Data to support which of my claims?
This claim. Again, Iâm only aware of data that supports this claim for disadvantage and/or low-income students.
Re correlation, there are lots of studies demonstrating the link b/w college reputation and future salary.
These graphs from the Economist provide a good visual of the relationship (broken down by major).
A WSJ article (from 2016) review literature and concludes that it matters for certain fields (humanities and business) but not others (STEM).
The starkest earnings differences are for business majors, where graduates from the selective institutions earn 12% more on average than midtier graduates and 18% more than graduates from less-selective colleges. Likewise, social-science majors from selective colleges earn 11% more than their midtier counterparts and 14% more than those from less-selective schools.
And this 2016 article in Higher Ed summarizes a study that teases out causation from simple correlation.
The authors are Paul Attewell, a professor of urban education at the City University of New York Graduate Center, and Dirk Witteveen, a doctoral student in sociology there.
âŠ
They then tried to control for all factors (other than the college experience itself) that might make some students more likely to earn more after graduation. They controlled for gender, age, race, parental income, parental education, SAT, college grade point average, college major, private/public institution, advanced degree after graduation and region of employment after college.
The results show that, after controlling for all those factors, graduates of the most competitive colleges earn (10 years after graduation) 8 percent more than graduates of very selective (but not the most competitive) colleges, 11 percent more than graduates of colleges that are competitive, and 19 percent more than graduates of colleges that are not competitive in admissions.
3 Likes
Oh look. It turned into a âIs it worth itâ discussion. Something novel for this site. Other than the 10,000 other times it has been discussed. Cue the bandâŠ
1 Like
I believe that the study often referenced is a comparison of students accepted to elite universities but decided to attend somewhere else vs the elite university. That is much different than a straight school to school comparison. In laymanâs terms they said that if you are accepted to Harvard but attended Ohio State, your earning (not wealth) will be more similar to graduates of Harvard than Ohio State.
Note: neither Harvard or Ohio State data was used in the study.
2 Likes
I only looked at the first Economist article prior to the Paywall, but it does not control for individual student characteristics. I suspect the others do not as well. It is essentially comparing the average student at Harvard who is gifted, high achieving, and comes from a wealthy well-resourced family to the average student at a less selective college who does not have any of these characteristics; then assuming that the difference in salary must all be because of the school name and not these large individual students differences. So if that gifted, high achieving, wealthy kid chooses to go to U Alabama over Harvard to save money, he/she would suddenly become the average U Alabama kid who has a 30% chance of failing to graduate.
If you look at studies that control for any individual student characteristics they usually come to a very different conclusion. I suspect the poster above is referring the classic Dale & Kreuger studies, which found.
when we adjust for unobserved student ability by controlling for the average SAT score of the colleges that students applied to, our estimates of the return to college selectivity fall substantially and are generally indistinguishable from zero. There were notable exceptions for certain subgroups. For black and Hispanic students and for students who come from less-educated families (in terms of their parentsâ education), the estimates of the return to college selectivity remain large, even in models that adjust for unobserved student characteristics.
3 Likes
I found the college score card data cited in the Economist article very useful. Thank you to whoever on this site (on another thread) pointed to it. I think it was Data10
The inside higher Ed article said the study controlled for a variety of factors, but then adds that the salary figures reported (from 2003) didnât control for that. Not sure how that makes sense, but I donât have the actual study to refer to.
Tell that to the high schools for the wealthy where itâs the worst
The ability to prepare for a very specific multiple choice standardized test does not tell you how well the student can think, nor how well they will do when they have to put in 40 hours a week on written essays. The essay will tell you whether this student can express themselves clearly and actually has basic language skills.
Itâs not the federal loans which re burdening students, but the private ones. If students could finance their education with insured federal loans, they wouldnât be in debt. We donât need less money for federal loans, we need more money in federal loans.
What about having wealthy parents who can pay for SAT prep, private space for studying, decent food, and things like that? Do you think that kids have control over the money their parents have?
So you DO think that character should be considered in accepting students? I though that you were against holistic admissions. I am confused.
There are 2,000 non-profit 4 year colleges in the USA. 80% of them accept students based on GPA and test scores and little more. You are expressing anger because you do not like the admission criteria used by which a small number of colleges, which serve maybe 7% of the population.
I mean, U Illinois Chicago is a great university, and it only considers GPA and test scores. Same for The University of Kansas, Kansas State, U Arizona, U Mississippi, Oklahoma State, and others.
Tulane University has auto-admission for students from Louisiana, based on GPA and test scores alone. Texas has this, and almost 70% of the students admitted to UT Austin, a top college, are accepted based on their GP alone. Auto-admissions exist for all Florida public colleges, for U Missouri, Iowa State, Arizona State, and more.
A good student has dozens to hundreds of excellent colleges to which they will be accepted purely based on their GPA and test scores alone. No holistic review, no legacy, nothing.
So if you donât want to play in what you call a âcharadeâ, you are in luck - the great majority of colleges are not playing that game. If you believe that âholistic admissionsâ are âa big Jokeâ, have your kid apply only to colleges where there are no holistic admissions, or where their GA ensures them automatic admission.
5 Likes
Why would you think the essays are the studentâs own work? There are so many websites offering essay writing now.
6 Likes
Is this really the purpose of the essay(s)? If it is, wouldnât it have been better if SAT/ACT had retained their essay options few seem to like? Having students write a short essay on their own would have alleviated the concern @roycroftmom expressed.
6 Likes
Or submit a graded paper the student had used for school.
3 Likes
I suspect that a lot of the morally outraged in the meritocracy crowd would like to imagine that the kid who got in to school X over their darling Harold or Sally on the basis of this pernicious âotherâ criteria terribly over-estimate the gap in academic chops between the athlete, musician, artist and their otherwise unremarkable super academic kid.
I suspect the gap is smaller than they think.
Big-time D1 sports is another matter. Lots of room for debate there.
6 Likes
Yes, the Economist article is about correlation, as I said. I have been very careful to distinguish b/w correlation and causation.
In contrast, the study reviewed in the Inside Higher Ed article does control for SAT scores and other individual student characteristics. And it concluded:
The results show that, after controlling for all those factors, graduates of the most competitive colleges earn (10 years after graduation) 8 percent more than graduates of very selective (but not the most competitive) colleges, 11 percent more than graduates of colleges that are competitive, and 19 percent more than graduates of colleges that are not competitive in admissions.
1 Like
Same as in everything, there are ways to cheat. I mean, people are getting fake diagnoses to allow more time on the SAT, paying proctors to correct answers, etc.
Writing something under a very short time constraint says little about a studentâs skills at writing in normal situations.
I think that this is actually a great idea.
1 Like
On this point I do have some sympathy. There is just too much room for gaming it, particularly for applicants with vast financial resources. There are other ways to check into the veracity of an applicantâs âlife circumstancesâ and I do think schools should do more verification work in this regard.
If I admit a kid with a 3.6 and less high school rigor over a higher stat kid in part because the kid has overcome the kinds of obstacles the derails people 99/100 times, I want to know the story is not made up, and I want to know it before her junior year at Penn.
Sadly, I do think people lie through their teeth and are bold enough to make up wild stories about who they are on the assumption that nobody will check or remember the basis upon which they were admitted. Sadly, I think they are right most of the time. I expect/hope that itâs a small minority.
Instead of hiring another Dean of Whatever, schools should devote some resources to fact checking their applicants. The stakes are too high, and the vulgarity of someone claiming hardship when in fact they had advantages is too much to tolerate.
6 Likes
I looked up the actual study rather than the article summarizing the study. Their model that controlled for certain individual student characteristics in addition to a division of 4 Barronâs selectivity values only explained 16% of variance in earnings. The other 84% of variance in earnings seems to depend on other factors that were not evaluated in the study.
I suspect the weak predictive ability of the model contributes to some of the odd results. An example is below. The model suggests that being Hispanic causes a huge decrease in earnings for a less selective college, but being Hispanic causes an increase in earnings for a selective college, but not a most selective. Rather than assuming Hispanic kids should attend selective colleges, but not most selective collages, and avoid less selective colleges at all cost; I think it is more likely they are missing another important variable that Hispanic kids who attend less selective colleges are more likely to have than Hispanic kids who attend more selective colleges. And that other important missing variable is primarily driving the earnings rather than being Hispanic. Being Hispanic may not be a powerful predictor, but rather some important factor among the 84% of variance that is unexplained is more common among Hispanic students than other races. Models that have weak predictive ability like this, often show this type of strange pattern with the limited available controls.
*Being Hispanic decreases income by 34% if attending a less selective college
*Being Hispanic increases income by 6% if attending a selective college
*Being Hispanic increases income by 7% if attending a very selective college
*Being Hispanic decreases income by 11% if attending a most selective college
This type of unexplained variable also fits with the previously referenced D&K studies coming to a different conclusion, in spite of using the same sample group with same earnings as this study (in one of the multiple sample groups listed). Same sample group with different controls can lead to a different conclusion. The latter D&K study found that income was more correlated with where the student applies for college than where the student is accepted to or attends. Being the type of student who applies to selective colleges may also correlate with being the type of student who favors a certain type of job after college that averages higher salary, being the type of student who is open to traveling away from home for a job, being the type of student who has parental support/assistance/encouragement in career search, etc. Controlling for applying to selective colleges led to a different conclusion in this example, and seems to be one of the more influential discussed factors for income.
1 Like
Itâs hard enough to do a proper fact-checking on an adult when applying for a job (yes, we can verify degrees, any claimed licenses, dates of previous employment but thereâs a lot of stuff you can no longer check) I canât imagine how tough it would be for a minor.
Me- fact checker for XYZ university calling the district social worker: âJoanie claims that she spent the last few years in foster care because her mother is a drug addict and her dad abused her. Can you verify that please?â That gets a hang up. Me- fact checker calling Joanieâs physician âJoanie claims that due to her muscular dystrophy, she faced significant physical challenges in her life and yet was an active participant in her HS. Can you verify her diagnosis please?â Thatâs another hang up. Me- âStan says that when his parents lost their home due to personal bankruptcy it created a lot of turmoil in his family which he has managed to overcome, and heâs taken on a significant share of family expenses by working at Pizza Hut. You are the manager of Pizza Hut- is it true he uses his entire paycheck to pay his familyâs grocery bill?â
Etc. How do you verify someone elseâs difficulties?
4 Likes
Iâm sure some situations are harder than others, but Iâm confident that if it were someoneâs priority to at least eliminate the low hanging fruit, it wouldnât be like summiting K2.
If itâs a chronic illness, with which one of my kids has to cope, you ask for a doctorâs note. Simple, and I would have been glad to do it if it meant helping to ensure some crappy kid and their crappy parents with no conscience wouldnât be able to use something that is a very real struggle for people for something as trivial as getting into a school.
If itâs a parent employment or financial matter, you ask for the parents to sign a certification or confirmation letter, wherein you point out that the school is relying on the veracity of their representations and that if itâs discovered later that they werenât being honest their kid would be thrown out and no tuition refunds. And you can do some collateral sleuthing on your own - the Washington DOL is wonderful at catching people on L&I doing physical activity that is not commensurate with work place injury claims. The internet is a wonderful thing.
As for privacy barriers, you have applicants and their parents waive them to the extent theyâre legally waivable, provide some evidence of things that are very important, and then do your best. It wonât be perfect, but who cares? Itâs an effort to weed out precisely the despicable people you donât want at your school. If they wonât waive, then you tell them that their circumstances wonât be considered and theyâll be evaluated on transcript and test scores.
Itâs like anything: if you want to do it, youâll figure out a way. God knows they made it next to impossible to send in fake transcripts and test scores.
3 Likes
Hugs to you for dealing with chronic illness- very tough.
FYI, there is a cottage industry right now of doctors selling fake Covid vaccine cards, writing phony medical âGeorge is allergic to X and therefore cannot be vaccinatedâ letters for employers who require vaccinations except for a medical waiver, etc. So I donât share your confidence that getting an MDâs letter is the way to get the cheats out of the system- somehow it seems to encourage MORE cheating. (I think thereâs a podiatrist in Florida who can retire based on the $100 he charges for phony medical lettersâŠ)