We wouldn’t be CC if we didn’t re-hash the same old topics every few weeks 
3 Likes
I think - when someone busted tail and got into or attends a certain school that is now - not top 20 or top 50 - there’s a defense mechanism.
The school is the same as it was when it was top 20 in the latest app.
It hasn’t changed.
We seem to “believe” what others say.
Anyone can start and have a rank.
In the end, did your child learn, have a great experience and outstanding outcome?
2 Likes
Yeah, I feel like I always learn new things about specific schools when these things are discussed. Some good, some less good, some just interesting, but always seems worth it to me.
1 Like
My post said a “There is a loose correlation, but are also many exceptions.” Among the top 10 ranked colleges, the ones that have the most students from >=60th percentile were as follows. This list does not correlate well with the WSJ output of giving a boost to Amherst for good diversity; penalizing Babson, CMC, and MIT for poor diversity.; and leave the others alone. Instead I’d call this a loose correlation with many exceptions, like I originally stated.
Portion in 60th+ Percentile Income (Chetty)
Penn – 90%
Yale – 84%
Princeton – 83%
MIT – 83%
CMC – 83%
Stanford – 82%
Columbia – 80%
Amherst – 79%
Harvard – 79%
Babson – Lower Chetty reported median income than others on list, but % not available
As I mentioned from the start, I continue to think it would help clarify things if you looked at 60-99, and not 60-100, as that last top 1% is a real difference-maker.
Redoing your list for 60-99, and adding the two I started with, it looks like:
CMU 76%
MIT 75%
Hopkins 73%
Penn 70%
Princeton 68%
Yale/Columbia 67%
CMC/Stanford 65%
Harvard 64%
Amherst 62%
Some other notables that did somewhat poorly in the overall ranking:
UVA 76%
Chicago 72%
Rice 71%
Brown 71%
I’m the same. I think many folks on CC are well aware of the USNWR T20 (and T50, for that matter) and the T20 LACs. Frankly, those lists are more or less the same year in and year out. Any new list is interesting if it is different - it’s interesting to see some schools that you might not have considered before - that’s the fun aspect; not whether or not XYZ lesser known school is actually better than perennial favorites like Tufts and Middlebury.
3 Likes
I listed the full range 60th+ percentile range because using 60 to 98th percentile with top 1% excluded doesn’t make any sense as a measure of having good SES diversity. The WSJ rankings also do not have the ability to capture anything well correlated with top 60 to 98th percentile income, with the top 1% excluded, but the do have the ability to look at things that are reasonably correlated with (not) 60th+, such as % Pell grants.
That said, even if you only look at 60 to 98th percentile and exclude top 1%, it’s still a loose correlation with many exceptions, like I have stated repeatedly. For example, in your list CMC was better than most of the top 10, yet it received a penalty. Why penalize CMC and not others?
There are also many other exceptions besides just CMC. For example, one of the schools that received the highest graduation rate score in the entire WSJ list was Berkeley The only school I’m aware of with a better score than Berkeley is Princeton. Berkeley did not have an exceptionally high root 6-year grad rate like Princeton. Instead WSJ gave Berkeley a huge boost. Berkeley has a 92% grad rate, which the WSJ ranking boosted all the way to a score of 98,. Yet Berkeley had 67% in the 60 to 99th percentile, with top 1% excluded, putting it midway through the top 10 list above – same as Yale and Columbia, worse than CMC, Stanford, and Harvard. So why give Berkeley such a huge boost, but not give a comparable boost for the rest? I could list many other exceptions like this.
I was not suggesting one measure was better than any other. I was originally responding to another poster who was expressing a confusion over whether the WSJ was really using the sort of statistical model it claimed to be using. I was pointing out that this sort of admittedly crude validity check supported the conclusion that “it looks like the WSJ is using the approach they said they were using.”
It appears that the WSJ is using graduate income data that is restricted to students who received federal aid, which would apparently include both Pell Grants and federal loans. Federal loans have a broader demographic than Pell Grants, but as I understand it, they are naturally very uncommon among the sorts of families who would be full pay, and that includes most if not all of the top 1%.
So it is not a matter of excluding the top 1% being a better measure of SES diversity. It is just a matter of understanding that the data being used by WSJ is very likely largely if not entirely excluding the top 1%.
And that in turn explains why people who were used to thinking of SES diversity in light of top 1% percentages might be “surprised” to see colleges like MIT, CMU, JHU, and so on not doing better. Once you understand that outside of the top 1%, these colleges are actually less, not more, diverse by income than many other highly-selective colleges, that helps explain why these WSJ results are really not that surprising.
But I am in no way suggesting the WSJ is doing things the best possible way. Again, I was just pointing out to this other poster they very likely were doing what they said they were doing.
So it is absolutely true that what we are seeing here is not a perfect correlation in the sense that however these colleges were ranked in the NYT by this measure, they were ranked in the exact same order by the WSJ’s model.
Of course that is not something I suggested we would find, and it does not make sense for many reasons. For one thing, the WSJ never claimed parental income was the only factor in their demographic model (or a factor at all). For another, that NYT data is old, and according to the WSJ, they got their demographic information from 2021 as per Ipeds and the College Scorecard.
So I definitely was not suggesting we should expect a perfect correlation.
But still, I think this is one of those situations where people have an opportunity to actually learn something new and interesting–certainly I did. I knew about MIT, JHU, CMU, and so on being relatively low in top 1% percentages as of the data dates in question. I actually didn’t know until looking it up that they all tracked higher in the 60-99 range, and lower in the 0-60 range, than many other highly selective colleges.
And personally, I am skeptical about trying to reduce socioeconomic diversity into a single number. So, to me, it is interesting to know all that information.
And then the WSJ ranking itself–eh, I am probably not going to be doing anything in particular with that.
I understand your point completely, and while it’s not a fight I’m inspired to fight, I certainly don’t want to impede the efforts of those who deem it worth while. My comment about advertising was a general reference to my personal view, and it is my personal view, that I am ultimately responsible for my choices and that of my children while they are my children. We’ve been bombarded with the same images, commercials and other means of persuasion to overeat, overspend, wear certain clothes, drive certain cars, and attend certain schools. I’m sure we’ve been influenced on the margin here and there, and perhaps more than I appreciate.
But on this topic, we were resourceful enough to do some basic due diligence on our own and I don’t feel like we were driven to make choices that were not in our best interests because of the multitude of rankings. In fact, I would say the existence of the multitude, and their sometimes wildly varying results, has helped us put those rankings into proper perspective. If there were only one, and everybody said “that’s the right one,” then I suspect it would have been much more powerful as an agent of suggestion. If there are legions of more vulnerable people who need protection from these crazy rankings, then by all means the critics can carry on with my full support.
The overwhelming majority of architecture at UW is comprised of collegiate gothic along with a healthy sampling of French renaissance revival, romanesque revival and Beaux-Arts revival. There are some post-modern and brutalist buildings, and of course all the new stuff looks like everybody else’s new stuff:
But the lion’s share of campus architecture falls into one form or another of classical campus architecture, although the new stuff (see above) is really starting to add up.
So have I. I don’t think they do either.
1 Like
I wholeheartedly believe this to be true. That general tendency applies to other things as well, like what we majored in or chose to do professionally.
1 Like
I have never been to Moscow. It just makes me think of the Russians…it’s was a symbol of strength, of communism growing up. So I am pre programmed to that. I never heard of the one at UW til I visited - but also I thought, not the view, but the aesthetic wasn’t the nicest.
But - we all think different things about different colleges. My son liked the campus fine, as an example. But I suspect decided it was a bit far from home - not really sure why he didn’t apply.
Thanks
I was talking about graduation score, rather than income. However, 99th percentile (top 1%) is not the only group that does not have a high rate of Pell grants and federal loans. For example, I also wouldn’t expect top 2% income to have a high rate of FA or top 3% or … Furthermore, while the CollegeScorecard stats only include federal FA recipients, they do not specify parental income directly, so it would not be a source for figuring out portion 60 to 99th percentile income. Instead they specify things that may be correlated with income, such as % Pell, % of students receiving federal loans, or % of students receiving FA in general
None of these stats or other CollegeScorecard available stats mesh well with the adjustments. As I noted earlier % Pell is reasonably correlated with portion not in 60th+ percentile, but doesn’t exclude top 1%. It only has a weak correlation, as discussed earlier. % federal loans extends higher than 60th+ percentile and might exclude top few %, but % federal loans also doesn’t mesh well with the WSJ adjustments. For example, example federal loan rates by college is below.
Princeton – 2% (possibly lowest of discussed colleges, but no penalty)
MIT – 7% (has penalty)
Amherst – 10% (gets boost)
Babson – 32% (has penalty)
I don’t doubt that WSJ is using a formula based on real stats to make adjustments to their graduation rate and other stats, rather than just making random adjustments. However, that adjustment formula only has a weak correlat8ion with SES diversity and/or racial diversity, with many exceptions. WSJ also does not specify how they are making adjustments, so this discussion is purely a matter of speculation. Nobody reading the rankings knows exactly what WSJ is doing in calculating the rankings, which is a problem.
Sure. If we had more granular data by school by income percentile, and then federal loans by income percentile, we could fold it all together into some sort of weighted average.
But since this was just illustrating a basic point, and not actually trying to reduplicate what the WSJ did, I am not personally inclined to take it farther.
Great, then we can agree to agree. Because that was the only proposition I was intending to support.
That said, in the process I learned something new about MIT, CMU, JHU, and so on as of that study period, so that’s cool too.
It’s sometimes difficult for people like you, me, and all of the regulars on CC, all who believe in due diligence for any major decision, to fully realize how uncommon it is. The only reason that I realize it is because I’ve taught students and have come to realize that many smart people don’t realize that research saves a lot of money, unneeded effort, and heartbreak.
It’s not that difficult to help people who are bad at that, at least the ones you know or meet casually. Just tell your story of how much you and your kid (or kids) were helped by doing some research on your own. Have them stop by here, even. If set even a handful of families on track to find colleges that are the best for their kids, academically, socially, and financially, you’ve already made the world a better place.
You’re already providing people with advice here on CC.
6 Likes