4 ms·

There’s no such thing as an objective standard. Professors change, course materials change, and as with here, circumstances can be different. That’s why we use

by willj 7y ago

There’s no such thing as an objective standard. Professors change, course materials change, and as with here, circumstances can be different. That’s why we use a curve.

WalterBright 7y ago
Caltech institute policy was no grading on a curve. The standard was to be decided on in advance. If half the students failed an exam (which happened) so be it. If half aced it, so be it as well. Stack ranking in employee evaluations is unfair, and is the same thing as curve grading.
- jedberg 7y ago
  Stack rankings and curves aren’t anywhere close to the same. Grading on a curve allows everyone to get a perfect score. Stack ranking means that if everyone is perfect only one person can get a perfect score. They are in fact opposites. Also, if half the class fails, that is a teaching failure, not a student failure. The teacher failed to teach the material and/or failed to write a good test. At Berkeley, if 1/2 the class failed, the professor usually wasn’t allowed to teach that class anymore.
  - WalterBright 7y ago
    > Stack rankings and curves aren’t anywhere close to the same. Sure they are. The best get a raise/A, the worst get fired/F. > that is a teaching failure, not a student failure Let's put it another way. Do you want the pilot flying you across the Atlantic 747 to be graded on a curve where he happened to be better than some of the other students, or graded against "met the standard / did not meet the standard"? Me, I want "met the standard" every time for someone I trust with my life. The teacher failing to teach is not a reason to pass a pilot who didn't meet the standard.
    - jedberg 7y ago
      > > Stack rankings and curves aren’t anywhere close to the same. > Sure they are. The best get a raise/A, the worst get fired/F. That's actually not at all how stack ranking works. The best might get an A, or a B or a C, and the worst might get a B or a C or a D or an F. An example: You give a test that is out of 120 points. Here are the scores: P1 = 100 P2 = 99 P3 = 98 P4 = 97 P5 = 96 Under a curve grading, they all get A's. Under objective grading, they all get B's. With stack ranking: P1 = A P2 = B P3 = C P4 = D P5 = F Should P5 be given an F or get fired? That's what happens in stack ranking. > Let's put it another way. Do you want the pilot flying you across the Atlantic 747 to be graded on a curve where he happened to be better than some of the other students, or graded against "met the standard / did not meet the standard"? A pilot is not the same as your undergrad physics class. The pilot was taught by instructor who has to teach in a specific way, often using specific words, developed over many years by a committee of pilots. They then take a test that 100s have taken over many years, written by a committee of pilots. The pilot instructor has very little leeway on how they teach the class and has no leeway in the test. Same with medical boards, or any other standardized test. But your physics professor probably wrote the curriculum themselves, teach it a different way every semester, and use a slightly different test every time. And as much as you wish it were true, none of your undergrad classes are life or death scenarios. So yeah, it makes sense to grade on a curve for an undergrad class.
      - WalterBright 7y ago
        > Under a curve grading, they all get A's. I think you misunderstand what curve grading is. > none of your undergrad classes are life or death scenarios. Except that bridges fall down and kill people when designed by incompetents. You bet your life every day on the competence of engineers. Employers are NOT interested in curve ranking. They're interested in a degree as a mark of meeting a standard. Caltech's policy is that their graduates meet a standard of quality. A curve is not a standard. Someone who gets sympathetic good grades because of a bad teacher is not of interest. If you don't accept that, there are plenty of other schools to attend. But I suspect that if you think about it, you expect a standard of quality in everything you buy - goods and services. You're not interested in a flexible curve.

by willj 7y ago

There’s no such thing as an objective standard. Professors change, course materials change, and as with here, circumstances can be different. That’s why we use a curve.

WalterBright 7y ago

Caltech institute policy was no grading on a curve. The standard was to be decided on in advance. If half the students failed an exam (which happened) so be it. If half aced it, so be it as well. Stack ranking in employee evaluations is unfair, and is the same thing as curve grading.

jedberg 7y ago
Stack rankings and curves aren’t anywhere close to the same. Grading on a curve allows everyone to get a perfect score. Stack ranking means that if everyone is perfect only one person can get a perfect score. They are in fact opposites. Also, if half the class fails, that is a teaching failure, not a student failure. The teacher failed to teach the material and/or failed to write a good test. At Berkeley, if 1/2 the class failed, the professor usually wasn’t allowed to teach that class anymore.
- WalterBright 7y ago
  > Stack rankings and curves aren’t anywhere close to the same. Sure they are. The best get a raise/A, the worst get fired/F. > that is a teaching failure, not a student failure Let's put it another way. Do you want the pilot flying you across the Atlantic 747 to be graded on a curve where he happened to be better than some of the other students, or graded against "met the standard / did not meet the standard"? Me, I want "met the standard" every time for someone I trust with my life. The teacher failing to teach is not a reason to pass a pilot who didn't meet the standard.
  - jedberg 7y ago
    > > Stack rankings and curves aren’t anywhere close to the same. > Sure they are. The best get a raise/A, the worst get fired/F. That's actually not at all how stack ranking works. The best might get an A, or a B or a C, and the worst might get a B or a C or a D or an F. An example: You give a test that is out of 120 points. Here are the scores: P1 = 100 P2 = 99 P3 = 98 P4 = 97 P5 = 96 Under a curve grading, they all get A's. Under objective grading, they all get B's. With stack ranking: P1 = A P2 = B P3 = C P4 = D P5 = F Should P5 be given an F or get fired? That's what happens in stack ranking. > Let's put it another way. Do you want the pilot flying you across the Atlantic 747 to be graded on a curve where he happened to be better than some of the other students, or graded against "met the standard / did not meet the standard"? A pilot is not the same as your undergrad physics class. The pilot was taught by instructor who has to teach in a specific way, often using specific words, developed over many years by a committee of pilots. They then take a test that 100s have taken over many years, written by a committee of pilots. The pilot instructor has very little leeway on how they teach the class and has no leeway in the test. Same with medical boards, or any other standardized test. But your physics professor probably wrote the curriculum themselves, teach it a different way every semester, and use a slightly different test every time. And as much as you wish it were true, none of your undergrad classes are life or death scenarios. So yeah, it makes sense to grade on a curve for an undergrad class.
    - WalterBright 7y ago
      > Under a curve grading, they all get A's. I think you misunderstand what curve grading is. > none of your undergrad classes are life or death scenarios. Except that bridges fall down and kill people when designed by incompetents. You bet your life every day on the competence of engineers. Employers are NOT interested in curve ranking. They're interested in a degree as a mark of meeting a standard. Caltech's policy is that their graduates meet a standard of quality. A curve is not a standard. Someone who gets sympathetic good grades because of a bad teacher is not of interest. If you don't accept that, there are plenty of other schools to attend. But I suspect that if you think about it, you expect a standard of quality in everything you buy - goods and services. You're not interested in a flexible curve.