@frazzled2thecore—well, like @ucbalumnus points out, the high-demand majors may have more visibility in this sort of thing than their actual number would merit (absent the high demand). I still stand by my earlier claim that for the vast majority of majors (and, in fact, the vast majority of students), grades required on prerequisites are going to be C or C- (if, in fact, anything beyond the default of a simple passing grade is specified).
We’re getting off-topic here, but -
Perhaps my data points are limited, but while it is true that there are plenty of majors well-represented among graduates in which a C or C- in prerequisites will do the trick, it seems to me that at some of these schools large numbers of students enter with every intention of pursuing a high-demand major (engineering, business, sometimes education) or applying to medical school and end up leaving with something else in order to preserve their GPA or keep a merit scholarship, and that is why there is more visibility.
Frazzled D changed majors a number of times, and that is how we became aware of these data points.
Outside of engineering, at some schools there are minimum GPA requirements for foreign language programs and other programs requiring a sequence of performance-based classes, but without a curve. (I took a brief glance at some of the flagship programs for studying critical languages.) It is something to consider if a student is invested in completing a specific major or sequence of classes and not just graduating with a high GPA.
This might matter a lot to some students and their parents. As @mathmom has pointed out, her S began as a “C” student in Arabic but graduated as an “A” student, and it obviously mattered to him that Tufts did not require a B grade for him to study Arabic in his study abroad program and then continue on to upper levels.
Well the Arab teachers at Tufts are notorious for giving out European style grades. The C did keep him out of some of the more selective language programs. Tufts doesn’t have their own Arabic language programs overseas, which had its plusses and minuses.
Note that the numbers above are for “grades given to students of identical quality.” When comparing grades between different majors, you sometimes have issues with the higher academic achieving students being more likely to pursue engineering and less likely to pursue certain other majors. To avoid this effect, the numbers above controlled for a variety of factors, such as measures of academic quality of the student, and then looked whether similar academic achieving students received different grades in different majors. They found that the engineering as whole appeared to have slightly more harsh grading than average, with notable variation between different departments. The typical pre-med majors of chem and bio were the harshest grading majors.
The standard, uncontrolled average GPA for different engineering majors at Columbia in 2006-7 are below:
Electrical – 3.30
Mechanical – 3.34
Chemical – 3.38
Civil – 3.45
Biomedical – 3.50
Environmental – 3.54
There has probably been about a 0.1 GPA increase since 2006-7 for grade inflation, so the overall engineering average would probably be near 3.5 today. This do not sound low to me, although it would probably be lower than the overall Columbia average. Highly selective private colleges often give A’s to most of the class and have high resulting average GPAs. For example, in Harvard’s senior survey, the average self reported GPA was 3.64. Whether this is good or bad depends on whether you consider grades as primarily a measure of how well students have mastered the material or a measure of distinguishing students from one another.
“TA confession: I’m sorry, but most of your children (my students) are average”
It depends on the college or secondary school. Exeter’s “average” runs circles around 99% of their peers. Georgetown U’s “average” runs circles around 99% of Maryland peers.
Sure, “average” varies by college. However, the higher the “average,” the more likely students are to leave STEM majors.
See this paper: http://sole-jole.org/13144.pdf
It makes sense, if you consider that we are sensitive to our standing relative to others. Students leaving STEM majors at Harvard, for example, might have been total stars at colleges with lower average SAT math scores. Had they attended a college with less competition in the major, they might have graduated with a STEM degree.
To relate it to the first post in this thread, at another institution, a student complaining about a B- might have received an A+. The B- is not a bad grade, but it signals the student’s standing in the subject, relative to their peers. If you share a curve with the next Stephen Hawking, you know where you stand.
Jeff Bezos on physics at Princeton:
http://www.achievement.org/autodoc/page/bez0int-2
I think, if you’re used to being the best at something, it can be a shock to suddenly not be the best. However, things didn’t end badly for Jeff Bezos, even though he’s not a physicist.
Actually the study suggests that a particular student attending Harvard would have an increased chance of graduating with a STEM (actually science) degree than a student compared to a student attending a college with less stellar students, rather than decreased since Harvard has so high CR scores. They found that peers having a high verbal score had a larger positive effect on sticking with the science degree than did the peer high math score’s negative effect. For example, if college A had mean SAT of 600 math / 600 CR, and college B had mean SAT of 700/ math / 700 CR, then a particular student would be more likely to stick with the science major at college B because the peer’s +100 CR has a larger positive effect on sticking with the science major than the peer’s +100 math’s negative effect. I expect the driving force of this relationship was not the score itself, but rather other characteristics that are correlated with score, including things like the degree HS academic prep in related subjects.
Note that the grade distribution is usually highly correlated with how exceptional the student body is, not just a reflection of how you compare to your peers. You brought up Harvard earlier. In Harvard’s senior survey, their mean GPA was 3.64… roughly an A-. Several years ago, Yale said ~2/3 of their grades were A- or better. It’s probably quite a bit higher today. Stanford has a history of even higher grades that HY. In a typical class at such colleges, few students get grades as low as B-. A B- often indicates you are at the bottom of the class. In contrast, at some less selective colleges, most of the class gets grades below B- and few get A’s. A B- may indicate you are slightly above average in the class. I wouldn’t assume that if you can get a B- at Harvard, you’ll have no problem getting A+ grades at a less selective public.
Princeton has less grade inflation than the rest of them. It’s not a secret on CC.
There doesn’t seem to be a whole lot of grade inflation at University of Chicago, either. GranitestareSon in enrolled in honors analysis this year, a class that students must apply to take. Most students are delighted to earn a B for a grade in any particular quarter. On the other hand, students who elect to take this class know what they are getting themselves into, so there shouldn’t be too much whining about grades.
Chicago has a smaller portion of students with A grades than HYS, but grades are still far higher than at typical less selective colleges. The page at https://canigraduate.uchicago.edu/statistics.php shows the grade distribution at Chicgao a few years ago. Grades should be a bit higher today. About 60% of grades were A or A- overall. One of the language departments had 100% A/A- grades. Math was one of the harshest graded departments, with ~50% A/A-'s, which fits with the Quora comment at https://www.quora.com/Are-University-of-Chicagos-honors-calculus-and-honors-analysis-courses-graded-on-a-curve mentioning up to half get A/A- grades in Honors Analysis. Similar comments could be made for Princeton. Even at their most severe past grade deflation, a far larger portion of students received A’s at Ptinceon than at typical less selective colleges.
According to Quora, Princeton gives about 35% As. So MIT is harder than Princeton and Princeton is harder than the rest. Stanford probably is the easiest of the bunch according to some of the posters in the above link.
In the past, when Princeton had grade deflation, the target was 35% A’s in standard courses and 55% A’s in independent work. While the official target was 35%, Princeton never reached that target. Most departments gave >40% A’s. A couple gave >60% A’s… The overall average during Princeton’s final year of grade deflation was 43% A’s. After Princeton grade deflation policies ended, these target A percentages were removed. And with this change, the A percentage has been increasing. It’s still a lower percentage than at HYS, but also a higher percentage than at typical less selective college.
I think the main reason grade deflation was unsuccessful was having an overall negative effect on both students and the college. In one Princeton report reviewing the policy, 1/3 of students said the policy negatively influenced their decision to come to Princeton, suggesting hindering number of applications, losing more cross admits, etc. And 60% of students, said the policy negatively influenced their experience at Princeton. Some students mentioned classmates refused to assist them or work with them, to increase the chance that they’d be one of the top ##% students that would get the A. I expect it also had negative repercussions on things like acceptance rate to selective grad programs. It’s difficult to have notably harsher grading than your peer schools.
@Data10 Thanks for the links. The grade distribution described on quora makes sense considering the group of students in honors analysis are initially self selected and then further narrowed down by the math department. That said, the syllabus for the class specifically states that there will be no curve. Not sure whether this means no curve in the sense of a predetermined distribution of grades, no grade adjustment in an exam with a mean of 50, or both.
FWIW, GSSon is a grader this year for a non honors section of calculus for mostly non math majors. Would be interesting to see what the grade distribution is like compared to the courses that majors generally take.
A few thoughts …
I think there is an issue of grade compression rather than inflation. A grade of B used to mean you performed better than most. Now it often means you performed worse than most.
Because there is so little difference in GPAs, colleges and the grad programs have to rely more on testing. They can use ACTs and GREs to determine if they are seeing an average 4.0 student or a really good 4.0 student. It’s not ideal, but it’s a sanity check on high schools and even university grading systems.
Because much of our society is unwilling to rank our students, the students who are hurt the most are the true top performers. If “good” gets an A, it’s hard to reward the person who did “superb”. It has to affect the motivation of top students.
I think the OP can read the last few pages and sort of see why current students focus so much on GPA.
I’ll bet big cyber dollars that the curve – median grades-- are lower for the non-honors section.
@Periwinkle The study you posted is another piece of the course selection/switching puzzle. In the Arcidiacono study, weaker students switch to protect their GPA. In the Wellesley study,
http://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.28.3.189
the same behaviour is noted. In your California study, URMs were not doing so and had to suffer the consequences. What gives?
My interpretation is that both the student body at Duke and Wellesley have a higher SES than those at the campuses of the University of California system. They know the game and they know how to protect themselves. Low SES students are not so well informed, unfortunately and they pay dearly for it.
More importantly, I feel the inconsistent grading practice across schools and majors is a serious problem. Perhaps a general exit exam like the CLA would solve the problem once and for all?
The story of Jeff Bezos is another interesting one. Here is an interesting quote:
Bruce Jones, a former Amazon supply chain vice president, describes leading a five-engineer team figuring out ways to make the movement of workers in fulfillment centers more efficient. The group spent nine months on the task, then presented their work to Bezos. “We had beautiful documents, and everyone was really prepared,” Jones says. Bezos read the paper, said, “You’re all wrong,” stood up, and started writing on the whiteboard.
“He had no background in control theory, no background in operating systems,” Jones says. “He only had minimum experience in the distribution centers and never spent weeks and months out on the line.” But Bezos laid out his argument on the whiteboard, and “every stinking thing he put down was correct and true,” Jones says. “It would be easier to stomach if we could prove he was wrong, but we couldn’t. That was a typical interaction with Jeff. He had this unbelievable ability to be incredibly intelligent about things he had nothing to do with, and he was totally ruthless about communicating it.”
There is smart, and then there is smart. I have heard the argument that scores beyond a certain range does not matter. They are wrong. It matters a lot, all the way across the spectrum.
SAT scores’ ceiling is too low to make a meaningful distinction between smart and super-genius, as well as not testing all of the various dimensions of smartness that could be of use. That the SAT is taken while in high school means that it does not reflect any post-high-school intellectual development (or regression) since high school.
Given the wide range of things college students study, there is rather little advanced (as opposed to frosh-level) subject matter across all students that one can cover in a test to get valid results across all college majors.
^^ My concern is not the cognitive elite, but simply the quantification of learning. I find it odd that we focus so much on the admission process but not the finished products. I find I am in completely agreement with Bhatnagar when he said:
The test is a “great leveler,”… It allows the company to be more fair in considering applicants who have graduated from different colleges, by relying on something other than an institution’s prestige.
…helps make the hiring process more meritocratic, by verifying what job-seekers know and can do. It’s a way of identifying talent, which is crucial for both companies and the millions of skilled yet underemployed workers…
https://www.insidehighered.com/news/2014/11/14/indian-companys-skills-test-college-graduates
If the SAT is too easy, how about the GRE and a subject test, if available? While I find the new products interesting, I doubt they will provide much more useful information than is available through the standardized tests we already have.
Interesting article. What I gain from it is that there are many testing companies salivating at the chance to open up the US “testing for employment” market. I suspect that Griggs vs. Duke Power Co. will stand in the way.
I am not convinced by the arguments put forward to justify the use of a college exit exam. I suspect the results of such an exam will align well enough with the SAT/ACT to render the exam superfluous. I would fear that institution of the exam would clutter up students’ schedules with required test-prep classes. As a parent, I would not want to pay for such classes. I would far rather pay for real college courses.
Do we really need to test the crap out of everybody? This constant need for assessment is driving K-12 education nuts. Employers know the reputation of various schools based on the folks they have hired in the past. If they thought it was needed for their success, they would be pushing for this.