New US News Rankings

[DiscoTech]

US News has put out their annual monkey-throwing-feces-at-a-dartboard rankings. Some schools game the rankings well, while others not so much.

Here's my Q: in your opinion, what is the most overrated program? Why?

For many years going I have been in awe of USC for being able to scale the rankings despite having a decidedly mediocre program. Here's why:

(1) They do a great job of massaging the numbers they send US News. Each school reports its funding totals to the National Science Foundation. Unlike what they send US News, NSF gets the true numbers. No double counting funding that goes to the ISI or the physics department and whatnot. USC clocks in at a whopping #42 with only $65M in externally financed research (https://ncsesdata.nsf.gov/herd/2016/html/HERD2016_DST_55.html). 

(2) Their peer and recruiter ratings (3.7ish) are really low compared to their overall ranking. Now that is just damn impressive. Their reputation should put them at #20 to #30 school.

(3) They have ridden their number fudging to build a great cash cow - foreign students fork over $50k for the honor of having USC name on their diploma. $50k for what basically amounts to a 5th year of undergraduate work, but with lax grading. This really is an impressive feat. A program with middle-of-the-road funding numbers builds a diploma mill, which probably generates $50M-100M in revenue. A good chunk of this money likely funds research at USC. MS students are probably the largest source of funding for research at USC! Give them some credit.

 

* On the flip side, UCSB and Harvard always struck me as underrated by US News' metrics.

 

On a more serious note, between ASEE data, NSF data, and citation metrics shouldn't is be possible to put together are more useful comparison of programs? The NRC used do something along these lines back in the day. Oh well. At least it is fairly straightforward, with a little digging, if a school's business model is built around a diploma mill.

 

[whitbee]

wow, that's a fascinating spreadsheet! where does one find a school's peer and recruiter rankings?

[DiscoTech]

1 hour ago, whitbee said:

wow, that's a fascinating spreadsheet! where does one find a school's peer and recruiter rankings?

You might have shell out for a copy of US News' rankings. Or find a bootleg copy on the web (you should be able to find last year's without too much work).

[Man_About_Town]

On 3/20/2018 at 5:48 AM, DiscoTech said:

US News has put out their annual monkey-throwing-feces-at-a-dartboard rankings. Some schools game the rankings well, while others not so much.

Here's my Q: in your opinion, what is the most overrated program? Why?

For many years going I have been in awe of USC for being able to scale the rankings despite having a decidedly mediocre program. Here's why:

(1) They do a great job of massaging the numbers they send US News. Each school reports its funding totals to the National Science Foundation. Unlike what they send US News, NSF gets the true numbers. No double counting funding that goes to the ISI or the physics department and whatnot. USC clocks in at a whopping #42 with only $65M in externally financed research (https://ncsesdata.nsf.gov/herd/2016/html/HERD2016_DST_55.html). 

(2) Their peer and recruiter ratings (3.7ish) are really low compared to their overall ranking. Now that is just damn impressive. Their reputation should put them at #20 to #30 school.

(3) They have ridden their number fudging to build a great cash cow - foreign students fork over $50k for the honor of having USC name on their diploma. $50k for what basically amounts to a 5th year of undergraduate work, but with lax grading. This really is an impressive feat. A program with middle-of-the-road funding numbers builds a diploma mill, which probably generates $50M-100M in revenue. A good chunk of this money likely funds research at USC. MS students are probably the largest source of funding for research at USC! Give them some credit.

 

* On the flip side, UCSB and Harvard always struck me as underrated by US News' metrics.

 

On a more serious note, between ASEE data, NSF data, and citation metrics shouldn't is be possible to put together are more useful comparison of programs? The NRC used do something along these lines back in the day. Oh well. At least it is fairly straightforward, with a little digging, if a school's business model is built around a diploma mill.

 

This is sadly not limited to USC.  Every single research University in the U.S. is following this model. Harvard has been doing this for years via their "Executive Education" program: https://www.exed.hbs.edu/programs/Pages/program-finder.aspx . Look at those tuition prices for week-long "courses" that give you nothing more than a certificate, not a degree.  All schools use their Masters programs as a cash-mill, but USC simply does this better than other schools.  This will only continue to expand with distance-learning.  Look at how many schools have created online Data Science Masters programs overnight with 50K+ tuition.

In the era of decreased, more competitive research funding, these schools are being industrious.  All that aside, I think there is a strong argument that depending on what program you are pursuing in engineering, USC is on-par with UCLA and Caltech.

Disclosure of conflict of interest: I recently interviewed at Caltech, UCLA, and USC for PhD programs and am likely going to choose USC as it has the best lab-fit for me.

[DiscoTech]

17 hours ago, Man_About_Town said:

This is sadly not limited to USC.  Every single research University in the U.S. is following this model. Harvard has been doing this for years via their "Executive Education" program: https://www.exed.hbs.edu/programs/Pages/program-finder.aspx . Look at those tuition prices for week-long "courses" that give you nothing more than a certificate, not a degree.  All schools use their Masters programs as a cash-mill, but USC simply does this better than other schools.  This will only continue to expand with distance-learning.  Look at how many schools have created online Data Science Masters programs overnight with 50K+ tuition.

My snarky tone aside, I do not have a problem with what USC is doing. It is actually impressive given that they are diluting the reputation of their M.S. degrees. The one tangible downside, if you are a doctoral student at USC, is that your classes are stuffed with these folks who are just hunting for a credential. I don't know how this affects instruction or grading, but the effect can't be zero. BTW, what USC does is clearly different from what HBS does. HBS isn't giving out online HBS MBAs or watered down MBAs. The folks you reference are getting certificates, which no one will confuse for an MBA from the world's preeminent business school. With regard to the data science degrees, there is no mistaking them from graduate degrees from the respective math or engineering departments.

To be clear, I do not mean to suggest that USC is alone in doing this. Columbia (probably even more so than USC) and Penn are in on the game. Cornell plays it cute by calling their coursework-based Masters an M.Eng.

 

17 hours ago, Man_About_Town said:

In the era of decreased, more competitive research funding, these schools are being industrious. 

Any idea why MIT and Caltech aren't being "industrious?"

 

17 hours ago, Man_About_Town said:

All that aside, I think there is a strong argument that depending on what program you are pursuing in engineering, USC is on-par with UCLA and Caltech.

The phrase "depending on what program" is doing a lot of work here. With the exception of maybe bioengineering, Caltech is by far the better program.

 

So all of this is a distraction from my actual point - USC standing in the overall engineering rankings is likely inflated relative to its actual merit. I gave you my two reasons - (a) external funding for engineering research (#42 in the country), and (b) academic and recruiter ratings on par with schools ranked 20-30. (b) is certainly the weaker part of the argument. Do these reputation scores actually correlate with program quality? The clearest counterpoint is UC - Santa Barbara, which has middling reputation scores, but is probably one of the best 2-3 schools in the country for materials science, physics, and electrical engineering. This is validated by number of graduates they place into faculty positions at top schools. Maybe this is true of USC in fields that are not materials science, physics, or electrical engineering. I don't know. I do think the ability to attract external funding for engineering research is a good barometer of program quality. Clocking in at #42 on this measure suggests that USC is overrated by US News' overall engineering rankings. People funding the research clearly aren't as enamored.

BTW, none of this is to say USC isn't a great program. There are world famous faculty there and I would be proud to go there. Good programs can be overrated relative to their peers. I do genuinely think what USC has accomplished with the rankings is impressive.

 

[Man_About_Town]

12 hours ago, DiscoTech said:

Any idea why MIT and Caltech aren't being "industrious?"

The phrase "depending on what program" is doing a lot of work here. With the exception of maybe bioengineering, Caltech is by far the better program.

 

When I interviewed at Caltech, the current students and faculty were constantly talking about the huge donation that they just got from a "Chinese Billionaire".  The actual term Chinese Billionaire was used.  So either way, the money is coming from overseas at all these schools.  My reason as to why Caltech is not creating a degree cash-cow is that the tiny size of the school simply could not sustain it.  Or being in the Billionaire's favor makes it unnecessary.

And you nailed it. My interests are in bioengineering.  I think depending on your sub-interest within BME, your best lab could be at Caltech, UCLA, or USC.