You know that the stat you are using is from promotional advertising for a company that sells “how to improve your SAT” on LinkedIn? https://www.linkedin.com/company/amikkalearning?trk=article-ssr-frontend-pulse_publisher-author-card
Check sources. Always.
On a related note, an article that looks at racial disparities in college-preparedness in Connecticut:
The College Board reports a 99.5th percentile score of 1560. If we assume 1560, 1570, 1580, 1590, and 1600 are equally likely that’s 1 in a 1000 which is consistent with the reported .07%. For Stanford’s claim to be valid, 100% of students who scored at least 1560 scored 1600, had a 4.0 gpa and applied to Stanford.
Yikes, that’s bad analysis.
These scores, are they continuous? Do you know anyone that got a 1287.234?
If not continuous, how “chunky” are they?
So many assumptions. So little fact.
At the core you are asked the question should I put more weight on a published article by an author with empirical evidence or ad copy? And you figure that the author is lying and the ad copy is factual. Amazing.
And your post was a direct copy and paste from the ad copy, every word and the exact same 0.017z figure. The quote appears on the first page of a simple Google search. And you want us to believe you triangulated various sources. That dog won’t hunt.
Fixed.
Have you ever heard of a discrete uniform distribution? 1.7 million test takers, with 1/200 scoring 1560+ (from the College Board). Thats 8500 scoring 1560+. Stanford admited 2075 students in the most recent class. Four applicants with perfecf SATs and a 4.0 is 8300. Please explain how that could be the case.
Other sources put perfect scores at slightly under 1%. That’s an order of magnitude larger than your figure.
And we are way off topic here. The point of my post was to point out that even if you wanted to rank order applicants based on a set of metrics, the very top schools are going to have to deal with differentiating those applicants in multiple lenses. Like AA, for instance.
This is my source. What are your other sources?
Just stop. I already gave you the link to the company website that wrote the exact same phrase you used in your post. Going on about all the research you did after the fact doesn’t change the clear indication you did a simple copy-paste.
I admit I did a cut and paste. It was that only source I could find that gave a specific number for the number of test takers that had perfect SATs. Now tell me what your source for the claim that slightly less than 1% get 1600 combined on the SAT.
Since you are attacking the other poster’s sources, you should probably link to yours.
When I looked, I couldn’t come up with anything from the College Board.
Nor I. Just a series of guesstimates by various parties. The college board is notoriously silent on their “proprietary information”. Not likely to change since the whole premise that high scores improv graduation rates is under scrutiny these days.
Curiously, that’s is the crux of the EU regulations on big tech data. The EU wants tech to make the data available to researchers. Big Tech is not keen on that
Seems like you dudn’t see this which I posted earlier.
I did. I just interpret the term “perfect score” a bit differently.
The tables basically say to me that at 1520+ there is no statistically significant difference. Hence the 99+ rating rather than the decimal version you quoted. Do I think the admissions officer at Stanford was disingenuous as you seem to imply? No, I do not. The decimal version is clearly a reference to the ad copy. I don’t see any 6 decimal figures n that paper. They stop differentiating at 2 decimals.
And again, the point is how should the elite schools deal with an over abundance of qualified candidates? The SCOTUS decision wrt AA leaves that unanswered.
Addendum: Stanford tried to reduce the number of legacy admits, but the alumni push back forced them to retreat. And to be fair, I am critical of all the elites in their collective refusal to increase enrollment to meet demand.
Increasing enrollments would decrease rejection rates and thus lower “selectivity” which is one of the main criteria used by US New & World Report for their college rankings. Top schools already tinker with their admissions policy to artificially inflate their rejection ratings. E.g., students who aren’t quite good enough to get into Harvard are more likely to get into Princeton than students who are barely good enough to get into Harvard, b/c Princeton understands that most students will pick Harvard over Princeton if given the choice and being rejected by students hurts their ranking. From Avery, Glickman, Hoxby and Metrick (2013):
I don’t get how selectivity makes a school better. If you open more spots and hire more people to retain the quality of education, does that make the school worse?
Isn’t it related to the (infamous) rankings fro m US News and World Report?
I would say it is more because of their undergraduate coop program where there is study and work terms intermixed with the degree vs the internship program. Being an alumni of there plus another Canadian university (both for Actsci) my only comment is that what makes waterloo successful is the coop program, not really the quality of education or their computer labs for that matter.
And yes most companies in Toronto hire coops from the waterloo program, mostly because it is known and it is easy to find a suitable hire (all the coop students I have ever hired have come from Waterloo). And then for entry level where there are 100s of applicants (at least for P&C), having coop experience + exams gives one the edge.
Quebec and Western Canada.. not so much.
Actually, I identified the source the other poster used. Verbatim.
My criticism was bad analysis. I stand by that statement. Adding precision without supporting data is suspicious
A perfect score is 1600, somewhat like a 10 on actuarial exam. It does not mean being in the top 1%. I have no idea what you mean by statistically significant difference . Difference between what? I have known students who have had 800 on the math SAT and they are clearly superior in math skills to those with 760. The + explicitly means the top half of the percentile so again I don’t know what you are referring to.
Elite schools should become less elite by establishing satellite campuses which is what they do for profitable MBA programs and what state schools do for their regular programs. The only problem I see is that a faculty member who is capable of tenure at these schools would not be willing to teach at these campuses. These schools could look at a two tier faculty system where those denied tenure at the main campuses could be offered positions with a teaching focus at the other campuses. Students would probably end up with a better education.
