USNWR 2009: Looking at the Data XI (Class Sizes)

College Search & Selection
1

<p>The publication of the 2009 USNWR College Rankings provides an opportunity to compare schools based on a wide variety of data points. In this and in other threads, I urge the reader to think less about the absolute rankings and more about the nature and value-added of the data point being discussed.</p>

<p>In addition to the straight recitation of the USNWR data related to class sizes that is presented below, I did some analysis to compare the relative classroom size experience at these colleges. To do this, I used the following weights:</p>

<p>3 points for % of classes with fewer than 20 students</p>

<p>2 points for % of classes with 20-50 students</p>

<p>1 point for % of classes with more than 50 students</p>

<p>After creating a score I then benchmarked each school against the top score. I believe that this scoring/benchmarking presents a clearer picture of the impact of smaller or larger classes at these colleges. Following is the full accounting:</p>

<p>Index , Class Size Score , National University , < 20 (3 points) , 20 - 50 (2 points) , > 50 (1 point)</p>

<p>100% , 2.68 , Harvard , 75% , 17% , 9%
100% , 2.68 , Columbia , 76% , 16% , 8%
100% , 2.68 , U Chicago , 72% , 24% , 4%
100% , 2.68 , Tufts , 74% , 21% , 4%
100% , 2.68 , Yeshiva , 69% , 30% , 1%
100% , 2.67 , Yale , 75% , 17% , 8%
100% , 2.67 , U Penn , 74% , 19% , 7%
99% , 2.66 , Northwestern , 75% , 17% , 7%
99% , 2.65 , Stanford , 74% , 16% , 11%
99% , 2.65 , Duke , 70% , 25% , 5%
98% , 2.63 , Princeton , 73% , 17% , 10%
97% , 2.61 , Caltech , 69% , 23% , 8%
97% , 2.61 , Wash U , 72% , 18% , 9%
97% , 2.61 , Rice , 68% , 25% , 7%
97% , 2.61 , Vanderbilt , 67% , 27% , 6%
97% , 2.6 , Emory , 68% , 25% , 6%
97% , 2.59 , Brown , 70% , 20% , 9%
96% , 2.58 , Brandeis , 66% , 27% , 6%
96% , 2.56 , Carnegie Mellon , 65% , 26% , 9%
95% , 2.55 , Dartmouth , 64% , 27% , 9%
95% , 2.55 , Wake Forest , 57% , 41% , 2%
95% , 2.54 , Johns Hopkins , 65% , 24% , 11%
95% , 2.54 , Tulane , 62% , 30% , 8%
94% , 2.52 , USC , 64% , 24% , 12%
94% , 2.52 , Case Western , 62% , 28% , 10%
93% , 2.5 , MIT , 64% , 23% , 12%
93% , 2.5 , U Rochester , 62% , 26% , 12%
93% , 2.49 , Georgetown , 58% , 34% , 7%
93% , 2.48 , UC Berkeley , 62% , 24% , 14%
92% , 2.46 , Notre Dame , 56% , 34% , 10%
92% , 2.46 , NYU , 58% , 30% , 12%
91% , 2.44 , W&M , 49% , 45% , 7%
91% , 2.43 , Cornell , 60% , 23% , 17%
91% , 2.43 , Rensselaer , 53% , 37% , 10%
90% , 2.41 , Boston Coll , 48% , 45% , 7%
88% , 2.37 , Lehigh , 47% , 43% , 10%
88% , 2.35 , U Virginia , 49% , 37% , 14%
87% , 2.33 , UC Santa Barbara , 50% , 33% , 17%
87% , 2.32 , U North Carolina , 44% , 44% , 12%
86% , 2.31 , UCLA , 53% , 26% , 20%
86% , 2.31 , UC Irvine , 49% , 34% , 16%
85% , 2.28 , U Wisconsin , 44% , 39% , 18%
84% , 2.26 , U Michigan , 44% , 38% , 18%
82% , 2.21 , U Florida , 41% , 39% , 20%
82% , 2.2 , U Washington , 35% , 49% , 17%
82% , 2.19 , U Illinois , 38% , 43% , 19%
81% , 2.18 , Georgia Tech , 40% , 38% , 22%
81% , 2.16 , Penn State , 33% , 50% , 17%
80% , 2.14 , UCSD , 44% , 26% , 30%
79% , 2.12 , U Texas , 35% , 42% , 23%
78% , 2.09 , UC Davis , 35% , 38% , 28%</p>

<p>Index , Class Size Score , LAC , < 20 (3 points) , 20 - 50 (2 points) , > 50 (1 point)</p>

<p>100% , 2.96 , US Military Acad , 96% , 4% , 0%
97% , 2.86 , Claremont McK , 86% , 14% , 0%
93% , 2.75 , Haverford , 76% , 23% , 1%
93% , 2.74 , Swarthmore , 74% , 25% , 2%
93% , 2.74 , Hamilton , 74% , 26% , 0%
92% , 2.72 , Amherst , 75% , 22% , 3%
92% , 2.71 , Pomona , 72% , 27% , 1%
92% , 2.71 , Davidson , 71% , 29% , 0%
91% , 2.69 , Williams , 73% , 23% , 4%
91% , 2.69 , Bryn Mawr , 71% , 27% , 2%
91% , 2.68 , W&L , 68% , 32% , 0%
90% , 2.67 , Middlebury , 71% , 25% , 4%
90% , 2.67 , Bowdoin , 69% , 29% , 2%
90% , 2.66 , Wellesley , 67% , 32% , 1%
90% , 2.66 , Vassar , 68% , 31% , 0%
90% , 2.66 , Oberlin , 70% , 27% , 2%
90% , 2.66 , Macalester , 68% , 30% , 2%
89% , 2.64 , Smith , 68% , 28% , 4%
89% , 2.63 , Colgate , 63% , 36% , 2%
89% , 2.62 , Carleton , 63% , 36% , 1%
88% , 2.61 , US Naval Acad , 61% , 39% , 0%
88% , 2.6 , Grinnell , 60% , 40% , 0%
88% , 2.6 , Harvey Mudd , 64% , 32% , 4%
88% , 2.59 , Bates , 64% , 31% , 5%
87% , 2.57 , Wesleyan , 63% , 31% , 6%
87% , 2.57 , Colby , 61% , 35% , 4%</p>

2

<p>Intersting. Thanks for putting it together.</p>

3

<p>I feel that the methodology for deciding what classes to include or exclude needs to be more detailed. I think there is evidence that schools may count similar classes differently. I don't believe it's with an intent to deceive, but rather lack of clarity, too much complexity in the data, and different interpretations.</p>

<p>Look at U-M and Berkeley. Same student-faculty ratio. Same research focus, no real difference in how many faculty are part-time, and I have to believe the same general expectation as to teaching loads. Yet Berkeley reports 62% of their courses are under 20; Michigan reports only 45%. Wouldn't we expect those two to be closer? </p>

<p>You could explain the difference (given their relative equivalence of faculty per student), if Berkeley profs are teaching a lot more classes than the average faculty, and/or Michigan profs are teaching a lot fewer than the average faculty. However, I haven't heard any kind of assessment like that from faculty or administrators familiar with the two institutions. Which leads me to wonder whether the schools are defining eligible courses differently. With so many course types to contend with, it's plausible that they could each make very different (yet justifiable) decisions about classes that are comparable.</p>

<p>I'm sure there are other examples like this (similar schools, similar ratios, notable difference in class size by this method); I just haven't looked into those. Obviously it's a topic if interest to Michigan because we rank 123rd in the country on this measure. No one is denying that at a big school, small classes aren't as common, and maybe schools like Michigan can and should make some changes that will increase small classes for undergrads. But as we look into peer data (i.e. other big public schools with faculty like ours), it's hard to know where we genuinely stand if we're not sure everyone accounts for classes the same way. </p>

<p>It doesn't seem that USNews does any of this kind of analysis.</p>

4

<p>Although top LACs do relatively well by the existing system, they would probably do even better if the data were parsed even finer. The average "small" class at a LAC is probably significantly smaller than the "small" class at a university, even if both are "< 20". </p>

<p>For example, Williams is well known in the higher education community for capping tutorial enrollments at two. But obviously, the "< 20" category doesn't do justice to this kind of effort.</p>

5

<p>A few observations on the data above:</p>

<ol>
<li><p>The median college in this comparison of 51 colleges had a score of 2.5 (MIT and U Rochester) and a benchmark of 93%. </p></li>
<li><p>Not one of the publics scored as high as the median level. The only ones close were UC Berkeley (whose reported data may have some questions as hoedown notes above) and W&M which is a very small public with less than 6000 undergrads. Class size is a true differentiator between the undergraduate experience available at private colleges and public colleges. The differences are not always great and may not be meaningful for all students, but the class size facts and differences are indisputable.</p></li>
<li><p>The spread among colleges in the top half above the median scored between 93% and 100%. The spread among colleges in the bottom half below the median scored between 78% and 93%. </p></li>
<li><p>Only one college with an undergraduate enrollment above 10,000 students finished in the top half of the comparison of national universities.</p></li>
<li><p>Cornell was the outlier/laggard of the Ivy colleges and was the only Ivy to finish in the bottom half of the national universities. </p></li>
<li><p>Lehigh was the lowest scoring private, but was surpassed by only 2 of the 17 publics ranked. </p></li>
<li><p>The spread among LACs from 2-25 was tight, from 97% to 87%, but not as tight as among the top 25 national universities (from 100% to 93%). </p></li>
<li><p>Of the 25 being compared, the only LAC with more than 5% of its classes having 50 or more students is Wesleyan. </p></li>
<li><p>8 of the 25 LACs indicate that they have no classes with more than 50 students. </p></li>
<li><p>The overall winner among national universities and LACs is the US Military Academy (West Point). 10 LACs had a higher score than the highest scoring national university.</p></li>
</ol>

6

<p>Note to Wesleyan: cut out all those film classes.(jk)</p>

7

<p>Hawkette --</p>

<p>The first problem with these numbers is that they fail to account for independent studies, honors work, or tutorials that many students take at some of these schools as upperclassmen. </p>

<p>The second problem is that I have yet to see any definitive proof that a smaller class size is always better for the student or the institution. For many introductory courses in the sciences and social sciences there is very much an economy of scale to be enjoyed in the presentation of this material. At Cornell, I doubt any student would tell you that their quality of education was compromised by taking a 1,300 person introductory psych course or a 300 person introductory wines course. Needless to say, humanities students at most colleges will experience very different teaching environments than science students, and a lot of the variation across colleges may be attributed to the relative amount of students and faculties in the humanities. This is why schools like MIT and Caltech come in a lot lower than what you would expect.</p>

<p>Beyond that, allocating different points to the different tiers of class sizes strikes me as completely arbitrary and without any empirical basis. A much better approach be a representative student experience calculation.</p>

<p>For instance, 10% of all Princeton courses have more than 50 students. This might reflect 100 courses out of 1000 offered. Assuming 5,000 undergraduates and 5 courses a semester there are a possible 25,000 student-class interactions in any given semester. Assuming that the average lecture with over 50 students has 100 students enrolled, then fully 10,000 of Princeton's student-class interactions each semester will be in lecture halls averaging 100 students each! That's 40% of a student's time at Princeton, and much more representative of the type of classroom experience they will encounter.</p>

8

<p>Cayuga,
While I appreciate your many points, don’t fire too hard at the messenger. This data is all culled from USNWR and presented here.</p>

<p>As for the specific criticisms, </p>

<ol>
<li><p>While I’m not sure how different colleges would be affectted by the reporting/non-reporting of independent studies, tutorials, etc, I’d be interested to read your explanations on this and how it differs among a group of elite colleges, eg, the USNWR Top 20 national universities (nearly all of whom are found in the top half of this comparison) and how this might affect the listing above. Why would the experience at Cornell be so different than the experience at someplace like Rice or Emory or even Lehigh??</p></li>
<li><p>On the issue of smaller class size, choose what you like, but I think that the vast majority of students, parents, maybe even professors, would prefer a smaller rather than larger class setting. I concur that larger sizes are more efficient for the colleges for the broad survey courses often found in one’s freshmen and sophomore years. Still, many colleges find a way to deliver these courses in more manageable sizes which only improves the opportunities for student contact with the professor and other students. </p></li>
<li><p>From the most recent CDS, Cornell reports that 371 classes (out of 2209 classes, or 17% of the total classes offered) have student populations greater than 50 students. So it is more than a few classes like Wine or Film or some of the science classes that are being offered in large class sizes. By comparison, the only other Top 20 national universities with double digit percentages for classes over 50 are Stanford (11%), Princeton (10%), Johns Hopkins (11%), MIT (12%), and Notre Dame (10%).</p></li>
</ol>

<p>As for Pyschology, did you mean 1300 students in one class/one lecture hall? That’s pretty big and, in fact, that size is nearly as large or larger than the entering freshmen at all of the USNWR Top 20 national universities, not to mention the huge difference with those at LACs. BTW, Caltech scores at the 97% level (same as Wash U, Rice, and Vanderbilt) and Carnegie Mellon is at the 96% level. </p>

<ol>
<li><p>Re the scoring methodology, there is nothing magical about it. It’s obviously pretty simple. I chose the 1-2-3 scoring to try to be fair to all colleges, knowing that if I more heavily weighted either the large or small classes, then someone would be upset. If you have a better way of presenting this information for all colleges, then please suggest and provide some results. </p></li>
<li><p>I don’t know for sure, but at Princeton, I doubt that the AVERAGE size of classes over 50 is 100 students. The school only offers 80 classes with more than 50 students. Of these, only 30 classes have 100+ students while 50 classes have between 50-99 students. I also highly doubt that 40% of a student’s time at Princeton is spent in classes with 100 students or more.</p></li>
</ol>

9

<p>There's tons of literature in learning research that addresses the enormous added value of engaged, interactive learning vs. the traditional and necessarily passive learning of the large lecture hall. While there are many variants in examining this issue of "quality" in learning, the primary fly in the ointment is most often numbers of students, i.e. student-faculty ratios, classroom sizes and attendee numbers, and engagement in social activities.
Pace and Astin talk much about time engaged on task and in interaction with peers and professors on meaningful activities and issues. And one of the great conclusions, was the too-often poor or lesser value necessarily provided at the public, lower cost institutions trumps being able to live and learn in highly interactive arenas.. Sadly, the perceived economics inherent in the public vs. private campus scenario see more and more families and their students being forced to select in-state public institutions ...not because they provide superior or higher quality education, but because that's what's perceived as affordable and too often needed. </p>

<p>Confounding this is higher ed's inability to successfully market "discounted" tuitions and financial aid, i.e. the "average" student paying substantially less than 100% tuition, or the importance and reality of getting a "better" education that aggressively engages each and every student vs. merely getting a degree. They are severely different things too many places.</p>

<p>I guess my point is ...size matters alot. If one is genuinely interested in learning vs. graduating. And that's a mountain of evidence supporting this notion. </p>

<p>So why are the Ohio, Penn, and Michigan states busting @ the seams? Again, money and the lack of evidence illumining that it's not who we know, but what we know. Think about it.</p>

10

<p>Please don't characterize my comments about Berkeley to mean their numbers are "questionable!" It's just as likely that the other publics have calculated their figures in a "questionable" way (although that term is not really correct here). The point is, we don't know. </p>

<p>And issues like how courses are labeled may also have an impact. What one school calls a seminar, another may call a practicum or clinical, and the guidelines may cause one kind of course to be included and another not, even though they are essentially the same kind of course. That's just an example--I don't have the definitions in front of me, so I can't say, but it's an example of how naming conventions could also play a role.</p>

<p>There doesn't have to be a speck of dishonesty involved--but there is room for inconsistency and in such cases even schools acting on good faith can end up sending in non-comparable numbers.</p>

<p>Whistle pig, what do you mean by ___ states bursting at the seams? Do you mean the state schools within those states?</p>

11

<p>I mean Ohio State ... Penn State... Michigan State ... which collectively may be written as Ohio, Penn, Michigan states (small "s"). And all are huge with the propensity to become ever more so. Why? Again, $ not quality.</p>

12

<p>
[quote]
Although top LACs do relatively well by the existing system, they would probably do even better if the data were parsed even finer. The average "small" class at a LAC is probably significantly smaller than the "small" class at a university, even if both are "< 20".</p>

<p>For example, Williams is well known in the higher education community for capping tutorial enrollments at two. But obviously, the "< 20" category doesn't do justice to this kind of effort.

[/quote]
</p>

<p>I think the question is whether you get any value added by being in a 3 person class vs. a 20 person class. Personally, for an English or History class, I'd rather have 15-25 people than 5 people. A class size of 15-25 assures ample speaking opportunities while you are still exposed to a variety of perspectives and opinions. Unless it's a review or remediation course, I see no reason to limit class sizes to under 10.</p>

<p>As for science courses, I would actually prefer a 400 person lecture to a 100 person lecture. I'm in medical school right now and our graduating class is around 150 which means it's small enough for people to ask questions during class. With no disrespect to my classmates or their questions, I absolutely do not want people asking questions during science lecture. I hate having to start and stop 10 times during lecture so the professor can have a two-way conversation with a student while the other 149 of us just sit there. If you have a question, go see the professor after class, during breaks, or during office hours. I don't want to be 15 minutes late getting out of class just because you needed to ask 5 questions in a row. </p>

<p>At Cornell, I was a bio major (the largest major at Cornell or any other university) and I never felt that the professors were inaccessible. I never had to wait during office hours and in fact many of our professors begged us to come talk to them during office hours. So, for science majors, I personally prefer large lecture-style courses.</p>

13

<p>
[quote]
I think the question is whether you get any value added by being in a 3 person class vs. a 20 person class.

[/quote]
Those who have actually tried "ultrasmall" classes -- and admittedly, only a tiny fraction of undergraduates get this opportunity -- tend to be enthusiastic. As [url=<a href="http://chronicle.com/free/v48/i23/23a01601.htm%5Dreported%5B/url"&gt;http://chronicle.com/free/v48/i23/23a01601.htm]reported[/url&lt;/a&gt;] by the Chronicle of Higher Education, for example:

[quote]
Nearly everyone involved in the tutorials at Williams loves them. The students, while maintaining that tutorials are harder than their regular courses, often say they're the best courses they take. For their part, professors enjoy the intense, personal nature of the class and the chance to worry more about the ideas than class dynamics ... In students' course evaluations, tutorials outrank other upper-level courses in their "educational value," and a 1997 survey of alumni who took tutorials found that 80 percent regarded the tutorial as the most valuable of their Williams courses.

[/quote]
</p>

<p>It seems likely that (at least) some of the students at other schools would appreciate the opportunity to (at least) try this approach.</p>

14

<p>This is not about what we think, whether we "prefer" 400 vs. 100...I hope he's not doing my brain transplant. Silly observation, imo. It's about what the research says. For anyone needing a start on this, check out the monograph "Involvement in Learning" edited by Kenneth Mortimer. Astin and all the biggies in the field were in on this one. While it was far from the early or even seminal word, it summarizes alot of the research to that time, most of which has merely been confirmed. </p>

<p>Wonder why tutors are usually 1:1 vs. 400:1? USMA doesn't have 400 kids in a calculus class. Usually a handful or 2 ...because you need to practice and perform every day. Wonder why recitals don't have 100 violinists? Wonder why labs are 10, even when the class is 150? Wonder why the Yale writing program, emulated at many top-rated institutions, usually only has 8 or 10 in a class? </p>

<p>btw, biology is rarely the largest major at most campuses.</p>

<p>There are many ways to defend those institutions that must have large sections of many classes, but one of them is not quality learning or even preferred method of instruction and student learning.</p>

15

<p>^I would beg to differ. While small classes may be optimal for some students, others learn better in large classes. Science courses lend themselves to being lectures. You need to pass a large volume of relatively objective information/facts onto the students. Lecture would be the most efficient way of doing that. Whether you're in a large state school or a smaller Ivy League school, your science courses are mostly lectures while your English courses are mostly discussions. Hence, it's silly to look school-wide data when the biggest difference in class size is major. If you are an English or history major, you will be assured of relatively small class sizes. You may argue that a history course at Berkeley might have 25 students while history course at Williams might have 10 students but I wonder whether 10 students is better than 25 students in a discussion course (for the reasons I stated in my first post). Meanwhile bio lectures are large in every school. Is a bio lecture of 200 students really different from a bio lecture of 400 students? It still consists of a professor trying to get through 30 powerpoint slides in 50 minutes. If the accessibility to the professor is not affected, then I don't see lecture-size being much of an issue.</p>

<p>Again, I'm not saying that my preference can be applied to all (or even most) students. But, it's silly to try to tell me how I would learn best. Each person has their own learning style. In medical school, you are certainly expected to learn on your own. Professors prepare 100 powerpoint slides that they know they won't get through in 50 minutes and you simply have the learn the rest by yourself. Heck, a large proportion of medical students don't bother attending lecture at all. I would much prefer a college that prepares me for that aspect of medical school education than one that holds my hand every step of the way.</p>

16

<p>
[quote]
Those who have actually tried "ultrasmall" classes -- and admittedly, only a tiny fraction of undergraduates get this opportunity -- tend to be enthusiastic...
It seems likely that (at least) some of the students at other schools would appreciate the opportunity to (at least) try this approach.

[/quote]
</p>

<p>I don't disagree with this. I think some students would LOVE these types of classes. Although, I think asking LAC kids whether they like small classes is kinda moot ;)</p>

<p>We actually have something similar in medical school called PBL (Problem-Based Learning). PBL consists of 6-8 students in a small group with a faculty facilitator. The students are presented with various cases ("50 year old white male comes in complaining of chest pain...") and the students are supposed to discuss various aspects of the case with little to no input from the faculty member. The students then divide up and research the learning issues and then present them to each other the next class. </p>

<p>Most med students do not like this class. I asked a lot of med students on my med school interviews about their opinions of PBL and there were around 1/3 who were positive or neutral and 2/3's who were negative about the course. They admitted that they put in very little effort and learned very little from the course. Administrators have even admitted that there is a general lack of student enthusiasm about this mode of learning. Now that I'm in med school, I actually tend to like PBL. However, my point being, for certain majors and certain personalities, small or ultra-small classes are not optimal.</p>

17

<p>
[quote]
Why would the experience at Cornell be so different than the experience at someplace like Rice or Emory or even Lehigh??

[/quote]
</p>

<p>Let me give you an example from Cornell. Probably half of all freshmen biology and physics students take the course in 'tutorial' mode where they are largely responsible for the pace of their own learning, but have access to professors, lab instructors, graduate student tutors, and upperclassmen undergraduates as a support network in their study.</p>

<p>Cornell has found that this type of educational experience is often the best way for students to learn this type of material. There is a body of knowledge that needs to be digested and often the lecture hall (even one with less than 50 students!) isn't the most conducive to this type of knowledge transfer. Most Cornell students in the life sciences consider this system to be a success and superior to sitting in a lecture hall. (These students are wildly successful, by the way... Cornell produces more students who go on for MDs and PHDs in the life sciences than any other school in the nation.)</p>

<p>Without knowing much about these other institutions, I'm not certain it is different. But I believe it should be taken into consideration. And it is currently not.</p>

<p>
[quote]
I concur that larger sizes are more efficient for the colleges for the broad survey courses often found in one’s freshmen and sophomore years. Still, many colleges find a way to deliver these courses in more manageable sizes which only improves the opportunities for student contact with the professor and other students.

[/quote]
</p>

<p>I would argue that all of the institutions listed above provide for a lot of opportunities for small class sizes and student-led learning. My friends who attended UCSD or UCLA didn't seem to have a meaningful difference in their classroom experience from my friends who attended WashU or Georgetown. </p>

<p>But more importantly, it's a question of balance and what is best for each institution. Some institutions will have more science and social science introductory courses, which, quite frankly, lend themselves more to a large-lecture setting. Others will have relatively higher concentrations of language and humanities courses, which need smaller class sizes. To blindly compare these institutions without taking these factors into account, as both you and U.S. News do, is disingenuous. </p>

<p>In a lot of cases, one lecture of 70 students instead of two lectures of 35 students each will free up a lot more time for a professor to spend interacting with students in a more meaningful setting --- office hours, lunches, lab work, etc.</p>

<p>
[quote]
As for Pyschology, did you mean 1300 students in one class/one lecture hall? That’s pretty big and, in fact, that size is nearly as large or larger than the entering freshmen at all of the USNWR Top 20 national universities, not to mention the huge difference with those at LACs.

[/quote]
</p>

<p>It is pretty big. But most Cornell students will tell you that both wines and intro to psych are fantastic classes that they learned a lot from. Cornell seems to be doing something right.</p>

<p>
[quote]
. I chose the 1-2-3 scoring to try to be fair to all colleges, knowing that if I more heavily weighted either the large or small classes, then someone would be upset. If you have a better way of presenting this information for all colleges, then please suggest and provide some results.

[/quote]
</p>

<p>The problem is that, like so many other of these arbitrary statistics you trot out, it has no intuitive meaning and isn't grounded in any sort of empirical understanding of the subject. You list 100% for Harvard and 91% for Cornell. Does this mean that Harvard is 9 percentage points better for educating students than Cornell? If not, what does it tell us?</p>

<p>You ask me to suggest a way I would rather see the results. I did so in the 'representative' student-class interactions I laid out for you. Unfortunately I do not have the time to calculate such metrics for each of these schools. Perhaps you would be interested in such an exercise.</p>

<p>
[quote]
I don’t know for sure, but at Princeton, I doubt that the AVERAGE size of classes over 50 is 100 students. The school only offers 80 classes with more than 50 students. Of these, only 30 classes have 100+ students while 50 classes have between 50-99 students. I also highly doubt that 40% of a student’s time at Princeton is spent in classes with 100 students or more.

[/quote]
</p>

<p>I never said that 40% of a student's time at Princeton is spent in classes with 100 students or more. I said that 40% of a student's time at Princeton is spent in large lecture halls, which probably average 100 students. (The distribution of large classes tend to be right skewed.)</p>

<p>On an anecdotal note, I have a couple of friends from Cornell who are now doctoral students in the biological sciences at Princeton. They aren't impressed with the quality of the undergraduate educational experience there, saying that the undergrads hardly have any interactions with professors of the type they enjoyed at Cornell. You can't really quantify that.</p>

<hr>

<p>One more thing about tutorials. I had the privilege of studying at Oxford, which operates exclusively on the tutorial system. It was wonderful for certain classes, like philosophy, history, and literature, but is very much a "luxury good" and I'm not certain if I learned all that much more than had I been in a 15-25 person class. And for economics or math, I failed to see the benefit vis-a-vis my experience at Cornell.</p>

18

<p>"On an anecdotal note, I have a couple of friends from Cornell who are now doctoral students in the biological sciences at Princeton. They aren't impressed with the quality of the undergraduate educational experience there, saying that the undergrads hardly have any interactions with professors of the type they enjoyed at Cornell. You can't really quantify that."</p>

<p>Dont' be ridiculous. Your friends are doctoral students and they want to judge Princeotn's undergraduate experience? I have access to all of my professors here, and they are very open to students. Princeton, first and foremost, is an undergraduate institution</p>

<p>Clearly any anecdotal note that your friends tell you are going to be skewed towards their Cornell experience.</p>

19

<p>
[quote]
Let me give you an example from Cornell. Probably half of all freshmen biology and physics students take the course in 'tutorial' mode where they are largely responsible for the pace of their own learning, but have access to professors, lab instructors, graduate student tutors, and upperclassmen undergraduates as a support network in their study.</p>

<p>Cornell has found that this type of educational experience is often the best way for students to learn this type of material. There is a body of knowledge that needs to be digested and often the lecture hall (even one with less than 50 students!) isn't the most conducive to this type of knowledge transfer. Most Cornell students in the life sciences consider this system to be a success and superior to sitting in a lecture hall. (These students are wildly successful, by the way... Cornell produces more students who go on for MDs and PHDs in the life sciences than any other school in the nation.)

[/quote]
</p>

<p>This is one of the things I like about Cornell. It lets YOU (the student) take responsibility for your own learning. The resources and help are there but no one is going to be getting on you for using or not using them. </p>

<p>I took the aututorial physics course at Cornell after having not taken any physics at all in my life before then. It was great because I could learn on my own and then seek help when I needed to. It's a great course for students who are self-motivated and are looking for flexibility. Not so great for students who need extrinsic motivation and structure. I ended up with A's in the course and aced the Physical Sciences section on the MCAT (99.7th percentile). That's why this is the predominant physics course for premeds at Cornell. It really mirrors the learning experience you will have in med school and as a physician. It teaches you to take a responsibility in your learning and to find appropriate resources when you encounter something you don't know or understand.</p>

20

<p>
[quote]
I would much prefer a college that prepares me for that aspect of medical school education than one that holds my hand every step of the way.

[/quote]
</p>

<p>I think you've been very insightful and I largely agree with your comments. I do take issue with the idea that a small-class format somehow constitutes "hand-holding." Small courses can be challenging, rigorous and motivating.</p>