<p>A couple of years ago, I got to go to South Africa and go on a safari led by an associate engineering prof at the University in Johannesburg. He had a bunch of exams to grade, so he asked me to help him. I graded quite a few. At one point, he asked how it was going. “Fine, but your students aren’t doing very well!” “Really? That surprises me. What are the grades like?” “A lot in the high 70s. A few in the mid-80s. One 90.” “Oh, that’s actually very good! Those are excellent grades for classes at my university.” So everything is relative.</p>
<p>It’s totally up to the professor as to how to assign grades. That’s life! If he or she makes his standards clear at the beginning of the course, that’s all you can expect. My dad is a professor, so I gues I see things more from his standpoint.</p>
<p>Even when the instructor does not “grade on a curve”, s/he may choose to make the standards different from the typical high school scale of 90% = A, 80% = B, 70% = C, etc…</p>
<p>There is no magic to 90%—if that’s the cutoff for an A, it’s an arbitrary cutoff. At least bell curves aren’t arbitrary.</p>
<p>Really, though, the question is of curving comes down to whether grades should be criterion-referenced or norm-referenced (or some sort of hybrid). There are pluses and minuses to all approaches, and you can’t say that one is always better than another—it depends on the class, the population of students, the position of the assignment within the course and the course within the wider curriculum, and so on.</p>
What is the example where criterion-referenced should be preferred to norm-referenced with the exception of assigning “D” and “F” grades which have an objective result of failing the class? I cannot think of one.
The classic example of useful criterion-referencing is a situation where a specific skillset is the desired outcome, but there are gradations of ability with regard to that skillset. Under those circumstances, an A could show, say, exceptional ability with that skillset while a B could show good but not exceptional ability, and so on. When such is the need, it doesn’t matter how others have done (i.e., norm-referencing), since all that matters is the learner’s own demonstrated ability.
How can you define “exceptional” or “good” except as a reference to others? If we view the level of mastery of the subject matter of a class to be continuous (as opposed to discrete) and extending beyond the abilities of even the single best student, then how can we define “exceptional” in absolute terms. I’ve never heard of a class where these assumptions of continuity and extending beyond the abilities of the students weren’t the reality (though I’ve had classes where ineffective testing has made the latter assumption hard to see) but if this theoretical class existed, it would require that it not be a subject matter that is capable of research in the subject and would lend itself better to a binary certification process rather than a college class in the way we think of it.
@Vladenschlutte, you said it yourself: “If we view the level of mastery of the subject matter of a class to be continuous (as opposed to discrete) and extending beyond the abilities of even the single best student…” (emphasis added)—despite the “if”, you’re assuming, as far as I can tell, that all classes work that way, except in the possible situation where there’s a binary pass/fail option. This is not the case, however—consider, if nothing else, some graduate programs, where the distinction between a “pass” and a “high pass” on one’s master’s thesis can determine entrance into the doctoral program.
I think it’s helpful to look at this in terms of a scenario that’s outside of academia. We all want our restaurants to be sanitary and clean. That’s why we have health inspections. If they score above a certain range they are considered up to date with health codes. If they score just below that range, they’re considered to be mostly up to date, but in need of some improvements that must be taken care of by the next inspection. Just below that range falls the restaurants with more concerning health code violations…etc…all the way on down to the restaurants that get closed due to failing their inspection. Clearly we all (I’m assuming) would prefer that our restaurants be up to date with health codes. So we desire our restaurants to fall in the first couple of categories. The first category is ideal, but the second category is also very acceptable.
How would you feel about health inspectors curving their findings? Does a restaurant with some more severe violations now get bumped up and given the all clear to keep running as is?
This is pretty similar. I don’t care about how I’m doing in a class in comparison to the rest of the people in the class. I care about how I personally am doing with the content itself…and I measure that by my grades. If I’m told that a restaurant got a C on their last inspection, I might think twice about eating there. If that C has been curved up to an A though, I wouldn’t really even have the opportunity to decide.
I might agree in that case, mainly because there are fewer students for whom this applies and thus you don’t have an appropriate sample to compare to. I suppose that should be another assumption, that norm-referencing should be reasonably possible. However, that’s out of the domain that we’re talking about, it’s not a regular course.
The health inspection is completely different, and doesn’t meet either the assumption of extension beyond the capability of the most competent restaurant or the assumption of continuity in capability the way it’s measured. It’s very clearly doesn’t apply.
How can one determine how one is doing with the content itself? There’s only a handful of people in the world for any subject who are going to know everything about that subject. By that logic, no one should have a top grade. Hell, hardly any more are going to “know more than 90% of a subject” (which I realize is a hard concept to define - just go with it). Almost certainly no undergrad in the country would. So, should no one in any class get an ‘A’?
Basically, @Vladenschlutte, it appears to me that you do not entirely understand what grade assignment involves (e.g., a grade of 90% doesn’t mean one knows 90% of a subject, it means that one scored 90% on an assessment—there’s a rather huge difference). This may well be the core of the problem here.
I realize that and it was part of my point… A 90% would signify that they got questions in the class 90% correct, that they scored 90 points on one particular measurement. How can you use that to find one’s knowledge of a subject.
Because the people designing the tests are PhD trained faculty who have demonstrated a level of expertise in the subject at hand. They are then tasked with creating exams that they think are good at assessing a student’s abilities.
Who is the faster runner: the person who comes in last place of the olympic finals or the one who comes in first place at the local YMCA saturday fun run? Grading on a curve puts a premium on the competition around you and there’s no real (good) reason to do that in grading in school.
I don’t think anybody, least of all myself, is suggesting that scoring a 90% in a class equates to knowing 90% of a field. One can complete an entire undergraduate degree in a given field and still know nowhere near 90% of a given field. I finished my Western Civilization course with a 96%…and clearly that does not mean that I know 96% of the history of Western Civilization. It means that I satisfactorily completed 96% of the material that was assigned. If I get 100% on a calculus exam, it doesn’t mean that I know everything there is to know about that given chapter. It simply means that I demonstrated a good knowledge of all the material over which I was being tested. It doesn’t matter to me that I scored higher than 80% of the class and lower than 20% of the class or anything like that. To me personally, that is not an accurate assessment of my ability to utilize the things that I’ve learned in my classes. An accurate assessment to me involves a percent score of the accuracy of my own exam.
Why even talk about nonsense like this? Who’s better at math, the best guy in the special ed program at your local high school or the worst math PhD candidate at any school? So what is it, should every college student get an A in every class because they’re smarter than the mentally disabled or should special ed programs only give out "F"s?
So say the professor made the test really hard, so now you scored 20% but it was still better than 80% of the class and worse than 20%? Everyone should just receive an “F?” Say your 96% was the worst in the class, you still deserve an “A?”
Say I’m an employer or something, reviewing someone’s transcript. If everyone gets an A in all your courses and I see you have all As, what can I do with that information?
If I scored 20% on an exam, I would indeed earn an F. If many people in the class scored in the same area, it would be indicative of the professor needing to change the exam format. I was in an introductory chemistry class several semesters ago, and this is basically what happened. The professor made the exam too long, and far too difficult. Only myself and one other person in the class (out of ~45) ended up scoring above a 70%. The professor took that as a sign that the test was unreasonably difficult, and rewrote the test, then had us retake it the next week. Then the grade distribution was about typical.
If my 96% was the worst in the class…then yes, I still deserve an A. Why shouldn’t I? Suppose that 1/5 of the class scored 100% on the exam, another 1/5 scored 99% on an exam, the next 1/5 scored 98%, the next 1/5 97%, and the final 1/5 scored 96%. Since the people who scored the worst in the class all scored “only” 96% (due to an arithmetic error, lost negative sign, 1 incorrect multiple choice answer etc.), do they deserve an F on the exam by merit of being at the bottom of the pile of -exemplary- exam grades? If an exam has 50 multiple choice questions at 2 points each, one could score a 96% on the exam by getting two questions wrong. If everyone scored higher than this 96%, it does not logically follow that the person with the 96% should get an F.
What employers are asking about whether or not class grades were curved? My transcripts only show my own grades. They don’t show the grades of my classmates or indicate where I fall in comparison to the rest of the class.
Had to submit my undergraduate transcript for a grant. No one asked if my classes were curved.
My point with regard to YMCA vs Olympics is that not everyone is taking the same test against the same people and so you’re using grades to compare in a way that they can’t be used. Curving does not make grades a better representation of a students ability. They make learning a zero sum game - a completely unnecessary practice and one that’s detrimental. I go to school to learn, not to compete.
I also can’t blanket say what grade a 20% should be. It depends on what the professor thinks a 20% on their exam should be but if everyone gets a 20% then everyone should get the grade the professor thinks 20% should get (which was stated in advance on the syllabus obviously)