Best class I have had. Too powerful for 50 people.
I thought it would be interesting to see which of the US News top dozen “Top Public Schools” also made the “Best Undergraduate Teaching” rankings. It is very interesting (yes, I know the latter ranking is based solely on a “peer assessment survey”). Anyway… get ready…. And fight! 
- Cal: #62 (tie)
- UCLA: n/a
- Michigan: #16
- UNC: #28 (tie)
- UVA: #48 (tie)
- UC Davis: n/a
- UC San Diego: n/a
- University of Florida: n/a
- University of Texas: #39
- Georgia Tech: n/a
- UC Irvine: #62
- UC Santa Barbara: n/a
- University of Illinois: n/a
- University of Wisconsin: n/a
Side note: Cal at #62 for “Best Undergraduate Teaching” is tied with UC Merced & UC Riverside, but behind UC Santa Cruz at #48 and Cal State Fullerton at #31. Unlike UCLA and others in the top dozen “Top Public Schools” rankings, though, at least they made the list!
4 Likes
IMO, where small class size matters the most are upper level courses where deep dives, tangential queries and divergent points of views and ways to solve complex problems are most important. Econ 101, CS 101, Chem 101, etc… generally are just laying a common foundation for more advanced courses. One very good instructor (could be tenured or not faculty) who is good at teaching and has a class of 200 or more is better than having 5 or more faculty of varying competencies teaching 40 person classes. If we look at class sizes for some level of courses and up, it evens the playing field between public/private, large/small to a great extent of something that is very important for advanced classes and an indicator of the quality of education being provided.
That doesn’t seem to be the sole purpose as they also separately rank liberal arts colleges.
I am somewhat surprised that people do not agree that smaller classes are generally better and preferred. For me, it is no different from student and faculty spend - generally better. Of course it is not better for every person and every situation, but that’s common sense and does not negate the facts and studies on the subject.
9 Likes
I.e. often where the content of the course approaches the frontiers of knowledge in the subject. Whether upper level undergraduate courses commonly do this does depend on the subject and particular course.
Of course, some subjects even at the lower / beginner level do rely more heavily on in-class discussion. Beginning foreign language is an example. However, the instructor for that level of course may not necessarily have to be a tenure track faculty member.
In isolation, holding everything else the same (e.g. the instructors and their teaching quality), that may be the case. But there often have to be trade-offs that may affect instructional quality or consistency, or affect access to a popular course. If only 33 students want to take introduction to CS, great. But if 1,000 students want to take introduction to CS, how should the school handle that?
We almost agree. The larger issue is that it is somewhat obvious that you go to a small liberal arts school and a larger university for different reasons and people do not typically apply to both styles of schools together. On the other hand, state and private colleges with 10k+ people should probably be combined because it helps people better understand that set of schools. The issue here IMO is that US News made some changes that likely resulted in some funky math and rankings that favored state schools. And that does not match up to people perceptions and more importantly admissions odds.
And also, I would prefer for them to figure out what the criteria should be to measure a college. For me, it goes back to input (quality of student), learning (quality of education), and output (ROI). State schools, private schools, specialty schools, liberal arts schools should all be accountable.
2 Likes
And that is why HigherEd is a business.
There are a lot of current or formerly ranked T50 schools with less than 10,000 students. These are primarily but not exclusively the private schools that dropped in the current rankings.
1 Like
Life can be hard for the “caught in the middle schools.” Interestingly, those are the schools many people prefer. not too big and not too small.
3 Likes
If this happens in this tech driven market, it’s probably because the school doesn’t have a CS major, or the department is terrible.
The class size argument, where small is better, is way too simplistic. If you don’t have a critical mass and have very small lower div classes then the upper div offerings are likely to be meager and diluted. And you won’t have an ecosystem of research labs and startups if you have too small a class. As I mentioned above, even at Harvard and Stanford intro CS classes run from 700 to 900.
2 Likes
And yet Stanford and Harvard do well on the class size metric so there is some difference between what they are providing and what the big public colleges are providing. And rankings should ultimately address many majors beyond computer science.
1 Like
I did the same for the top dozen “Top National Universities” and their “Best Undergraduate Teaching” rankings:
- Princeton: #4
- MIT: #48 (tie)
- Harvard: #48 (tie)
- Stanford: #24
- Yale: #16
- Penn: #62 (tie)
- Cal Tech: n/a
- Duke: #9
- Brown: #3
- Johns Hopkins: n/a
- Northwestern: n/a
- Columbia: n/a
- Cornell: #39
- Chicago: #48
1 Like
Fixed that for you. 
Any ranking/rating only makes sense if it aligns with the criteria the consumer believes those things should be rated by. In 875 responses to this thread I’ve yet to hear a single person say they agree with the criteria/weighting USNWR used.
For Colleges, all should be ranked together (if at all).
You brought up the CS specific class size issue at Berkeley. At Berkeley, across all majors inclusive of both lower divs and upper divs, 70% of classes have less than 50 students. And if you just look at upper divs, 80% of classes have less than 50 students.
Given that your entire argument is driven by metrics that are based solely on the large CS/DS class sizes in the lower divs, you then can’t argue to keep aside that facts that show these large sizes are a testament to the strengths of those programs.
1 Like
I’ll bite. I approve.
1 Like
And the same for the top 13 “National Liberal Arts Colleges” (taking out the military/service academies) and their “Best Undergraduate Teaching” rankings:
- Williams: #12
- Amherst: #1 (tie)
- Pomona: #1 (tie)
- Swarthmore: #4 (tie)
- Wellesley: #20 (tie)
- Bowdoin: #4 (tie)
- Carleton: #1 (tie)
- Barnard: #25 (tie)
- Claremont McKenna: n/a
- Grinnell: #6 (tie)
- Middlebury: #12 (tie)
- Wesleyan: #29 (tie)
- Davidson: #6 (tie)
- Hamilton: #36 (tie)
- Harvey Mudd: #32 (tie)
- Smith: #20 (tie)
3 Likes
100% correct.
How “percentage of large/small classes” is looked at in this way does not mean that this is what a student sees. Looking at it in this way makes it seem like all colleges have smaller class sizes than what an average student will encounter.
For example, suppose a college has one class with 910 students and nine classes with 10 students each. Clearly, 90% of the classes are small. However, out of the 1,000 students, 910 (or 91%) are in the large class, and 90 (9%) are in the small class.
Average class size is also misleading. In the above example, the average class size is 100. But if you take the average class size from the student point of view, it is 829, because the 910 students in the large class weight the average so much more than the 90 students in the small classes.
2 Likes