Harvey Mudd reports just 58% of classes have under 20 students while only 4% have 50 or more students.
Dartmouth College is at 65% and 7%.
Back to the faculty:student ratio. Most schools game that system by counting graduate TAs and part-time adjunct faculty as faculty. In a research university, there may be as many TAs as there are faculty with terminal degrees, and many departments, like English departments, have most of their intro classes taught by temporary faculty, who are hired on a semester by semester basis, for a class or two. These adjuncts are purposefully kept <1/2 time, so that the university won’t be required to provide healthcare. That of course further increases the official number of faculty, further reducing the student:faculty ratio.
Which brings us back to the class size. While most classes may be “small”, those classes are mostly taught by inexperienced graduate students or by faculty who have few resources, often lack any time of office, are underpaid, have zero job security, and have rarely taught the class more than twice. If they actually have a tenure track or tenured professor before their Junior year, that is for a large course, in which the main point of contact is still the TA.
So while a college will advertise small class sizes and large faulty:student ratios, these are often not classes with tenured, world famous professors teaching a personalized intimate course for 15 students in a small classroom. It’s TA’s in lab and discussion sessions for huge courses, or smaller intro classes with a none tenured lecturer, the lower caste of FT faculty, or an adjunct who is likely either teaching in another college in the next hour, or waiting tables that evening, who does not have an office, has not taught this class for very long (or is teaching for the first time), and is likely teaching the most bland and non-challenging version of the class, because they have no job security, and can be fired at the whim of any administrator if a student complains about anything, from their grade, to the instructor’s dress style. So yes, most of those colleges are gaming the system.
This is rule for large public universities, no matter how elite, and for many private schools, especially those with graduate schools and those in large urban areas with large pools of unemployed or underemployed PhDs and MAs.
Compared to the Claremont colleges, Mudd has a different distribution of class sizes. Specific numbers are below. If you look at classes with less than 20 students, CMC is first and Mudd is last. However, if you instead look at classes with less than 10 students, Mudd is a distant first and CMC is last. Some students at CMC claim this relates to capping/targeting class size at 19. I expect the larger math/science/tech enrollment at Mudd also plays a role.
CMC-- 84% under 20 – 76% of classes have 10-19, 8% have <10
Scripps-- 80% under 20 – 63% of classes have 10-19, 17% have <10
Pitzer – 71% under 20 – 56% of classes have 10-19, 15% have <10
Pomona – 71% under 20 – 53% of classes have 10-19, 18% have <10
Mudd – 58% under 20 – 26% of classes have 10-19, 32% have <10
Deep Springs waves hi
@Data10 - where do you find the data for class size <10? I’d love to see more data like that.
Class size <10 and class sizes of various other ranges are listed in the CDS.
If they are grading on a curve, I never thought smaller was better. A couple of naturals show up and there goes your “a”.
@privatebanker If somebody is trying to grade on a curve in a class that is smaller than 10 students, or even 20, the curving is invalid. Curving a class assumes that the deviation of the mean grade of the class from the expected mean of the school is the result of the difficulty of the tests or the harshness of the grading. This assumes that the class is a representative sample of the entire student population. However, small samples are rarely representative of the population, and sampling error is a much more likely explanation, as exemplified by your scenario.
Of course this also assumes that the grade distribution in school is normal, which is definitely not the case. Due to grade inflation, grade distributions tend to be heavily left-skewed.
Very small college classes are rarely graded on a curve. Instead curves are more common in large freshman intro type classes. Of course, there are some exceptions to this generalization. For example, several years ago Princeton tried to implement a grade deflation policy in which no more than 35% of students received A’s. The policy didn’t go well as has since been abandoned. Most departments averaged far above 35% A’s, and among the professors who did try to follow the policy, some applied the 35% A rule to small classes. One of the student comments in the policy review report at https://www.princeton.edu/sites/default/files/documents/2017/05/PU_Grading_Policy_Report_2014_Aug.pdf reads, “The grading policy is particularly unreasonable in introductory language courses. On the first day of classes, my [language] teacher said that only 3 of us in a class of 11 would receive As. This often means that despite receiving an overall grade of 90+ a student cannot receive an A-grade because some other student got a 91 or 92.”
Regarding grade inflation. It sometimes isn’t grade inflation. In grad school, I TAed for the intro Bio course for premed kids (I know, there is no pre-med major, but they all wanted to go to Med school). The A cutoff was 85/100, and about 40% of the class would get As regularly. So they decided that 85/100 was too low, and raised it to 90/100, without otherwise changing anything else about the class. The 40% As remained - kids just put in more effort.
I’m sure that this would be a lot less likely in more difficult classes, like an intro Algorithm class in CS, but it does indicate that sometimes an increase in As indicates that there is an increase in number of hard working students and in the work that students are willing to put into classes. The Baby Boomer narrative of Kids Today Are So Lazy Compared To Us has so permeated our consciousness that we are rarely willing to even consider any argument which is based on Millennials or Gen-Z kids being more hard working than their parents.
So TAs are excluded entirely, and the assumption is that an adjunct faculty is teaching only a 1/3 load (which at many schools would be 1 class per semester).
^^^Although one can glean some good information from the CDS reports, remember that filling this out is optional… there are few guidelines, no oversight, and no consequence/penalty for reporting inaccurate information (whether by honest mistake, or not)
On the Internet, everyone is a dog.
What is kinda wierd about these numbers. Stanford and UChicago have roughly the same size undergraduate (Stanford at 7k, UChicago at 6.3k) enrollment. The same(ish) student to faculty ratio, but UChicago is at 80% of classes less than 20 and Stanford is at 68%. I’m not sure that tells me anything other than the numbers aren’t intuitively making sense.
Before someone jumps and says “They’re gaming the system,” I’m not sure that is the answer. I also am skeptical 19 student classes are gamed. What if they were capped at 18, to not give the appearance of gaming the rankings, but in fact game the system? How does one organically arrive at class size? I really believe that CC cares more about college rankings than schools do. But, if higher rankings lead to larger pools of candidates which in turn lead to more selectivity to allow schools to get better classes, I say game away! If it works, instead of complaining about, do it too. All this whining to me sounds like people complaining about the Yankees or Red Sox always winning because they spend money on players. Nothing is stopping you from doing it too.
@BrianBoiler In fact, the number don’t really mean anything. In fact, the number of “small classes” doesn’t mean anything, nor does the faculty:student ratio. These are only important because USANews includes them in their calculations of college rankings because their people decided that these are indicators of some measure of “quality”.
High faculty:student ratios and number of percent of small classes are, supposedly, indicators of how much personalized attention a student gets in a university, which is, in turn, assumed to have a positive effect on student success. I have been unable to find any study which actually demonstrated that this is true at the college level.
However, what these factors do demonstrate is how wealthy a college is. Colleges which do not have the funding to actually hire a very large number of faculty, and provide the facilities for teaching many small classes will, of course, do their best to game the system. I do not fault them for gaming the system, I fault that so many people are invested in such a faulty system.
Just more evidence that USANews and other ranking systems are using a poorly measured indicators, which have never been validated, to measure and rank the quality of colleges. Business as usual, nothing to see here, people.
Hmmmmmmmmm… class sizes are the basis for public education funding K-12, so you are invalidating the entire K-12 funding system?? You need to dig a little deeper to find the many links between class size and quality of education, there are a lot out there.
@CU123 I was pretty specific, I think, because there is research on the effects of class size in elementary schools. However, since college classes are taught is a profoundly different way than K-12 classes, it is not valid to extrapolate from one to the other.
You mentioned Stanford in your post. At Stanford, classes are rarely capped . Instead class size increases to however many students choose to enroll. There have been a few cases of ~1000 students enrolling in a class in the past, such as when Zuckerberg lectured in intro CS. The format is generally all students who enroll have a single lecture class led by a well regarded professor in the field. The class also has smaller sections led by a TA where students can more easily ask questions, ask about exam prep, etc. Sections are usually reasonable size, often under 10 students. You can look up the sizes of specific classes at https://explorecourses.stanford.edu/ . Some examples are below. I am listing enrollment for a specific quarter, rather than sum of the year.
First Year Tibetian – 2 students
Intermediate Tibetian – 0 students enrolled
Aerial Robot Design (AE) – 5 students
StoryCraft: Athlete Relationships – 19 students, split in to 2 sections
Longevity (Psych) – 50 students, split in to 7 sections
Lowest Level Freshman Physics – 119 students, split in to 13 sections
Intermediate Level Freshman Physics – 483 students, split in to 35 sections
Most Popular Intro CS Class – 427 students, split in to 70 sections
@MWolf the same principles apply to education across the board. Again dive a little deeper and you will find that smaller class size is ALWAYS beneficial to education.
Not if you are the student who cannot get into the desired class because it is capped at 19 and you are the 20th student who wants it. Or if the class is suitable for large lecture format and is taught by the best instructor, but only 19 out of the 300 students who want to take it are allowed in.