got it - my bad - i might be thinking of someone else.
That’s a separate, practical matter. Doesn’t mean that the prof directly teaching each small section isn’t better.
Would a law of diminishing return help when considering class sizes? That is, instead of looking at percentage of classes not exceeding N students which can be easily gamed, compute for example 1/n with n being the size of a class and look at the average of 1/n over all classes, where the higher this average the better. This method disincentivizes capping class sizes at N, still rewards private schools, and does not exceedingly penalize public schools. (1/n is just an example — I’m sure there is a better function and a better way to combine the 1/n’s than simple averaging.)
Yup something like that could be a way to incentivize it without the ability to so easily game it. Nice idea 
Ya, things have changed because they’ve cut enrollment significantly. And they also split 61A into two separate classes - one with 1000 and another with 400. The course staff is very large with more than 35 GSIs/TAs.
Outside of 61A, only 61B gets close to that number. Generally the #s have been trending down because CS enrollment has been cut to 1/4th and the major is now direct admit.
2 Likes
Practical matters do have an impact. If you have 10 great professors, would it be better to use their time to teach 10 individual sections of an introductory course, or instead have one teach a large introductory course and the other 9 teach upper level courses?
1 Like
In addition to 61A, this class has 1,264 enrollment listed for Fall 2023. Your statement that “There is literally only 1 class at Berkeley that has that many students [1,000+]” is literally wrong.
https://classes.berkeley.edu/content/2023-fall-data-c8-001-lec-001
2 Likes
Except I wasn’t approaching it as a practical question, simply as a comparative one.
As to your current question, what would be better still would be hiring 20 professors. Or whatever.
Ok, you have 20 great professors. How would you allocate them?
You seem defensive about this. Why? I am stating a simple preference, which is that a great prof teaching each and every small section is better than the alternative. That’s it. If you feel otherwise, have at it.
Moving on.
1 Like
Truly, not defensive at all! I appreciate all kinds of schools. For example, my son also liked Cal Poly which has the model of “professors teaching each section.”
Just wondering how you would allocate the N professors. It’s not easy to find and hire great professors, especially in CS, etc. The tradeoff is that if you allocate them all to small sections, you don’t have as much depth and breadth in the higher end classes.
1 Like
Sure, I am happy to accept that although I was primarily talking in the context of CS. It doesn’t change anything with respect to the efficiency and value of these classes + the advantages accrued by critical mass.
More relevant to your example, Data Science as a field was pioneered at Berkeley and most other schools borrow liberally from the Cal curriculum. I would rather be in that large intro level data science class at Berkeley than at some small school where 1 professor is teaching my small class about data science based on curriculum and materials borrowed from Berkeley.
I previously posted the weightings that had the largest change over last year. Some of the changes that I’d expect to favor publics include:
- Class Size weighting decreased from 8% to 0% – An 8% difference in weighting is quite significant, by far the largest single component weightings change this year. Publics tend to have larger class sizes than privates, so removing class size is more likely to favor publics than privates.
- Alumni Giving decreased from 3% to 0% – Selective publics tend to score lower on alumni giving than selective privates, so removing alumni giving is more likely to favor publics. Berkeley was removed from USNWR a few years ago by lying about their low alumni giving – claiming 12% when actual was 8%. In contrast some selective privates average close to 50%.
- 2 First Gen Variables Added at 2.5% – Publics tend to have a larger portion first gen than privates.
- Test Scores weighting decreased from 5% to 0%, if fewer than 50% submit scores – Some publics, such as UCs switched to test blind. Others do not meet the 50% threshold. Test scores are more likely to be a relative weak point in the rankings criteria for selective publics than selective privates.
4 Likes
As a fellow NJ resident, I would love to see this happen. The challenge is those peers have been perceived on a different level for many decades. Going back as far as the USN ranking existed in 1983, UCLA, UCB, UNC, UMich and UVA have never been ranked worse than 30, and often a meaningfully lower number. So the bar is high to break into that top public tier.
If Rutgers wants to be serious about it, they should stop bleeding money into sports spending that is massively unprofitable. When we toured Rutgers, sports is all the tour guide wanted to talk about. He didn’t even know the names of academic buildings. My kids walked away very turned off.
Of course, what must be realized is that graduation rates mostly proxy:
- Academic strength of incoming students. A college with incoming frosh who were 4.0 GPA HS students will typically have higher graduation rates than one with incoming frosh who were 2.5 GPA HS students.
- Student ability to afford the college. A college can affect this with its financial aid and scholarship policies. It can also affect this by using admission criteria that favor students from higher money families.
Other factors that can matter include the mix of majors, volume of non-traditional students, volume and nature of academic requirements, and academic advising for undecided / undeclared students. Students who do co-op or other gap semester activities during college will have extended calendar time to graduation even if they do not need extra semesters of school.
An individual student’s likelihood of graduating within a given time frame is mostly based on the student’s characteristics like prior academic performance, and the student’s characteristics relative to the college in terms of financial affordability. Choice of major may have some effect, and undecided major may increase the risk of late graduation.
2 Likes
Not sure what the word “position” means. In teh first 3 years of USNews ranking, Cal was ranked 5th, 7th, & 5th, respectively. Michigan was also top 10 in the first couple of years.
1 Like
By “never been higher” I meant it has always been a rank of 30, 29, 28, etc. – not 30, 31, 32, etc.
1 Like
What I keep hearing is we shouldn’t use this factor because it disadvantages public universities or vice versa. Let’s just recognize that they are apples and oranges and use different criteria to rank so one can have “useful” criteria to rank private colleges relative to one another and same with public colleges. What doesn’t make sense is ranking small schools with criteria that on their face favor large schools and vice versa for public universities.
1 Like
If the basis of comparison is:
- 1,000 lecture by a faculty member with 30 discussions by TAs (typical intro class model at big research university)
- 30 lecture/discussion classes by faculty members (LAC model)
- 30 lecture/discussion classes by TAs (English composition, beginning foreign language model at big university)
It is likely that instructional quality will vary in the latter two cases much more than in the first case. In the latter two cases, some students may get good instructors but others may get bad instructors.
In the first case, the student also gets two instructors, and could also seek different TAs during office hours if they still have difficulty understanding how the faculty instructor and the TA in the discussion explain things.
So it is not necessarily the case that what you are proposing is better for all students in all types of classes. Also, in popular majors, it may not be possible to accommodate all interested students without increasing class sizes (e.g. some LACs like Pomona and Swarthmore are having difficult accommodating all students who want to take CS classes).
1 Like
Except that is the sole purpose of USNews rankings – to compare small and large Unis, public and private. And also, baccalaureate-granting institutions against each other (primarily LACs).
Something must be working of them as they just passed the 40-year mark of conducting rankings.
2 Likes