A new Top World U rankings

<p><a href="http://www.businessinsider.com/best-universities-in-the-world-2014-7"&gt;http://www.businessinsider.com/best-universities-in-the-world-2014-7&lt;/a&gt;&lt;/p>

<p>So the 3 big weights are the first 3. It looks like it’s not linearly weighted. If they just add up points, and the difference in points between 13th & 30th is far bigger than the difference in points between 65th and 151st or 49th and 105th (which is what I’d expect in a lognormal world), then I can see Northwestern outranking Duke. In a lognormal weighting system (or world), the best could have more points than all schools below 50 added together.</p>

<p>And indeed, it looks like whatever they’re using has a lognormal distribution, as Seoul (which kicks butt at producing CEO’s, placing 9th, but is terrible in the other 2 big categories, way in to the hundreds, way below a big chunk of the world), is slightly above UW-Madison (which is very steadily around 30th in all 3 big categories). </p>

<p>You’re right. I don’t agree with their approach at all.</p>

<p>Well, we live in a lognormal world. Harvard (all its schools) produces more CEOs than the bottom 200 colleges in the US added together.</p>

<p>I’ll let the math do the talking. </p>

<p>Let’s compare two American schools in accordance with the methodology which is publicly accessible on the organization’s website. </p>

<p><a href=“http://cwur.org/methodology/”>http://cwur.org/methodology/&lt;/a&gt;&lt;/p&gt;

<p>One of these schools is out performs the other in 7 out of the 8 sub-fields. But the other is ranked 3 spots higher. How is this possible? It isn’t. At least not mathematically. </p>

<p>Criteria 1 (weight - 25%): School A - ranks 65th
School B - 151st</p>

<p>Criteria 2 (weight - 25%): School A - ranks 30th
School B - ranks 13th</p>

<p>Criteria 3 (weight - 25%): A - 49th
B - 105th</p>

<p>Criteria 4 (weight - 5%): A - 22nd
B - 28th</p>

<p>Criteria 5 (weight 5%): A - 23rd
B - 32nd</p>

<p>Criteria 6 (weight 5%): A - 13th
B - 25th</p>

<p>Criteria 7 (weight 5%): A - 15th
B - 27th</p>

<p>Criteria 8 (weight 5%): A - 42nd
B - 61st</p>

<p>Weighted average rank: A - 41.75 (
B - 75.90!</p>

<p>B is ranked 3 spots higher than A despite the massive difference in the weighted average rank!</p>

<p>In case you’re curious, B is Northwestern. A is Duke.</p>

<p>These rankings are obviously flawed. Surely no one thinks that NYU should be ranked higher than northwestern or that Berkeley is better than a whole host of schools such as yale or princeton</p>

<p>The rankings measure what they measure. If you look at the criteria, a full 50% of the total score is determined by “major international awards, prizes, and medals relative to the university’s size” of either the faculty or the alumni. A bunch of the other criteria measure faculty quality as well.</p>

<p>By those criteria, it would not surprise me to know that Cal has about the same or more all-star faculty (and faculty that produces more influential/cited research papers) than Yale or Princeton.</p>

<p>You do realize that Cal is an absolute powerhouse in research, I imagine.</p>

<p>@purpletitan of course and thats why Berkeley’s grad programs are some of the best in the world but in terms of undergraduate no one in their right mind would argue that Berkeley is better than Princeton or Yale. In my opinion the best way to measure a university is by the strength of the students relative to other institutions and no one is really going to argue that on average, a princeton or yale student is smarter than a Berkeley student</p>

<p>@spuding102‌:</p>

<p>I prefer undergrad student outcomes (where Swarthmore/Amherst/Williams are up there with HYPSMCaltech).</p>

<p>In any case, this seems to be a ranking of universities in totality with a heavy emphasis on faculty prowess, not an undergraduate ranking.</p>

<p>The rankings aren’t flawed, they are what they are. Interpretation of the ranking may be flawed, or at least highly debatable, but at least the inputs are objective.</p>

<p>Peer assessment survey and high school counselors’ rating? Messy…</p>

<p>@PurpleTitan‌ A big difference in raw scores in one category should not be able to outweigh all the other categories combined. Specially when that one category is something as unpredictable and volatile as the number of alums who are CEOs of major corporations. Two Duke alums (Terry Myerson and Mark Reuss) recently narrowly missed out on the CEO positions at Microsoft and GM. They currently serve as integral members of the leadership team at these organizations. I don’t see how Duke would miraculously be a much better school if these ‘ill-fated’ alums had made the cut. Specially when you consider how many variables can impact decisions like the ones made my GM and Microsoft. For example, GM’s decision was clearly motivated by considerations that had nothing to do with individual merit (they were desperate for a female CEO). </p>

<p>@hidall1:
Yes, but Duke grads could have narrowly beaten out some candidates for other CEO positions as well (and some school could have narrowly won or lost a Fields medal). The rankings are what they are and measure what they purport to measure. Do I think that some of the criteria are absurd? Sure, but people will weigh the way that they want to weigh. Nobody says any one ranking has to be the end-all and be-all.</p>

<p>@YZamyatin: the biggest problem I have with the USNews ranking isn’t actually the surveys, but that so many of the numbers are self-reported, and USN evidently makes no effort to verify whether that are truthful/accurate or not. I read somewhere that one LAC (high up in that ranking) has been lying about their numbers for years. Plus, so many of those numbers are gamable. That’s why I prefer ranking by outcomes. Very hard to game whether your grads make CEO or win prestigious awards.</p>

<p>Of course you’re right! That’s one of the reasons why the number of CEOs that a university produces is a very imperfect measure of quality! Honestly, I can’t fathom why citations, h-index and broad impact aren’t accorded significantly more weight. They seem to be the best indicators of overall research excellence. They are also harder to manipulate. You can’t just hire a couple of superstars (Nobel laureates and the like) and change your score overnight!</p>

<p>Here’s a link to the methodology:
<a href=“http://cwur.org/methodology/preprint.pdf”>http://cwur.org/methodology/preprint.pdf&lt;/a&gt;&lt;/p&gt;

<p>The “alumni employment” factor is based solely on how many CEO’s of the Forbes Global 2000 have attended the school at any level, relative to the school’s size. I have a few problems with this. For one, it doesn’t take into account the many, many ways an alum can be successfully employed. Judges, senators, entrepreneurs who haven’t taken their companies public-they don’t count at all. This leads me to my second issue with this way of ranking alumni employment: schools with business schools will outstrip those without simply because successful CEO’s often receive MBAs after they have become successful. So a school like Yale, with no business school, will do poorly in the rankings, while Harvard will get a big boost. None of this reflects the value of an undergraduate degree from either school.</p>

<p>As an example, CIT (no B school) ranks 303rd in alumni employment, while NYU (Sterns B school) ranks 21st. Do we really think a student coming out of NYU is much more employable than one coming out of CIT? I find this doubtful.</p>

<p>Sue22:</p>

<p>Yep, I like the “American Leaders” subranking of the Forbes ranking better for measuring alumni success in the professional world.</p>

<p>Duke University, please move on. You don’t rank schools solely on the strength of the undergraduate student body. There are a lot of factors that measure the overall strength of the school. And, Duke not in the top 25 isn’t surprising to me at all. </p>

<p>You’re the one that needs to move on. Duke is ranked in the top 20 on a ‘meta university ranking’ that averages results from the 3 most credible world rankings. As a research institution, few schools can surpass Duke on a per capita basis. Duke is home to more highly cited researchers than all but 3 schools in the country. Our faculty is the 4th most productive in America and we receive more research dollars than any private university not named Johns Hopkins.</p>

<p>Stop being blinded by hatred. </p>

<p>Yale has a B School.</p>