US News 2018 Ranking Predictions

@pupflier wrote:

Yes, and that includes doing everything possible to raise your sticker price while complaining how hard it is to recruit First Gens and Pell Grant recipients.

My prediction is that the rankings will continue their fascination with ROI, based solely on self-reported earnings and a very limited time horizon (5 to 10 years). To this extent, I think @Chembiodad’s comments reinforce some of the frustrations around ranking students who come out and make $80k/year higher than those who go to med school and earn double+ more than those students for the final 25-30 years of their careers. The schools that generate Phd’s and Med school students keep “slipping” in the rankings.

Has anyone taken a close look at the predicted graduation rates for US News (accounts for 7.5% of the ranking)? They seem to be arbitrary and inconsistent from year to year. My opinion is that US News uses the predicted graduation to maintain the pecking order and to move schools up or down at its discretion. So, to increase a school’s ranking, US News can assign it a lower predicted graduation rate and alternatively, to decrease a school’s ranking, US News can assign it a higher predicted graduation rate. Let’s consider 2 examples: UChicago and JHU. By giving UChicago and JHU lower than expected predicted graduation rates (relative to peers), UChicago and JHU are able to do better in this category than they should. I will show some comparisons from last year’s rankings.

Let’s compare UChicago to Cal Tech / MIT. In last year’s rankings, UChicago had a predicted graduation rate of 94% but yet it’s selectivity is ranked #1. Now, let’s compare that with Cal Tech - selectivity is ranked #1 (tied with UChicago) and the predicted graduation rate is 98% (compared with 94% with UChicago). Based on the student body strength of UChicago, it’s predicted graduation rate should mirror Cal Tech’s and be closer to 98%.

Assigning UChicago predicted graduation rate to be 94% gives UChicago a definite (and unfair) advantage in the rankings because it is receiving a nice boost in the 6 year graduation rate performance category or in other words, its ‘under-performance’ with respect to the 6 year graduation rate is not being captured fully. If UChicago’s predicted graduation rate was closer to 97% or 98%, it’s under-performance would be -4 or -5, as opposed to -1 or -2. This could easily push UChicago back to #6 or #7 in the total rankings.

And another example - MIT’s selectivity rank is 3 and its predicted graduation rate is 97%. Again, UChicago should be closer to 97%-98% for the predicted graduation rate, based on the main inputs (selectivity, student body strength, etc.), similar to MIT and Cal Tech.

Let’s also consider JHU - selectivity rank of 10 with a predicted graduation rate of 93%. Cornell, which has a selectivity rank of 20 (significant lower) has the same predicted graduation rate of 93%. This seems very inconsistent. Another way of looking at it, compared to Duke (with a selectivity rank of 13), it has a predicted graduation rate of 95%. This suggests to me that JHU’s predicted graduation rate should be closer to 94% or 95% and that the current predicted rate of 93% is giving JHU an unfair advantage in the rankings.

Another inconsistency --why does Northwestern have a higher predicted graduation rate (96%) than Stanford (95%), Duke (95%), Dartmouth (95%), Brown (95%)? Alternatively, why do Columbia and Penn have predicted rates of 97% compared to the peer group? How does this make any sense based on the proposed methodology about predicted graduation rates?

Curious what @Alexandre and @Penn95 think?

Here’s another recent opinion: QS World University Rankings. These U.S. schools were in their top 20. 1. MIT 2.Stanford 3. Harvard 4. CIT 9. Chicago 13. Princeton 14. Cornell 16. Yale 17. Johns Hopkins 18. Columbia 19. Penn

It will be interesting to see how it compares with U.S. News. They have to make changes every year, or who would be the magazine?

@StanfordGSB00

This is pretty basic stuff. Predicted graduation rate is based on entering class stats from 6 years ago. Only recently has Chicago and JHU caught up in consistently high enrolled SAT scores (which is the primary criterion this predicted graduation rate is based on). It’s not related to current selectivity rank (which takes into account current enrolled freshman class). There’s no funny business here. Chicago and JHU will both have a higher expected graduation rate soon.

@StanfordGSB00

See more here:

https://www.usnews.com/education/best-colleges/articles/how-us-news-calculated-the-rankings

Graduation rate performance (7.5 percent): This indicator of added value shows the effect of the college’s programs and policies on the graduation rate of students after controlling for spending and student characteristics, such as test scores and the proportion receiving Pell Grants. U.S. News measures the difference between a school’s six-year graduation rate for the class that entered in 2009 and the rate U.S. News had predicted for the class.

  1. So are you saying that USNWR is colluding with UChicago , JHU and others to keep their ranking higher?
  2. "This could easily push UChicago back to #6 or #7 in the total rankings." Could you explain (mathematically) how a percentage of a percentage could move a University's ranking that much or is this just an assumption with no basis?

@stevensPR - I beg to differ. Please look at selectivity ranks from 6-7 years ago. Northwestern didn’t have a higher selectivity rank than Stanford 6 years ago nor did it have a higher selectivity rank than UChicago or Brown.

@StanfordGSB00

How about you pull the data? Predicted graduation rates are based on test scores. Northwestern’s test scores are or had been higher than Stanford, Brown, Duke and many other privates for a while now. Predicted graduation rate is largely based on test scores, not necessarily selectivity rank (which is derived from acceptance rate as well).

@CU123 - everything I wrote is my opinion and I have no idea how the under-performance vs. over-performance affects the 7.5% total. If a school scores -2 vs. -1 vs. +2, there are many ways that US News could calculate the numerical value a school receives, as % of the 7.5% total. I outlined below one possible way the calculation works and I welcome other views on this metric. My assumption is that once scores are scaled, 7.5 is the maximum number of points a school can receive as the total maximum score (from all categories) is 100 points (i.e., Princeton last year).

Over(+)/Under(-) | Total Score out of 7.5 Points
-4 | 3.5
-3 | 4.0
-2 | 4.5
-1 | 5.0
0 | 5.5
1 | 6.0
2 | 6.5
3 | 7.0
4 | 7.5

As you can see from the example above, if my methodology was indeed similar to US News’, a school like UChicago which received -2 last year would have lost an additional 1.5 points on its total had its predicted graduation rate been closer to 97% (like Cal Tech). That would have pushed UChicago to #4 or #5 last year. Again, this is my opinion and I have no idea how the 7.5% is actually being calculated. I welcome opinions.

OK so a lot of assumptions here.

You do realize that Stanford (who I assume is your alma mater) will never move into the top 3 due to its Division 1 athletic programs…and there is nothing wrong with that.

@StanfordGSB00

here’s data from the graduating class of 2014 (high school class of 2010) for Math + Verbal SAT Ranges:

Northwestern SAT: 1380 to 1530
Brown: 1330 to 1530
Stanford: 1360 to 1550

http://enrollment.northwestern.edu/pdf/common-data/2010-11.pdf

https://www.brown.edu/about/administration/institutional-research/sites/brown.edu.about.administration.institutional-research/files/uploads/CDS_2010-11_Brown.pdf

https://ucomm.stanford.edu/cds/2010

You’re also completely failing to account for pell grant distribution in freshman enrollment from this period as well.

where do you guys think vanderbilt will place?

@CU123 He’s a GSB grad i presume, so not sure why he’d care so much about Stanford’s undergrad ranking.

I know for a fact Duke and JHU will increase in rankings due to higher enrolled freshman profiles and enhanced class sizes through noticeable increases in faculty hiring.

@stevensPR, unless you know something that the general public doesn’t know, I am quite confident that Duke’s selectivity rank will not increase beyond where it is now. And that is mainly b/c of the fact that US News uses the average ACT and SAT Scores, plus the top 10% and acceptance rates. And Duke’s SAT average hasn’t changed materially nor has its ACT composite average. Looking at Duke’s common data set, its ACT range is 31 - 34 (same as last year). Last year, Duke, Northwestern and Dartmouth were tied at 13 for selectivity rank. And given the latest common data sets compared to last year, I can see Northwestern increasing in the selectivity rank because its acceptance rate went down and its ACT range went up from (31 to 34, to 32 to 34) while the SAT range stayed constant. An increase in the composite ACT average (which accounts for more than 50% of the student body) from 32 to 33 would likely bump up Northwestern’s selectivity rank relative to Duke and Dartmouth. It’s unclear what Duke’s composite ACT average is but if I were to guess, I would say it is a 32. And US News apparently converts the averages to percentiles to calculate the selectivity for test scores (weighted for both ACT and SAT averages).

@StanfordGSB00

Go to collegeboard and look at duke’s sat scores for enrolled freshman for the old SAT. It is now 1450 to 1570 - which is definitely a material difference. These scores are what the rankings for this year will be based off of.

@stevensPR, the collegeboard statistics are incorrect and there is a post somewhere on CC that talks about this. It’s the common data set numbers. https://finance.provost.duke.edu/common-data-set US News consults the common data set numbers and then the schools. Collegeboard.com is not used. If you look at last year’s ranking data, all of the numbers will match perfectly from Duke’s common data set (2015-2016), and this year’s numbers will match Duke’s 2016-2017.