<p>I agree with mathmom-if a class is well-taught and students do the work, no one should have to fail. In one class I took, towards the end of the semester the professor said the class as a whole had done better than his classes usually did, so he was going to make the class B-centered instead of C-centered that term.</p>
<p>Would it stand to reason that the use of curves, and the severity of curves, depends on the selectivity and size of the university?</p>
<p>Clarification: If the selectivity is already high, professors have no need to weed out students, etc. If selectivity is lax, there are LOTS of students and a need for some kind of process to direct them toward appropriate majors/studies.</p>
<p>doubleplay,</p>
<p>Here in Ontario, universities try to maintain a C to C+ average in all larger classes as a matter of tradition, so the use of curves in one form or another goes as far back as I can remember. The severity of the curve do vary depending on the selectivity of the university, and to an even greater extent, the limited size and high selectivity of specific programs. </p>
<p>Unfortunately, it is not done to weed out weak students as much as to weed out strong students in favor of even stronger students. In short, the more competitive the program, the more severe the curve. </p>
<p>My younger sister, for example, was accepted, back in the 1980s, into the business program of one of the top three universities in Canada. To continue into the second year of the program, she had to take three required courses in the first year and all she had to do was to maintain a suspiciously low C- average on each one of them.</p>
<p>One of the courses was a math course designed specially for these “future” business students and she remembered that there were well over a thousand students in that class at the beginning of September. By Christmas, it was down to a more manageable 280 or so.</p>
<p>Naturally, these 280 “survivors” continued to battle the curve and each other throughout their three remaining years. When my dear sister graduated, she had a B average, a whole grade lower than the solid A average she had in high school. </p>
<p>After graduation, she was looking at the possibility of entering corporate law. She wrote the LSAT and scored in the 92%tile range without prep. (Nobody prepare for those tests then, did they?). It took us a long time to make sense of this discrepancy.</p>
<p>Pragmatically speaking, using a curve means you can issue hard tests. The curve (or grade shift) compensates for the fact that no one scores above, say, 70% on the original test.</p>
<p>I am a huge fan of curving grades in larger classes. A secondary issue is the particular distribution of As, Bs, Cs, Ds, and Fs for the class … the prime benefit is the ability to differentiate the differences in performance of the students. </p>
<p>There are many discussions on this board about how the SAT does not differentiate well among the really high scoring students … how the upper tail of students is smashed into th 780-800 range. To me gving a test where the vast majority of students score 90+ seriously questions the precision of any differentiation among the students … were the questions and the grading precise and consistent enough that the prof can say the 93 was really better than a 91?</p>
<p>Personally I think the best test average on a curved test is a 50% score. This allows the full range of student performance to show through. I’ve lived this from both ends … when I was a goof ball sophomore having a curve with an average of 50% or so saved me when I got a 25% score (25% compared to a 90% average would have cooked me) … and going the other way getting a 92% when the mean was 45% showed I really knew the material.</p>
<p>Jessehl hit it on the nose … a mean near 50% allows the spread of student’s performace (standard deviations from the mean) to be well understood and more precise … then the question rises of what is the appropriate grade scale given a now clearer picture of the ranking of th student’s relative understanding of the material.</p>
<p>When the need is to weed the curve is brutal. For example, Chem 101, 450 students in the first quarter (mostly Pre-Med), 69 was good enough for a B. Chem 102, 300 students left, 76 was a C. Chem 103, 150 students left, 81 was a C. For Organic Chemistry, sophomore year, there were only about 100 Pre-Meds left. In the huge freshman seminars we were told there were guaranteed Fs in the standardized tests. It was a straight curve (bell) I guess with a lot of room for Cs.</p>
<p>Some teachers give extra hard tests that cover everything, no student is expected to answer every question, so two grades of 82 do not mean that the students even answered the same questions. Students can show their areas of expertise (or lack of). Highest grade becomes 100, curve goes from there.</p>
<p>Some teachers even told you that not even the author of the textbook would get 100 in the test (not even an A, unless it was curved). There are even some that would say that only the teacher would get an A if he took the test.
I can tell you those were not very good teachers or very pleasant classes.</p>
<p>How each teacher curves is up to them, but at least they should explain their “method” in the beginning of the course.</p>
<p>For less quantitative courses, the “faster than a speeding bullet” approach works pretty well, although I would probably use phraseology more like, “A=exhibits mastery of the concepts taught in the class and is able to apply them correctly and creatively.” If you set out expectations like this, it is my experience that students do group themselves pretty well into the categories. Also, when you do it this way, it’s OK to have 30% A’s one semester, and 10% the next semester, and you only have to fail somebody when they really have failed to learn anything.</p>
<p>I was surprised at a UC in the bio major, they still followed that “weeder” curve even in senior lab classes- so with less than 20 kids, they still had a number of kids who would not earn a passing grade, who would need to repeat it. I actually verified this policy, in person, at a meeting for another purpose- I can see hte weeding in lower div classes, but once you’ve survived Ochem, why can it not be a situation wherein every one COULD succeed??</p>
<p>If a student gets an F even if he has mastered the material, he should sue the school for fraud.</p>
<p>somemom, my thoughts exactly! Just what IF there is a small class of exceptionally hard-working and bright students? I know I’ve had several hs teachers talk about how the make-up of their classes varies widely from year to year. Over several years, there’s probably a bell curve lurking in there, but for any given year, especially in a small class, I don’t think it is realistic to expect it.</p>
<p>My D’s nursing school is brutal. A passing grade is 77, and an A must be over 93. There are no curves. Two weeks ago 51% of the class failed an exam. Several days ago 33% of the entire year’s students were given letters telling them that they are failing out. And these are extremely motivated students. To me, that shows some pretty bad teaching. A curve would look pretty good to a large portion of that school right now! I think that if tests are hard enough for a large percentage of a class to fail, either the test is badly made or the material is badly taught.</p>
<p>Nursing programs are famous for weeding out their students. All the programs aim for a 100% passing rate on the state nursing boards, so they don’t string students along. It may be that the students didn’t realize the tremendous amount of memorization required. The material may be taught well & understood completely, but finer details were not committed to memory. (An example might be: Does XYZ drug act in the ascending limb of the loop of Henle or the descending limb? Often green nursing students are just not expecting they will be required to master the fine points of pharmacology or A&P.)</p>
<p>As a music professor in applied music, this was my (not altogether) joking version of criteria for grading student degree recitals: </p>
<p>A: would pay to hear the performance.
B: am glad to hear the performance.
C: would attend performance if paid to do so.
D: would pay not to attend performance.
F: would not attend, regardless of the consequences. </p>
<p>If you consider the C the lowest passing grade that the professor should expect to hear and is being paid to attend, it seems reasonable.</p>
<p>In the science classes S has encountered there is a curve for each test, but the final grade is from a total numerical curve, so the “grade” on the exams can be deceiving. One high scorer really throws off the curve when it is curved down from the high, rather than as a distribution curve. In a sociology class, each individual exam was curved, and then after the final, the entire class grade was recurved based on total points. There are lots of ways to do curve grading. It seems like mastery of material in the discipline should set the grade. Curving seems to allow lots of non-mastery, or poor teaching.</p>
<p>“Just what IF there is a small class of exceptionally hard-working and bright students?”</p>
<p>The answer is you lose. At UT-Austin (a couple of decades ago) there were two identical classes in economics offered by the same professor, but at different times. The professor gave identical tests and applied a curve based on the class, so that an 80 in one class might be a B and an 80 on the same test in the other class might be a C. The early morning class attracted the more serious students and was therefore the more difficult class in which to score a good grade.</p>
<p>Now that would be a crummy thing. One could ask if he could have spread the curve over the 2 courses…would that have been more fair, or am I missing something?</p>
<p>“Just what IF there is a small class of exceptionally hard-working and bright students?”</p>
<p>With the exception of the “small class” part, that is what happens sometimes in large classes, after the drop period. All the kids who got Ds and Fs on the first exam drop, then the curve is based on the A through C kids. Word to the wise in large pre-med classes- make SURE you ace the first test, because after that, it’s going to be a bumpy semester.</p>
<p>Hmmm, another D’s OChem class at Cal had 1200 kids in multiple lecture sections, but the group for the curve was the entire 1200 kids!</p>
<p>Ah, the curve. Funny story - I recall leaving an exam in Advanced Calculus (4th semester of Calc) knowing I’d flunked and wondering what on earth went wrong, as I had studied the same as I usually did and was doing B work in the class. Got the test back, saw my percentage score - 22% - thought F, then saw the letter grade - a B!</p>
<p>The professor began to speak. “The average score on the exam was a 19%”. No one ever spoke in this class, but at this news a few began to grumble. It seemed across the board in engineering/math that they shot for 50% as an average; this was low. The professor cut them off immediately, saying “But that fellow in the rear row got a 97%, and there were two others in the 90’s”. As he pointed them out to us, these guys had their heads buried 
but it was in interesting lesson in ‘the curve’.</p>
<p>Just to complete my thought in #23…</p>
<p>For a long time, I was wondering what happened to those who were weeded from the program. Years later I learned that the students would cascade down the hierarchy to the next most competitive/popular program which would in turn use the curve to force the extra students to cascade down to the level below until we reach programs that have open admission. So, by knowing the Ontario university and the major, I can pretty well tell how academically strong a given individual is.</p>
<p>Fast forward to the present…</p>
<p>My D recently graduated from another one of the top three in their highly competitive business program. Unlike the previous school which accept all students with an A average and then eliminate as many as necessary until they get down to a number they can accommodate, this one accepts students directly out of high school. In my D’s year, they enrolled a little over 200 students out of an application pool of 4500. Once you are in, they do not actively trying to weed you out, although some did not make it to graduation.</p>
<p>Instead of keeping the average grade at C or C+, this one keeps it at B- for individual courses. Since it is an “honors” degree, a student can not stay in the program if the average drops below a B-. IOW, if you are unfortunately enough to get a C, you better have an A, or more than one B to “neutralize” it.</p>
<p>Since we don’t do ACT/SAT here, the program accept students based on grades (with certain required courses) and EC. My D went in with a 94% average (A=80), very high for high school graduates but just average for the program. You know, when she graduated, she had a B average, exactly average for her graduating class. In short, she was average going in and average coming out. I guess a teacher can make a mistake with a student but their collective opinion is pretty accurate. What quality control though!</p>
<p>I used to really disagree with using the curve for what is already a highly selective population. Now, after being on CC for a while, I can see that it has some merit. A lot of the debates here would never have occurred under this system. I still, however, prefer the British system, where they have government inspectors that ensure an A is an A, wherever the degree is granted. At least this is what I am told.</p>