<p>here's my analysis:</p>
<p>top 10 - no way
top 20 - no way
moving up in top 30 - may move up 2-3 spots
moving down in top 30 - may move down 2-3 spots</p>
<p>so, I'm giving it 50-50 to move up or down 2-3 spots. Please note, the idea of this thread is not to analyze what the ranking should be, just trying to figure out where the ranking will fall based on what we know of the us news ranking methodology.</p>
<p>I don’t think Michigan will move much. It may go up or down one spot, but that’s about it. Unless the USNWR formula changes significantly, Michigan will remain in the 28-30 range for some time.</p>
<p>yes, at first blush I thought the dip in acceptance rate would help but after looking at it, I don’t think it’s going to be meaningful for the rating.</p>
<p>Acceptance rate and selectivity play a small role in the overall rankings. The main issue is the way the USNWR compiles data (non-standardized, and not adjusted or audited for accuracy or consistancy), and the weight it assigns to several factors that are not necessarily indicative of quality of education, especially when you consider the differences in budgetary and fiscal matters between private and public universities. The most blatant of those are the class size and student to faculty ratios, where some universities include hundreds, if not thousands, of mandatory freshmen seminars and conveniently omit thousands of graduate students from their ratios. Properly adjusted to reflect reality, there would be virtually no difference among the top universities, and the better public universities would be ranked significantly higher.</p>
<p>I don’t think Michigan will move much, if at all, but there could be some small shake-ups elsewhere in the rankings. The main change in methodology affecting “national universities” this year is somewhat greater weight placed on “graduation rate performance,” i.e., the degree to which the school’s actual 6-year graduation rate exceeds or falls short of a US News-calculated “predicted graduation rate” for that school, based on things like percentage of Pell grant recipients and SAT/ACT scores. (The idea is that high-income students and students with high test scores can be expected to graduate at higher rates than lower-income, lower-test-score students, so a school’s actual graduation rate should be measured against a predicted rate based on those kinds of factors).</p>
<p>That category will still be a small part of the overall ranking, but it could break some current ties or possibly move some schools up or down one place, depending on the rounding. (Each school is assigned a “raw score” expressed in whole numbers, based on aggregating all the various factors on which they’re measured, so Michigan’s raw score of 74 is currently 29th highest, just below Tufts’ 75 and just above UNC-Chapel Hill’s 73; but Michigan’s 74 could actually be 74.45, rounded to 74, and Tufts’ 75 could actually be 74.55, rounded to 75, so that even tiny changes could tip the schools into a tie or even leapfrog Michigan over Tufts).</p>
<p>One likely loser is Caltech, whose graduation rate underperforms its predicted rate by a whopping 7 points. Caltech is currently tied with Dartmouth (raw score = 92, good for 10th place). Dartmouth is +2 in overperforming its predicted graduation rate. So if nothing else changes for either school, it seems the tie could be broken and Dartmouth could edge ahead of Caltech. Some other schools that could be hurt a little by the increased weight attached to this metric are MIT (-3), WUSTL (-3), Rice (-2), Carnegie Mellon (-2), USC (-2), and Tufts (-2). Schools that could be helped a little are UCLA (+3), UVA (+4), and UNC-Chapel Hill (+5).</p>
<p>Michigan (+1) won’t be affected much in its raw score, but it could move up by virtue of Tufts (-2) moving down or by UNC-Chapel Hill (+5) moving up. But I don’t see Michigan being anything other than 28th, 29th, or 30th, and there could be no change at all.</p>
<p>USNews posted a “preview” of the top 25 and Michigan was not on the list. And the top 25 is actually about 27 schools due to ties. New results out tomorrow. It doesn’t look like there are any real changes.</p>
<p>Here’s preview:
[2014</a> Best Colleges Preview: Top 25 National Universities](<a href=“2014 Best Colleges Preview: Top 25 National Universities”>2014 Best Colleges Preview: Top 25 National Universities)</p>
<p>28th - tied with Tufts.</p>
<p>“The most blatant of those are the class size and student to faculty ratios,”</p>
<p>I think those may be runner up, I would assign most blatant to the following. The magazine models/predicts graduation rates. Those above the line get credited some factor and those below the line debited. In other words, models are inherently noisy and inherently imperfect predictors, so what is the magazine’s approach: they assign the model error to the various schools. That to me is, pun intended, rank stupidity and gets my award for most blatant.</p>