Yup, and by definition that equals GIGO. (As every student in AP stats knows)
2 Likes
And as every member of my weight-watchers group knows.
The gap between “I stayed within my points balance!” self-reporting and ACTUAL results as measured on a digital scale… Self reporting is a very optimistic activity!!!
3 Likes
- Did this take differences in COL into account?
- Did this take popularity of different majors into account? (self-selection)
Perhaps more accurate. At least more interesting:
Northwestern’s starting salary for the typical 22 year old graduate of its undergraduate School of Education & Social Policy is even greater at $79,765 than for its arts & sciences graduates. (Most regard Northwestern’s SESP as the best undergraduate school in the world for this area of study.)
Still horrified by the poor editing of Poets & Quants For Undergraduates article which is the subject of this thread with respect to Northwestern University. Does anyone really believe that Northwestern University’s undergraduate tuition is over $21,000 per year higher than that for Columbia University ? Of course not (except for P&Q’s sloppy editors).
P&Q For Undergraduates doesn’t even have the decency to correct this article which compares Northwestern’s total COA (tuition, fees, room & board) to other school’s tuition only.
If P&Q For Undergraduates used the correct figure for Northwestern University’s undergraduate tuition of $67,128 per year (instead of the incorrect figure of $89,544), Northwestern would be among the schools with the best ROI, not the worst.
I’ve not heard back from the author. I wonder if they “edit” or simply took the data from their source - which is a European company I believe (Edumentors).
It’s interesting how many put them now at the forefront of b school rankings - yet after hearing from you about sloppiness, maybe they shouldn’t be that source.
1 Like
It’s also unclear what schools they are even considering for this list. It refers to “top colleges”. So are they only looking at “top 20” schools? Top 50? 100? The whole article is garbage.
2 Likes
P&Q does provide a lot of useful information, but impactful mistakes are not uncommon when they do rankings. Usually, it is corrected after reader comments point out the mistake. Corrections are usually done by the website owner and acknowledged in the comments section on the MBA site.
FWIW There are two P&Q websites–one for undergraduate business schools and one for graduate business schools (MBA).
2 Likes
It’s a small sample size, but no one in my son’s math cohort from his college at Oxford took a job paying 30K GBP after graduation, including the person who went to work in the public sector (the FCA).
2 Likes
This doesn’t appear to justify a separate topic, but Poets & Quants also did a recent ranking of colleges with the highest (reported) crime rates:
One of the more conspicuous methodological shortcomings of the analysis may be the small size of the pool considered — just 30 schools.
There are at least a dozen methodological problems here…it could be a B-school case on “bad statistics make for terrible decisions”.
It does not distinguish between a dorm party where someone spiked the punch and an actual violent attack.
I’ll let the fine posters on CC continue the list… but will note that colleges which encourage reporting crimes to law enforcement always look worse than significantly more dangerous campuses where a Dean or other administrator sits down with the victim to suggest “do you really want to be known on campus as the freshman who was raped?” Why not take a few days to think about it before calling the police ". At which point…the physical evidence is gone and it’s a "he said she said ".
Why post something so misleading and erroneous? Bowdoin and Swarthmore, really?
Because the second site appears to support the questionable reliability of Poets & Quants, as was discussed here regarding the site of the primary topic. Also, I’m not positive that the second site is erroneous. I haven’t considered it particularly deeply. However, I wouldn’t judge it by the schools it names — statistical results often do not align with intuition.
1 Like
Tell the FBI that their metrics are based on intuition. That will be a fun discussion.
The concept of determining ROI without considering the individual student is inherently flawed. If a particular student goes to goes to a highly selective college, they don’t suddenly become the average student at that college who is likely gifted, high achieving, highly motivated and well connected; just as if the student instead saves money by getting a full ride scholarship to a less selective college they don’t suddenly become the average student at that college who is likely less high achieving and may struggle to graduate. Similarly majors and planned career paths can have dramatic impacts on earnings. If a college has higher median earnings because it has more tech majors that doesn’t help the kid who isn’t planning to be a tech major.
The denominator of the ROI equation also depends on the individual student. At highly selective private colleges in the US, a good portion of students have $0 expected cost to parents, and a good portion of students pay >$80k/year. Whether the student pays $0 or pays >$80k per year impacts ROI.
While nearly any estimate of ROI that does not consider the individual student has little value when estimating ROI for a particular student, this ROI ranking is especially flawed, perhaps the worst one I have ever seen.
Only 10 colleges are listed, with only 4 colleges in the US. With only 4 colleges, a student cannot compare stats for the college they are considering. Is the idea supposed to be that these 4 colleges have the worst ROI among all 2000+ colleges in the US? If not, how many colleges are considered? Regardless of how many are considered, I very much doubt the listed colleges are going to top a legitimate worst average ROI type ranking.
More importantly, the listed stats for those 4 colleges make little sense.
“Average tuition” implies after FA, yet Northwestern’s $90k price tag implies no FA. However, Cornell and Columbia have a much lower $68k price tag, so is it after FA for Cornell/Columbia; but not for Northwestern? No methodology details are listed, so one can only guess.
Among the listed US colleges, Berkeley is listed as having the lowest cost of living. Silicon Valley is not exactly known for having a low cost of living, so that doesn’t make sense. No methodology details are listed, so I can only guess as to how they computed that particular figure for cost of living.
Perhaps the most important figure is the earnings. It’s phrased as “median annual starting salaries.” Northwestern is highest at $61k and Berkeley is lowest at $48k. Other sources list completely different values and relative rankings. For example, CollegeScorecard has Northwestern as lowest median earnings, rather than highest. Regardless of what source you choose, the specific median salaries are completely different than the the ones that are listed. No methodology details are listed, so unclear how this is computed.
The ranking is completely useless.
3 Likes
P&Q goofed up. Listed Northwestern’s total COA (tuition, fees, room & board) and compared it to other schools tuition only.
This point is 100% accurate. The P&Q ranking is more than pure garbage, it is intentionally misleading as no reputable website would miss such a glaring error (at least not without an ulterior motive).
1 Like