"Why Can’t Everyone Get A’s? Excellence is not a zero sum game." NYT op-ed

Parent Cafe
41

An instructor can use non-curve grading, but set thresholds that are not necessarily the same as the K-12 scale of 90% = A, 80% = B, 70% = C, 60% = D (or similar). The typical K-12 scale encourages instructors (in K-12) to load up the graded tests and assignments with easy problems, with only a few hard ones to distinguish between different levels of passing. College tests and assignments may have different percentages of easy, medium, and hard problems, so that the typical K-12 grading scale may not be appropriate for them.

For example, if only a third of a college test’s problems were easier ones suitable for C students, then a suitable non-curve grading scale would have one third correct being in the C range.

42

In some of D’s engineering courses the exam material is all about applying what they’ve learned in class in a novel way. Hard to predict the mean. Sometimes the profs nail it, other times, not so much. It all works out in the end and FWIW, the hardest courses with the lows means were the ones where she learned the most.

43

There can be different reasons for such issues. I had several very problematic courses at the graduate level. I once had a 4% on an exam. Of course, it resulted in my having to retake the course. I often have courses where the average is in the 50’s, but there are generally those who manage to be well above the average, especially in a class of 24+. If there are those who can ace the exam, then I don’t see it as a problem with the course as opposed to the students themselves.

44

@sylvana8798 just because a couple of students can ace the exam doesn’t mean that there isn’t a problem. What if those 2 students are beyond exceptional? Should the rest of the class be penalized in comparison? What if there are 2 sections of a course and one section contains a couple of extraordinary students. Should the a student in that class who scores the same as a student in the other section receive a lower grade even though the work was comparable simply because they were being compared to the exceptional students but the students in the other class were not? If that were the case you would have to score both class sections as a whole and as far as I know that doesn’t happen.

45

^^^Maybe, but if only a handful kids (who may have special knowledge or are exceptionally bright) score well above a 60, then something is not right. Either the professor is not teaching the material in a way that is understandable to most of the students, the test does not reflect what was taught, the students in the class are not properly prepared, or it is a class full of slackers. Yes, there need to be standards and accountability, but the goal is to impart the necessary knowledge. If most of the class isn’t getting it, that is not effective teaching.

Sometimes teachers make the first test really hard so students know that they need to up their game and focus on the class.

46

Or the instructor may not be the best at calibrating the difficulty of the test problems. An instructor who is immersed in the subject day to day may not have the same view of what is “easy” and what is “hard” as a student learning the material for the first time.

47

Did not read the article- did not try to log in et al. But- the thread title about A’s got me. A’s would be meaningless if everyone got them, all of the time. A’s are to connote superior work or mastering all of the material presented. Getting the degree, license, whatever means enough mastery. I can see honors classes weighing heavily to the high grades as well.

I remember our honors freshman chemistry course. It was wonderful, the high score on the blue book exam might be an 87 for one or two with the next top scores in the 60’s- also an A. Wonderful because our brains were stretched doing those exam problems. A lot was taught but we also realized how much more there was to learn. It would have been disappointing if the tests were easier. I also recall not feeling I had mastered the material in any college class I did not get an A in. There was more to have learned.

You will note people get licenses and degrees without grades. Letter grades are ambiguous enough, basically only four (F is failure) showing better/worse performance. So much better than a 100 scale.

A note on excellence not being a zero sum game. Huh? By definition excellence is above average, otherwise it is a meaningless term. Sorry, but we are not, nor never will be, equal in so many parameters one can measure.

48

Zero sum game is best represented by situations with benefits accruing disproportionally one versus another.

Options contracts. Auction bidding. Two candidates for one job. That sort of thing.

An A in Chem and a B in Chem bears no resemblance. Neither wins or loses in the process.

49

“When I hear about classes where the highest grade is a 60 there is something wrong with the class not the students. Either the material is not being taught or the tests are not reflective of what is being taught or the expectation of what can be done in the limited test taking time is unreasonable.”

US students are too used to being able to get 90%+ on everything. But I don’t see anything wrong with the mean being in the 40s for a long tail test if the pass mark is set appropriately. For example, the AMC-12:

“The average score is typically around 65 out of 150.21 About 6.5% of students taking the AMC 12 scored 100 or better.”

The Cambridge undergrad math exams are similar but harder, the mean is about 9 out of 28 questions correct (total on four 3 hour exams over 2 days), with the top third getting 50%+ correct (the pass mark is about 20%). But there are perfect scores every year, because the aim is to find the best mathematicians in the world. You can understand the material, but still not be able to make the intellectual leaps needed to find the complete answer to a question.

50

@ucbalumnus What you are describing in post #40 is one why that many instructors curve.

51

However, #40 describes the possibility of a non-curve grading scale that does not have the 90/80/70/60 percentages that K-12 in the US typically has. For example, the course whose information is at https://sp19.datastructur.es/about.html#grades has pre-set non-curve grading thresholds of 86% = A-, 68% = B-, 47% = C-, and 25% = D-.

52

This is the Lake Wobegon philosophy of education, where every student is above average.

53

“When I hear about classes where the highest grade is a 60 there is something wrong with the class not the students. Either the material is not being taught or the tests are not reflective of what is being taught or the expectation of what can be done in the limited test taking time is unreasonable.”

Most STEM programs have weed out courses, especially engineering where as others have mentioned, you want the person designing the bridge to do well in mechanics or working on circuits to do well in electricity and magnetism, two typical weed out classes. Organic chemistry is another one discussed on cc for pre-med, that course is probably the first course med schools look at, and most colleges do not hand out many Bs and As there. There are a few theory courses that serve similar purposes in math and computer science. You need that kind of rigor in these disciplines.

54

This argument almost seems like a thought experiment to me: if smart students should get A’s, and all the students in the class are smart, should they all get A’s? I agree with both sides to a point, but the reality is, not all students, even at Harvard, are going to be equally smart or knowledgeable all of the time, so there is a limit somewhere – not necessarily an arbitrary limit (eg, 10%), but a natural limit (eg, 25/80 students have the capability to get an A).

What about the system used to score AP tests? Because there are specific numeric cuttoffs for each score, the percentage of students getting 5’s varies each year based on students’ performance and preparation, yet there is always a mixture (usually uneven across each score, year, and test) of 1-5’s. I’m not sure how the system works, though – I think they “norm” the test by giving it to college students taking an equivalent class, so the mileage may vary in an actual classroom. Still, I think it’s a good way to mix the flexibility of a benchmark with the rigor of a curve.

I don’t think it’s always needed, though; if every student who tries gets an A in gym or art 101 or underwater basket weaving, that’s a win-win in my book.

Also, I don’t agree with the idea of adjusting grades based on effort. If part of the grade’s based on participation or easy points, that’s fine (obviously, as a student, I love that), but I hate when people imply that “working hard,” which is often a very subjective term, is a substitute for actual achievement. Maybe it’s because I’ve heard that a lot, but I’m sorry – if I studied two hours for the ACT and someone else studied for two hundred hours, my 36 is still earning me way more merit money than their 25. It’s not pleasant, but it’s the truth. My sister complains that I spend less time on homework than her and still get better grades. I admire how hard she works, but what am I supposed to do, assign myself more homework?

55

I am fine with a weed out class or a class that requires what you call rigor. But in those classes there should be students who earn As & Bs without a ridiculous curve. In the example given, nobody got higher than a 60 on the test - nobody was able to “pass” (considering 65 a passing grade) I still argue that something is wrong in that example.

My daughter’s college calc 3 class kept on shrinking all term long with lots of students dropping out. However, she earned an A by actually getting an above 90 average on the given tests/assignments. So, still a weed out class, but one where it is possible to do well.

Some posters mentioned inconsistencies across sections of the same class. In my daughter’s college, many of the introductory classes (where there were multiple sections taught but several different teachers) avoided this problem by having all homework, tests, and the final/midterm be the same across the sections. They were all graded and curved the same to be fair to all students. Higher level classes typically don’t have this issue as their are usually only 1-2 sections taught by the same teacher.

56

You are just used to a US-style high school grading system (A = 90+, etc.) and therefore you think that is the right model. It’s actually rather flawed in that it compresses the high achievers in a narrow range and therefore becomes more about perfectionism rather than demonstrating knowledge.

I think that the best exam model would be one where the median grade is 50 out of 100, where passing is 25 and anything above 75 is exceptional. Some of my daughter’s college courses ended up like this, and while it makes for people thinking that the exam is hard, it truly separates the strong students from the weaker ones.

57

If the students are not properly prepared or are slackers, it is not the teaching that is the problem, but the students themselves. You can only dumb it down so much.

Should the test problems be “easy”?

That’s where 20 years of experience comes in. I don’t generally “curve” exams or grades. Everyone is pitted against the material. If everyone were to get 90% of the course points, they would all get A’s. If I have two sections, all the exams are graded together and everyone just gets whatever points they get regardless of section. I don’t put letter grades on exams, so it’s not like a 90% in one section got an A, while a 75% did in another section. Faculty teaching different sections don’t coordinate our exams in my current school, so what other professors do is their problem. Students may do better or worse in a different section.

When I give an exam where no one scores well, I can evaluate it to see if it was too long or the problems were more difficult than the time allotted for and make some adjustment. That has been rare in the courses I’ve taught over the years.

58

Nobody is saying the course should be dumbed down, but pointing out (as you acknowledge) that if most students in a class don’t do well, the professor needs to evaluate the reasons why. That is what a good teacher would do - was the test too long, is there one part where a majority of students struggled, was the wording ambiguous, or just the question too hard for the time allotted. If students are allowed to register for the class who are not prepared, the college needs to fix that. You are also right that a professor should not be required to lower standards to such an extent that the course goals aren’t met even if the students all do well. I remember my kids taking about AP Psych saying one teacher was harder and the other much easier. The kids in the harder section got worse grades, but did better on the AP exam.

59

^For many of the “reasons why” there is nothing the professor can do about it. Knowing the why is one thing, fixing it is something else entirely.

60

Isn’t it common to write a test with some “easy” problems (for C students), some “medium” problems (for B students), and some “hard” problems (for A students)?

Of course, if a fixed grading scale using K-12 thresholds (90, 80, 70, 60) is used, then the majority of the test needs to be “easy” enough for C (and D) students to be able to solve.