to the AP Physics B test I took (algebra based physics - our school didn't offer C - calc based physics).

They don't include calculus in Physics, though the calculus we do in A-level Maths + Further Maths is easily able to handle it. I think it's government interference to encourage the uptake of Physics.

I doubt the AP tests are not up to par with A levels, but are simply different. AP's are supposed to take the place of college level classes - not to be used for admission requirements. The point of AP is to take a college level class in college and the way to prove you've actually learned it is to take the AP test. It's possible to just take the test, but it's generally accepted that you should probably take an AP class as well. Out of curiosity, do you know if IB tests are accepted?

OK. IB tests are accepted, they are considered to be equal to A-levels but you need to do more work for the same consideration.

Basically, APs are for getting ahead so you don't have to pay so much money in college for entry level classes like Calculus, Physics, and Literature.

Right. Here, we don't have such entry level classes except for the very weak. How many years is typical at universitiy for you (and from what age) up to a bachelor's degree?

And yeah, it sucks that APs are not available everywhere, but that's America and capitalism for you.

I do think the UK state education sector is better and more evenly funded than America's.

I would be much more open to a system like that - tests that you can actually study for and know what will be on them. However, I do think that GPA and some extra curriculars and such are needed. Grades are typically a good indication of someone's work ethic and when paired with the results from the subject tests could probably be a very good indicator. Extra curriculars are used to show that people can handle extra work and sometimes allows for extra skills to be gained that may actually be useful for college. For example, I know many people are very impressed if someone is an Eagle scout - it shows multiple years of dedication and hard work - both are needed for college.

The top universities here, (Oxford, Cambridge, UCL, Imperial) which can easily match America's Ivy League type universities, pretty much only accept based on A-level grades and interview*. When I went on open days those universities repeatedly stated that they didn't care about scouts etc., they'd actually rather you didn't mention it at interview, and they didn't want well rounded people (verbatim).

I don't believe they affect you ability to study and from what I can see it's unfair on the academically brilliant but not well rounded people in the US who have to waste hours on that kind of thing just to look good on the application when they could be studying. Especially for the sciences/maths.

*Interviews are almost purely academic, asking you to talk through given problems.