# Grading on a scale is stupid

1. Mar 21, 2016

### bluemoonKY

I have always thought that grading on a scale is stupid. When a teacher/professor grades tests based on a scale, the teacher grades each student based on their performance compared to other students. This is a stupid way to do things because it can both cause some students who know the curriculum well to fail a class, and it can cause students who don't know the curriculum to pass a class (or even potentially get an A in a class!). If everyone in, say, a calculus class knows how to solve all types of calculus problems on the tests compentently, the teacher/professor should pass everyone in the class. If nobody in a calculus class knows how to solve calculus problems on the tests at an adequate level, the teacher/professor should fail everyone in the class.

Many times in my life I have heard people say that teachers/professors should grade rigorously and fail incompetent students for two primary (and related) reasons: 1) passing incompetent students harms the reputation of the school or college and 2) if you pass incompetent students, it will result in incompetent /workers in the real world. I agree with both of these reasons. However, if teachers/professors grade the way i'm saying they should grade, there is no danger of teachers/professors passing incompetent students harming the reputation or the school or causing incompetent workers in the real world caused by the teachers' grading.

Let's take the situation of students majoring in aeronautical engineering at a university taking a calculus class. I've heard people use as an example the idea that teachers/professors should not pass incompetent calculus students because " it will result in planes falling out of the sky and crashing." These people are (correctly) implying that the engineers' ability to design planes that won't fall out of the sky and crash is predicated to an extent on the engineers' calculus skills. If everyone in the calculus class in our example ( including the aeronautical engineering majors) is competent at solving all types of calculus problems, and the professor thus passes everyone in the class, then this will not result in planes falling out of the sky due to incompetent calculus students being given passing grades because everyone in the class is competent at solving calculus problems. Conversely, if everyone in the same class (including the aeronautical engineering majors) were very incompetent and awful at solving calculus problems and the professor graded on a scale and gave the least worst calculus student (who is still VERY incompentent) a grade of A and that least worst calculus student is an aeronautical engineering the major, then the professors' grading on a scale could result in planes falling out of the sky and crashing due to the incompetent student receiving a grade on a scale of an A.

2. Mar 21, 2016

### Staff: Mentor

You are completely ignoring the main reason behind the idea of the grading to scale. Can you at least explain what it is, so that we know your opposition doesn't come from ignorance?

3. Mar 21, 2016

### bluemoonKY

If you asked teachers/professors who grade on a scale why they grade on a scale, they would probably say something like they think that there should be as many A's and B's as there are D's and F's (and vice-versa). But the real (off the record) reason that teachers usually grade on a scale is because the students are so incompetent that if the teacher didn't grade on a scale, the teacher/professor would end up failing just about the entire class. Probably another (less common) reason that some teachers might grade on a scale is to prevent grade inflation. However, both reasons are stupid reasons to grade on a scale for the reasons I said in the original post.

4. Mar 21, 2016

### Staff: Mentor

5. Mar 21, 2016

### Staff: Mentor

And you know that from... ?

6. Mar 21, 2016

### micromass

I agree. It's stupid. I try never to curve the grades or to grade on a scale. But this requires a lot more effort in designing exams and tests. You'll need to design an exam which is hard enough to make sure incompetent students fail, but easy enough so competent students don't fail. This is a tricky balance which is quite hard to achieve.

7. Mar 21, 2016

### bluemoonKY

Berkeman, yes, grading on a scale = grading on a curve. I've seen and heard teachers/professors use both of them, but they both mean the same thing.

8. Mar 21, 2016

### bluemoonKY

Borek, I know that the real (off the record) reason teachers grade on a scale is so that the teacher doesn't have to fail the entire class from my years of observation at both public schools and in college. I think the reason micromass gave could be another reason, especially in upper level college courses. In public school, the reason micromass gave makes less sense to me because the range of student abilities is/was so large.

9. Mar 21, 2016

### bluemoonKY

micromass, based on the *extremely* wide range of abilities I observed in most of the classes I attended when I was in school, I wouldn't have thought it would be that difficult to design an exam that would ensure that incompetents fail and competent students pass, but you would know better than I would since you are a teacher.

10. Mar 21, 2016

### micromass

Sure, I know it sounds weird. But believe, designing good exams is difficult. The very first exams I designed were much too difficult. They took way too long to solve and required some ingenious techniques which are very much obvious to me, but not to students of that level. I think every teacher should grade on a curve the first few years until he knows the abilities of the students more.

11. Mar 21, 2016

### bluemoonKY

micromass, are you a public school teacher (K-12) or a college professor?

12. Mar 21, 2016

### Staff: Mentor

For the record: I don't care about grading approach, I just hate skewed/partial arguments no matter which way the go

And micro hit the nail on the head. It is like with taking photographs. If you choose wrong parameters and you end with either over- or underdeveloped pictures, then the only way of saving is to modify the histogram. That's what grading on the curve does.

13. Mar 21, 2016

### micromass

I did both.

14. Mar 21, 2016

### Mondayman

I have quite a few friends in engineering at my university, and they have a saying, "Ride the Curve". They have even gone as far as to design a T-shirt that has a stick man riding a surfboard along a normal distribution. While the shirt is funny, it is a little disturbing to hear that in some of their classes the average grades hover around 30-40%. But I wouldn't know if this is the instructors intention or the fault of the students.

15. Mar 21, 2016

### bluemoonKY

I can see how a teacher would need to grade on a curve to separate incompetents from competents in courses that are subjective like history. But in a math course, I don't see why a teacher would need to grade on a curve. Why can't mathematics teachers/professors just use their mathematics textbooks as a source for the problems on the course and just change the coefficients? Mathematics textbooks usually feature a math problems in a range of difficulties.

16. Mar 21, 2016

### micromass

Good question and I'm glad you're asking this. First of all, this holds for rather computational courses like calculus only, not for more advanced math courses.
Second, how many problems should you give? Let's say the exam lasts two hours. How do you make sure the exam is not too long or too short. If the exam is too long, everybody will get bad grades even though they don't deserve it. It is very tricky to make an exam with the right amount of problems.

17. Mar 21, 2016

### bluemoonKY

I still don't see why it's difficult to choose the right parameters in mathematics. I mean, if a teacher is teaching calculus II, why would it be difficult to choose the right parameters? There are limits to the amount of fudging that can be done with the course content. The teacher must teach integration by parts, integration by partial fractions, integration by substitution, parabolas, and sequences and series. The teacher must include problems on tests requiring integration by parts, integration by partial fractions, integration by substitution, parabolas, and solving for whether or not different types of power series and other types of series converge or not. If all the students can solve all the types of calculus problems that I just listed, they are all competent. If none of the students can solve any of the types of calculus problems I just listed, they are all incompetent. If the students can solve the different types of calculus problems I just listed to different degrees, the students are individually as competent or incompetent to the degree to which they can solve the different types of calculus problems I just listed, not competent or incompetent based on how well they do in comparison to each other.

18. Mar 21, 2016

### bluemoonKY

You should err on the side of giving them more than enough time.

If the exam last 2 hours, you should give the amount of problems that you think that they should be able to do in an hour.

19. Mar 21, 2016

### micromass

So then some people will be able to solve it completely while being too slow.

20. Mar 21, 2016

### bluemoonKY

Micromass, since when was speed supposed to be a determining factor on someone's grade on a course on anything other than keyboarding? I always thought that when teachers give math tests, the math tests are strictly supposed to test whether or not the student knows the course content, not the time it takes the student to answer it.

21. Mar 21, 2016

### micromass

You don't think speed should somehow be a determining factor? I'm not saying everybody should be super fact problem solvers, but if you need to take 2 hours for solving $x^2 - x +1=0$, then you should fail the course even if you do find it after 2 hours.

22. Mar 21, 2016

### bluemoonKY

I think that it wouldn't be that difficult for a teacher to determine what would be a reasonable amount of time to solve mathematics problems. I don't think it's any reason for a teacher to need to grade on a curve.

23. Mar 21, 2016

### micromass

It is that difficult. I have made errors in judging this multiple times.

24. Mar 21, 2016

### Staff: Mentor

I second micro's comment. It's hard, at least at first. The first intro physics exam that I wrote 33 years ago was a disaster. With time, you get a feel for how your students will do on different questions. Also, certain types of questions are very difficult to design and chose effectively according to difficulty: true/false and multiple choice. For those questions, grading is binary, you either get it right or you don't. That's why I've always preferred short-exercise questions, because I can adjust partial credit to some extent.

25. Mar 21, 2016

### Staff: Mentor

Does it take you exactly the same amount of time to find every integral using integration by parts, or do you find some integrals harder than other?

Have it ever occurred to you that you have spent several hours trying to solve a problem, only to find out it is in fact trivial, you just looked all the time from the wrong side? Have it ever happened to you to fail an exam because of that?

If you think it is all simple and trivial you just lack imagination.