<p>One other factor here is to consider how different an student's experience can be depending on their course of study. For example, language classes and classes in the less popular majors tend to be smaller. If you are a psych major or an econ major, chances are good that you will be sitting in a lecture hall a lot more often than your room-mate majoring in German.</p>
<p>^I was about to bring up this exact point. In my opinion, school-wide class size data is meaningless because class sizes differ so much b/w majors. Bio courses are going to be huge at almost every school and science courses in general lend themselves to large lecture-style classes. English, History, etc. courses are going to be much smaller. At some point, if you are taking liberal arts courses, you are going to have to ask yourself if it's going to make much difference being in a 5 person class vs. a 12 person class. That's over a 200%difference in class size and yet I'm not so sure that the educational experience is much different. In fact, a 12 person class might offer a broader range of perspectives than a 5 person class.</p>
<p>Also, I don't know what we're quantifying as "small", but I've had classes with 60-100 people that have offered the best discussions/arguments in my entire college career... and classes with 15 and fewer people that were TERRIBLE and discouraged student engagement. </p>
<p>My point is: smaller is better in some instances, language classes, research seminars, etc... but for some of us, being able to sit in the back of the lecture hall and just TAKE IN the material is quite beneficial.</p>
<p>One of the previous times that this issue came up (i.e. there are more small classes but due to smallness, less chance an individual student will have one, and vice versa) the conversation went the same way--adding smaller classmates was immediately dismissed as "gaming the numbers." </p>
<p>I challenge that. </p>
<p>A university might have very sound reasons for deciding that, say, every freshman should have a Greats Books class (for example) with 20 or fewer students. Naturally, doing something like that is going to mean a lot of sections being added--if you have 2000 freshmen, you have to offer 100 sections to make that happen. That may have a marked effect on your class size statistics, while only increasing the number of small classes on any freshman's schedule by just one. </p>
<p>It's completely appropriate to point out how the math works. But it's not necessarily some weaselly move on the institution's part.</p>
<p>The "weaselly" move is whether or not the school honors the Common Data Set instructions not to count tutorials, directed study, or any other type of one-on-one teaching. </p>
<p>A school that has widespread one-on-one teaching and honors the letter of the Common Data Set rules will be statistically penalized, especially when such one-on-one tutorials are so common that they are assigned course numbers by virtually every department. That school would be better off mandating a minimum of two students so that it gets "credit" for personalized instruction!</p>
<p>Also, in your great book example, are these 100 sections of twenty students still going to a single large weekly lecture? If so, isn't that really one large class with small sections under the Common Data Set definitions?</p>
<p>Some people are shy...they get intimidated when a professor/GSI is intently waiting for them to answer a question.</p>
<p>A lot of my engineering friends learn best by sitting in the back of a large lecture class and absorbing the material like sponges (they also do the reading and the homework).</p>
<p>How do the stats account for the possibility that many classes are small because because no one wants to take the course? If classes have 4 or 5students because the subject is one that few people find relevant or because the profs are just plain awful (good profs tend to fill their classrooms), the stats might be a bit misleading.</p>
<p>Ha! Good question wobudong.</p>
<p>
[quote]
Aren't you really saying that 50% of students at Stanford will incur a class of 50+ students during their four years? It's inconceivable that they will incur a class of 50+ 50% of the time, i.e., two years' worth of classes.
[/quote]
</p>
<p>Bluebayou: Indeed, this means that almost 50% of all “students’ class experiences” at Stanford are in class sizes of above 50. Yes, that was eye-opening to me too. That’s one of the main reasons my wife wishes she hadn’t gone to Stanford as an undergrad but had gone to a small LAC. Of course, as mentioned in other posts, an individual student’s class size experience can vary a lot depending on which major they choose, so you’d have to overlay that factor on the overall statistics. And even more depends on what you’re hoping to get out of the college experience – S was clear that he wanted to go to a LAC, but D may or may not: I’m sure I’ll be asking questions on this in the future.</p>
<p>My intention here is not to plug for one kind of institution versus another – class size is obviously only one of many variables, and for some of you it may not even matter. But for those to whom it does matter, it would be useful to have more meaningful stats than what I’d seen earlier. (I think the logic of my suggested method is ok, but if I’ve goofed up, please post a counterexample to clarify.)</p>
<p>Alexandre: Awfully nice of you to give me the green light to do the #s for Michigan! ;) Here you go: 58% of student experiences in classes > 50 students, 14% in classes < 20</p>
<p>One correction btw: Harvard’s avg that I reported above should have been roughly 100, not 120, to be comparable with the others. But I would caution again that I would not use the average very much since it can be significantly affected by the tail of the distribution – I would suggest instead using the >50 and <20 numbers which are much more robust. In retrospect, I should have bagged posting the averages.</p>
<p>
[quote]
I think the logic of my suggested method is ok, but if I’ve goofed up, please post a counterexample to clarify.
[/quote]
</p>
<p>I'll have to think about this for a while. I appreciate your call for specific examples to clarify how the statistics can be summarized most helpfully.</p>
<p>dadx2:</p>
<p>Your numbers are plausible for Swarthmore. My daughter had 1 out of 28 courses over fifty students. So, she was below your average. However, her first-year science course was a seminar on Special Relativity with 15 students that met for three hours one night a week, plus the lab section, plus the Sunday night student problem set group, plus the Wednesday afternoon professor's office hours (for vexing problem sets). The prof. taught three sections.</p>
<p>A pre-med student doing both semesters of intro Bio and intro Chem plus Orgo could have four courses of 50, which would top the 9% figure you got.</p>
<p>So, overall, that number is plausible, increased a bit by the relatively large number of science and engineering majors at Swarthmore.</p>
<p>I believe that her smallest course (aside from one-on-one senior thesis) was five students in a course on Research Design in social sciences. That ended up being a very good experience as the five students worked with the professor to design and do some kind of campus survey project the school was interested in. Five plus the professor made for a nice collaborative-learning team experience.</p>
<p>
[quote]
Also, in your great book example, are these 100 sections of twenty students still going to a single large weekly lecture? If so, isn't that really one large class with small sections under the Common Data Set definitions?
[/quote]
</p>
<p>No, that's not what I meant. </p>
<p>I was talking about a university initiating a new course with small enrollments for all freshmen, not a new large lecture class with small discussion/recitation sections.</p>
<p>I gave it a try with the data that I have collected for colleges (either via CDS or the college's website) in the USNWR Top 30 National Universities and came up with similar results to what the OP got. Below is a fuller set of data, but there are still many gaps. If anyone has the class size data for any of the missing colleges, please post the information or a link. Thanks.</p>
<p>Here are the colleges in the USNWR Top 30 national universities that provide the highest % of student experiences for classes with more than 50 students:</p>
<p>% of Student Experiences with 50 or more students , College</p>
<pre><code> All Data from 2007-08 CDS or website
</code></pre>
<p>67% , UCLA (06-07)
58% , U Michigan
58% , UC Berkeley
58% , Cornell
55% , MIT (06-07)
52% , U Virginia***
51% , Brown***
50% , Johns Hopkins
50% , Stanford***
48% , Princeton***
46% , U North Carolina***
43% , Northwestern (06-07)***
42% , Yale***
41% , Carnegie Mellon
34% , Dartmouth***
33% , Rice***
31% , Emory***
31% , Vanderbilt***
9% , Wake Forest***</p>
<p>Information Not Available </p>
<p>na , Harvard***
na , U Penn
na , Caltech***
na , Duke***
na , Columbia
na , U Chicago
na , Wash U***
na , Notre Dame***
na , Georgetown***
na , USC
na , Tufts***</p>
<p>Colleges that have a *** after their name denotes a colleges that was recognized in a USNWR survey as providing excellent classroom teaching.</p>
<p>
[quote]
Colleges that have a *** after their name denotes a colleges that was recognized in a USNWR survey as providing excellent classroom teaching.
[/quote]
You still like that ancient survey, don't you, Hawkette?</p>
<p>Why don't you normalize the data for the amount of math, engineering and science majors at each school? It is likely that lower-division math, science and engineering courses will have the largest class size due to the nature of the material being taught.</p>
<p>UCB,
Based on this data, it does look to me like there is a relationship between the great teaching colleges and those colleges that offer smaller classes. Sorry. </p>
<p>Here is the data for the same group of colleges as measured for class sizes of fewer than 20 students. </p>
<p>% of student experiences with class sizes of fewer than 20 students , College
All Data from 2007-08 CDS or website</p>
<p>36% , Emory***
35% , Yale***
33% , Wake Forest***
32% , Princeton***
32% , Dartmouth***
32% , Northwestern (06-07)***
32% , Vanderbilt***
31% , Rice***
30% , Stanford***
26% , Carnegie Mellon
26% , Johns Hopkins
25% , Brown***
21% , Cornell
19% , MIT (06-07)
18% , UC Berkeley
16% , U Virginia***
16% , U North Carolina***
14% , UCLA (06-07)
14% , U Michigan</p>
<p>Information Not Available </p>
<p>na , Harvard***
na , U Penn
na , Caltech***
na , Duke***
na , Columbia
na , U Chicago***
na , Wash U***
na , Notre Dame***
na , Georgetown***
na , USC
na , Tufts***</p>
<p>Colleges that have a *** after their name denotes a college that was recognized in a USNWR survey as providing excellent classroom teaching.</p>
<p>
[quote]
Based on this data, it does look to me like there is a relationship between the great teaching colleges and those colleges that offer smaller classes. Sorry.
[/quote]
No need to be sorry...I agree that the data shows a relationship. What about my other point though?...it looks like colleges that are known for science/engineering programs have larger classes...this is likely because you don't need a small intimate environment to teach fact-based math and science like you do humanities and social sciences...there is less need for student interaction/debate/exchange of ideas in lower-division math/science courses.</p>
<p>
[quote]
a USNWR survey as providing excellent classroom teaching.
[/quote]
A survey conducted in 1995...</p>
<p>I don't know about the relationship you are asking about. If you want to do the work, I'd be interested to see the results. </p>
<p>As for the USNWR classroom teaching survey, I agree that it's old news and perhaps not applicable to today's campuses (and not even sure if it was accurate back then-I mean, who can accurately measure such a thing???). But it is interesting to see that the colleges that were recognized for this have smaller classes on average (today and probably also back then).</p>
<p>I assume that the classroom teaching excellence survey was filled out by an entirely different breed of academics than the ones hawkette regularly disparages as cabalistic, narrow-minded, conspiratorial, elitist and uninformed about undergraduate education. In other words, the ones who fill out that peer ranking. :)</p>
<p>I would imagine that faculty get their impressions of an institution's undergrad teaching via their contact with grad students who got their baccalaureates there; articles in the popular and higher er press about undergraduate curriculum initiatives; research and case studies on pedagogy that make their way into academic journals. I am sure they are also influenced by their own beliefs about the role that class size, school size, admissions selectivity, and overall reputation play in "classroom teaching excellence," whether or not they know anything specific about the teaching at the institution.</p>