In this instance,
Again that is subjective (despite the fact that they say it “allowed us to objectively compare newer and smaller schools with older and larger schools”). Use a different dynamic/scoring system and you get a different list. Which one is “correct”?
Yes, you are likely to get a somewhat different order in each list. Yet, pretty much the same schools are likely to show up near the top whether you measure admission rates, average SAT scores, average HS GPA or class rank distributions, average class sizes, 4y graduation rates, major student awards per capita, average instructional spending per capita, need-based aid, endowment per capita, alumni-earned PhDs per capita, alumni mid-career earnings, average faculty salaries, percentage of faculty with doctorates, major faculty awards, etc. etc.
It may be possible to explain away high performance (or excuse low performance) on any of these individual measurements. When a college performs well in measurement after measurement, it’s reasonable to conclude it must be pretty good. Whether a college ranks 7th or 9th is immaterial. Whether it ranks 11th or 25th doesn’t make much difference either. 25th v. 75th, according to some subset of the above metrics, probably does indicate a significant quality difference, in my opinion. Whether it’s enough to override other considerations (cost, “fit”, etc.) is a matter of personal judgment.
@PurpleTitan It may be true that universities include a higher proportion of students in professional majors, but there are probably political science majors or biology majors who want to go into journalism at liberal arts colleges. Maybe this does not hold true for students interested in nursing, but many liberal arts majors also become professionals: teachers, lawyers, doctors, social workers. Even at Reed and Swarthmore, schools known for PhD production, only about a fifth of students actually go on to get PhDs. Even if not all PhDs work in academia, this would still mean that roughly only four-fifths of Reed or Swarthmore graduates are not in academic careers. If pre-professional majors should not be factored into a per capita study of natural scientists, then the study should probably not include humanities and social science majors at the liberal arts either.
@saillakeerie Any list you can conceivable think of involves selection. The authors of this particular study were very specific about how they defined impact on science, by examining those criteria considered highly by scientists (the Nobel, the Fields, and membership in the NAS and etc.) and what schools produce alumni who go on to receive such honors. What is selective is whether or not you would consider those criteria as the same as having an impact on science. But honestly, if such honors are the highest as considered by the scientific community, I’m inclined to believe to believe that they are acceptable proxies for “impact on science.”
@jademaster, I agree. For STEM achievement, the denominator should really be STEM majors.
^ Perhaps even better would be by department major, to not incorrectly disadvantage, e.g., small schools strong in science but without engineering, i.e., compare physics grads, per capita.
Too bad it only lists American schools. I’ve long thought the University of Edinburgh played the same role 200 years ago that places like Stanford, Harvard, MIT and Berkeley do now.
@tk21769 I agree. Then you get into how much of a difference is relevant? Some may say 25th to 75 is significant. Others may think its 5th to 10th. So to the extent various types of formations yield similar results, you still need to determine what delta is significant. And there are some differences when you use different formulations. Ultimately the subjectivity is reduced somewhat though its not clear how much.
@jademaster Just looking at Nobel winners, is it necessarily clear every time who should win? Or is there some room for debate? Two or more may have been deserving? Are all winners equal in terms of contributions to science? Are there other awards/designations that could also be relevant? And if so, how should they be considered/weighted?
@saillakeerie I’m confused about all your questions. We are not debating the criteria of being selected for a Nobel.
I’m sensing that your gripe with the rankings is due to some notion that there are certain intangibles that cannot be measured objectively, like “impact on science.” But in this study, “impact on science” is defined as a set of measurable quantities. You can say, who gets to define what is impact on science. But the study isn’t interested in providing a rationale for why they chose what they chose. The authors are following consensus and taking those honors deemed important by scientists and mathematicians as an objective measurement of impact to discern what schools have produced successful alumni at a high rate. And that’s what the study does: measure the rate that schools produce alumni who go on to be great scientists and mathematician. If you don’t believe that the Nobel is relevant for measuring impact on science, then you at least have to concede that your opinion isn’t shared by scientists.
@jademaster I am not debating the criteria for being selected for a Nobel either. Only the validity of the binary approach in evaluating what the results of that selection process mean in terms of impact on science/advances in science.
My issue with the approach is this. You take something subjective like the greatest impact on science/advances in science. There are any number of things that can be used to measure that. Some are objective. Others are subjective. Still others are a mix of both. Which factors you include and the relative weight you attach to each is subjective and can have significant impacts on results. Subjectively selecting which factors to include and the relative weight of each (even if each of the factors can be objectively determined) doesn’t produce an objective result even though you can produce a numerical sorted list of results (because if you had included different factors and/or weighted them differently, you get different results).
Do you think the rationale for what the authors of the study included and excluded is relevant?
Successful alumni. Great scientists and mathematicians. Do you believe either of those can be determined objectively? And if so, rates of success fullness or greatness can be objectively compared?
This discussion reminds me of a Calvin and Hobbes strip. Hobbes: What’s the point of attaching a number to everything you do? Calvin: If your numbers go up, you are having more fun.
Between perfect objectivity and voodoo, there’s a lot of space for reasonable judgement calls.
Which is more important, (a) the number of Nobel laureates on the faculty, or (b) the average class size?
In most cases I’d pick “b”, because in most cases “a” equals zero so it isn’t even a factor. But … if I could pick a physics department with multiple major prize winners on the faculty, then I’d probably be willing to sacrifice some class size. That kind of judgment may be reflected in the apparent preference of many top students for Ivies over top LACs. However, as you run down through the top N national universities, the number of faculty Nobels quickly drops to zero. So the next best choice may be a college with small classes often taught by professors a couple degrees of separation away from a major prize winner. That’s what you get at a lot of little LACs that may not present high numbers.
It’s useful to be able to check intuitions about these things against data.
That can’t be a 100% objective process, but it doesn’t need to be voodoo either.
@saillakeerie Like I said, any list you can conceivably come up with involves selection. Any weight accorded to what’s selected is subjective, but in this study, the criteria reflect the biases of scientists and mathematicians. So while things like “impact” and “greatness” are not measurable quantities, the important thing to remember is that there is a general consensus among specific groups about what “impact” and “greatness” are in the confines of those groups. Those concepts themselves may not be objective (because they’re qualities, not quantities), but the criteria used to measure them are. In science and math, the Nobel and Fields are considered honors for great scientists and mathematicians. You can count medals. You can count memberships in academies. This is the definition the study gave to “impact” and one that is a general consensus among scientists and mathematicians, but by no means the only way you can define a successful scientist or mathematician.
So while I agree with you that the list is subjective, I disagree with what to make of the criteria chosen. I do believe those criteria are suitable proxies for “impact.” Taken as what it is, the study shows certain schools churn out more impactful alumni in science and math as measured by honors given to scientists and mathematicians for impactful research. That’s the only thing it shows. What people then make of the list is their own interpretations.
Of course you can use a different set of criteria to measure “success” and “impact,” but it doesn’t render the findings of this study any less relevant. I agree with you that numbers are fun when they go up, and believe that the authors probably intended this list to be for bragging rights. I find nothing wrong with that. Caltech, a very small school, deserves bragging rights for producing so many prominent alumni in the sciences. Should this factor into where a student should attend college if they wanted to be a great scientist or mathematician? Probably not, since the vast majority of students won’t actually go on to be Nobel or Fields winners.
Another thing too is that a study like this doesn’t say anything about honors colleges at giant public research U’s, many of whom win prestigious national science awards/scholarships at impressive rates. MSU may not have alums enter the National Academy of Science at an impressive rate, but its (tiny) honors college may. But it would be impossible to tell from a study like this.
@jademaster I think Caltech desires bragging rights because it is a great school. Not because someone can create a numbered list which puts it at or near the top. And the reality of such lists is that people will use them to determine where to go to college.
@PurpleTitan It’s sad that honors colleges are often not considered an “honor” to attend by students. Many people forget that elite schools educate a very small portion of the country’s population. Arizona State alone has more undergraduates than all the ivies put together.
@saillakeerie How does anyone know what is a great school? Most of us are unlikely to have personal experiences or connections with more than a couple. What we know of other places is often second-hand, and for many, probably influenced by rankings. I’d personally be more worried about students using the US News to determine where to go to college. I don’t think this specific list is anywhere near that level of notoriety.
@jademaster, it depends on the student and how wise they are. Yes, may HS students are lemmings and want to reach for the bright shiny thing or do things for peer approval or just to bolster their self-confidence. That is to be expected at that age, really. Furthermore, not all honors colleges are the same. And the experience would not be the same as at an Ivy/equivalent RU or an elite LAC.
I wouldn’t recommend them for someone looking to get in to management consulting, as an example. For someone looking to get a PhD, especially in a STEM field (maybe those pursuing an MD as well), however, some may be terrific cheap options.
And yes, the American elites can only take a tiny percentage of all college-bound kids. As I pointed out in another thread:
"Another exacerbating circumstance is that the private elites here have relatively small undergrad populations while the US is a big country. I count 30 American unis and LAC’s as as Ivies/equivalents, but add up all the undergrad places in all those schools, and it is still smaller as a percentage of the population than the per capita places at Oxbridge+LSE+Imperial in the UK or UTokyo+Waseda+Keio in Japan.
Oxford, Cambridge, and UTokyo offer as many or more places than the biggest Ivy (Cornell). Imperial is equivalent to our Caltech but many times bigger. LSE has no true American equivalent, but you could say it’s close to Wharton, except it’s many times bigger than Wharton. Keio and Waseda are the size of our giant state schools. And both the UK and Japan are much smaller countries."
Plus Japan has the other 6 former Imperial universities.
That means that there really bright kids with high potential in the US at near-Ivies, honors colleges, etc.
A big problem with the qz.com table for major awards is that it is counting 3 very rare distinctions.
A school needs at least 3 awards even to be considered. Indeed, 6 of the 25 “top” schools have only 3 winners of the Nobel, Fields or Turing prize over the entire period 1966-2013. Even after adjusting for school size, very large schools have a big advantage in meeting that minimum. The CUNY system currently enrolls 516,000 students. The schools clustered in the bottom third are all large research universities.
The number of alumni-earned STEM PhDs is a much less rare distinction. On a per capita basis, Amherst and Swarthmore rank in the top 50 … but so do many other, similar LACs such Carleton and Haverford
(see table 4 in http://www.nsf.gov/statistics/infbrief/nsf13323/).
Most schools in the qz.com top 15 (for awards) also rank in the top 50 for per capita alumni-earned STEM PhDs. CUNY does not. Nor does any school ranked 18-25 in the qz.com “awards” table. By what other metric, plausibly related to undergraduate STEM education, does CUNY or Oregon State rank in the top 25?
Let’s remember what this list purports to be-not a list of the best schools, or even the best schools at which to study science, but the schools responsible for the greatest advances in science. For that metric I think using awards like the Nobel, Fields & Turing is reasonable.
Isn’t the qz.com table highlighting one specific CUNY college…City College of New York (CCNY), which currently have a total enrollment of about 15K and was a fairly small college when most of the Nobel Laureates attended the school (1933-1963 time frame)? CUNY-Hunter College had two Nobel Laureates and CUNY-Brooklyn College had one. CCNY had a total of ten Nobel Laureates.
FYI:
The Top Ten Schools with Oscar Winners
- USC (34)
- NYU (25)
- Cal (25)
- Columbia (18)
- U of Vienna (12)
- CUNY-CCNY (11)
- California IOA (10)
- Harvard (10)
- UCLA (9)
- Boston U (8)
^ all schools located in or near centers of movie-making (except Vienna which may be a hotbed, IDK). Not a surprise at all.