“Of course, the other explanation is that employers don’t need 1000 clones - they need some people who are good at some things and some people who are good at other things. It matters not if the physics major is academically “smarter” than the psychology major if he can’t write a press release or devise a marketing campaign or come up with a better retention strategy for employees.”
@pizzagirl love your posts and your sarcasm. If you had a fan club I would run for president.
That was me. Of course it’s not a good idea for 99.9% of people. I said it was “very extreme”. It was a class especially designed for students who were very talented and passionate about the subject … Everyone knew what the expectations were going into the class; students took it because they genuinely loved the subject and enjoyed spending that much time on it. There were several other courses offered for people who weren’t that motivated/crazy. No one gave a crap about their grades. I think pretty much all the students my year got A-'s, A’s, and A+'s anyway.
I think a lot of the disagreement is in what our opinion is of dealing with professors who are bad at their job of teaching and fairly evaluating students. Some of the defense of a forced normal distribution is, “I don’t know how to make a fair exam so I will just hand out grades on a quota and say that that’s fair.” I think that if that’s a justification, then the prof in question needs to learn how to be a better evaluator of student performance. It’s not an easy skill to design fair questions for new topics but it is a skill that can be learned. Similarly, a history professor should know about what an A paper and a B paper looks like, within a margin of error. Perhaps profs should be required to take a handful of courses on how to teach students.
You’re never going to be perfect and even veteran teachers can make tests that are harder than they were meant to be. It’s fair to give people a score bump if that’s what happened. But if you write bad questions on a test that no one really understands and can answer well (like 40%=A exams) then I wonder what it is that you are actually testing because randomness and noise always creeps into tests of that design.
This discussion has provided reasons I had never considered for “where average grade is failing.” Some perspectives make sense. But to the student, it can be a different story
My daughter is living this right now. And she just feels disheartened. She is a graduate student in a top engineering program that I expect only accepts top students. So likely no slouches there. Yet average exam results can be quite low.
Her test scores of 40% earn Bs, which she I suppose she should be happy about. But she doesn’t want just good grades, she wants to have mastery of the subject. And when you can’t even correctly answer half the questions on an exam, do you have mastery? She worries about sufficiently learned the subject in a manner she will be able to use in a future job.
Re marlene, #343–many times students cannot solve a problem on an exam, within the time constraints of the exam, but they are perfectly capable of solving it in an untimed framework. Sometimes they have forgotten one equation or concept that they can quickly locate in a book or in their notes. If your daughter can solve the problems retrospectively, then she is doing fine. She might talk with the professor, who may be able to look at her exams and spot categories of problems that are giving her issues. How are the A students doing on the exams? If they are scoring in the high 90’s, she should talk with them. If they are scoring in the high 50’s, she should not worry about her percentages.
Going back to Canuckguy’s post #526, I think it is best to look at the actual transcript, rather than relying on the major. Students who major in my department have about half of their courses in common, over 4 years. Maybe it is only 1/3 of their courses that are in common. What the students elect to take with their other available credits is a really good indication of where they are intellectually, and how committed they are. Engineering tends to be more constrained, with fewer electives, so the engineering majors have quite a few more courses in common. But that can sometimes put a limit on exploration on the high end, in coursework.
There are a number of different curving strategies. Personally, I would never use a forced distribution unless ordered to do so. Perhaps in the case of a very large course section, or one in which we were supposed to forcibly weed people out. I have used a “High Score to 100%” approach to the overall course grade. It does reward better students with more points, but I think it is more fair than just adding X points to everyone or significantly changing the thresholds. If things are not too bad, I just adjust thresholds in some manner that seems reasonable.
Academia tends to be a self propagating system where the people who get academic jobs are most like their professors (the result of personal references being the core of what gives you a chance to advance). This tends to mean hyper competitive (even if not especially talented), hard working (even if not especially efficient), and encouraged rather than discouraged by being treated poorly with low grades and generally unacommodating professors. People who are not like this tend to drop out, and there are two unfortunate trends that you could almost instantly observe:
By attrition, you are left with the kind of people who propagate the previously mentioned qualities and so the classes are designed for the benefit of people who are similar to them.
As far as understanding of the material goes, I wouldn’t worry about it. Those tests may measure aptitude poorly but if she learned the material to a level that would let her do well on a standard scale, she will be fine in a real position. Profs just sometimes put material that can be solved using what is learned in class but usually requires a more advanced level of intuition (I.e. A few classes further into the major) to recognize the means of the solution, and they call it fair.
Here is an example of a major where two students in the same major may have only one course (out of twelve required for the major) in common:
http://guide.berkeley.edu/undergraduate/degree-programs/history/#majorrequirementstext
Engineering is very standardized and the presence of the ABET organization usually means that you should have a pretty good idea of what a fresh graduate is capable of, with some bounded margin of error. They might have done work beyond their requirements, but not necessarily all that much if the program is really tight on requirements.
Most other majors tend to be highly self-serve in nature and the degree means less than how you went about obtaining it - what classes you took, what specialist knowledge you developed. For example, a mathematician that specialized in statistics is much different than one who specialized in abstract algebra. That makes an evaluation of the transcript pretty important.
I haven’t read through this 20+ page thread, so I may be repeating earlier comments. I have several engineering-related degrees, including in EE. I’ve taken EE classes at both a state school (UCSD) and private (Stanford), as well as non-EE STEM classes at several other colleges, including much less selective ones. My experience was that grading on the curve is fairly common, particularly for larger, lower level classes. A minority of professors will give out challenging enough tests for the exam average to be below 50%, and curve the exam such that the class grading is reasonable. I don’t think there is anything wrong with this.
However, with modern grade inflation, curving an exam to make the average class grade a C is not at all common at selective colleges. At Stanford and various other highly selective private colleges, exams may be curved to as high as A- in a class full of students who do exceptional quality work in the class, such as an accelerated version of an intro class or an upper level engineering class, where nearly everyone is a major. I’d also expect that at most less selective colleges, it is more common to curve exams to a grade in the B range than the C range, although there are numerous exceptions. Along the same lines, you’ll generally see a much higher withdrawal rate in intro classes and a much lower withdrawal rate in upper level classes. If you want grading information about a specific college, many do publish information about grade distribution on exams or as overall class grades. For example, the first university than came it a web search was University of Wisconsin at https://registrar.wisc.edu/documents/Stats_distribs_2015-2106Fall.pdf . It looks like the mean grade for EE classes is in the B+ range, with maybe an average of 10-15% getting a grade of C or below. Even at more harshly grading colleges, I haven’t heard of that rate of withdrawal being common.
I think the reasons for engineering majors often taking >4 years to graduate more relate to engineering majors often requiring a much larger numbers of specific courses than other majors with ABET requirements, some students being unprepared in math/science during HS and needing to start out behind in near remedial classes, engineering courses often requiring specific long sequences so if you miss one you fall behind, and engineering majors being more likely to do coops or other work experience that takes time from classes. At some colleges, limited registration can be an issue.
I think a lot of the disagreement is coming from a misunderstanding of how grades are being “curved.”
Most of the time we are not talking about curve fitting the grades to a specific (usually normal) distribution (aka bell curve), where 16% get A’s 34% get B’s 34% get C’s 14% get D’s and 2% get F’s (or some other predetermined variant where the top x% get A’s and everyone else is out of luck.
The “curves” I am familiar with in classes with tests averages in the “failing” range is handled differently. What happens is that everyone gets a bump of x points to shift the curve of the class towards 100%. In essence, it is the equivalent of adding a handful of questions that everyone answers correctly. There is nothing about this method that restricts the number of A’s that are awarded.
People on this thread seem to be asking for professors to make half the test fairly easy questions so that nearly all the students get full credit on the first 50 points, and then make the other half of the test harder questions where many students may only be able to answer half of them. But why not just cut to the chase and assume that the students know the basic stuff (if not this will be very evident when they try to tackle the harder problems) and give them the full time to work on the harder problems? That gives them a better chance to show how well they do with the harder material. The prof can then “curve” the grades to the beloved high school standard by adding in those 50 points and give them for signing the name rather than by completing the plug and chug problems at a basic level.
@mathyone What you are suggesting is exactly what happens in all of the cases of test averages in the “failing” range that I’ve seen.
Weeding out students who are passing because your college over enrolled a particular major speaks volumes, but not about the children. If they only want x number of engineering majors they shouldn’t accept x + 60% because other majors are under enrolled. A family who would pay $65k/year for a particular college if their child is enrolling in engineering may not be so willing to pay that for a different major.
The alternative is to admit frosh directly to majors, but schools (or engineering divisions) which do this can be very difficult to change majors at. Some schools do a hybrid, admitting some students directly to majors, while giving others general admission to compete in a sometimes-brutal weed-out process for the remaining space in the popular majors (e.g. UIUC, Washington).
The reason many schools have to make this choice is that they cannot afford to maintain reserve capacity in every major to accommodate yearly changes in demand. Of course, this is not unique to engineering majors. Even a school with an enormous endowment may have a capacity-restricted major (visual and environmental studies at Harvard).
Of course. VES at Harvard. Good example.
The only course I was involved with that I knew was intentionally “weeding” people out was an intro physics course mostly populated by (~370) pharmacy school hopefuls. The exams were so difficult that even we T A’s had a hard time getting them right. This was a state flagship, not a private school, and they could only accept so many pharmacy students. It was an ugly process and the students hated us, but somehow we got tasked with this.
But isn’t it better to let everyone try and see who washes out than to determine that a priori somehow?
No. University is a commitment of at least 4 years and uncertain quantities of money. The effective opportunities for transfer are limited. If you aren’t pre-screening a priori for who really belongs, then weeding people out is akin to a very costly bait-and-switch.
Of course, if demand far outstrips the space available in the major, it does not require the courses to be extremely difficult to have a “weeding” effect. If the admission to the major is highly competitive, then even normally-respectable grades like B or B+ (regardless of whether grading is done on an absolute scale or on a curve) could put the student on the path to be “weeded out”.
Some students see it differently, in that it can be more costly to be “weeded out” after spending two years at a college and have to transfer to another college to find your major (if there are any other colleges where you can transfer into your major). Particularly if one had earned normally-respectable grades like B or B+ that are more than sufficient to stay in good academic standing and succeed in the major, if one were already in the major.
I disagree. One alternative could be taking a reasonable number that your resources could support knowing you’ll lose some along the way as interests change, just like every other major.
Another alternative is to post prominently on your website that 60 of every 100 students will be weeded out, no matter what grades they earn. Tell families that students won’t be graded based on what they know, they’ll be graded based on a comparison with other students to create a predetermined ranking with only a specific, limited number permitted to earn each grade. Then take your chances and see how many students decide to enroll in your school.
In broadcasting we call these type of tactics bait and switch. Our industry considers them unethical.