US Supreme Court curbs consideration of race in university admissions

And what the percentage of applicants were black versus the past.
Plenty of, “Oh, no more AA? Guess I won’t even apply,” folks out there.

A new motion filed Monday in a lawsuit spanning back two years accuses elite colleges of secretly favoring the wealthy in the admissions process and demanding $685 million in damages.

lol at “secretly”

https://thehill.com/homenews/education/5045305-elite-schools-mit-penn-georgetown-wealthy-students-admissions/

No… that’s a 49.2% decrease.

They do this throughout the article and their actual meaning is obvious, but it hurt my brain to read it.

Is it “… of the total enrollment”?

Obviously… they just worded it wrong.

Or calculated wrong. Either value is a valid stat, but their words and numbers are not in alignment.

And if they’re going to go with what percentage of total enrollment went down they should provide more info.

I mean if Hispanics as a percent of the total went from 20.3% to 16% that’s extremely different from if they went from 4.4% to 0.1%.

I could back into that, but that’s putting too much work on the reader.

1 Like

I tried to read it but I was still pissed with the way elections used only percentages so I looked into it and I saw that the total enrolment had dropped from about 560 to about 490.

I then stopped reading it because I wasn’t sure if there were offers made but students didn’t show up explaining the drop in enrolment.

It’s a poorly written article.

BTW, 63-32=31 Hispanic students being 4.3% of the total implies 721 total students, so Hispanics would have gone from 8.7% to 4.4% if the class size stayed constant.

But if 721 is accurate then Black students went from 6.0% to 2.6%, which is less than a 4% decrease and the article said more than a 4% decrease.

So I guess the class size must have changed materially. I didn’t get around to checking the Asian numbers.

The writer and editor at The Crimson suck.

Any program that requires 95%+ high school avg is just filtering based on grade obsession. Also actuarial science isn’t that challenging, let’s be honest. Certainly nothing to sneeze at but the fact that chatgpt can probably pass the first two exams is pretty indicative of what kind of intellectual skill is actually required here. Actuarial science majors aren’t doing stuff like PDEs, organic chemistry, quantum mechanics, etc. If you’re good at calculus and understand a geometric series, you can probably get by. So the idea that waterloo actsci students are somehow a cut above the rest in Canada is pretty funny stuff.

This is a complete myth that you get “launched into affluence”.

Yup, they are trying to predict who will become leaders so they can retain their powerful network.

chatgpt has surprising ability to solve difficult math problems. Last year someone tried it for the 2023 Putnam exam and it scored in the 50’s. That would put it in the top 50. The test was repeated for the 2024 exam with the latest pro version and it scored 80-90… Maybe 10-20 students will score higher. If you are not familiar with the exam, it is taken by only the top math students, and the median score out of 120 is often 0.

1 Like

Wow, when did this improvement happen? Last i looked it was having trouble with basic algebra.

2 Likes

Obligatory Algebra sweatshirt:

It’s weird how these algorithms work. Sometimes you think it’s like a miracle but then you introduce a small twist in the question and it goes cray.

That is amazing.

My second year analysis prof at Queen’s was on the U of T team that won the Putnam in 1938 (U of T won frequently in the early days of the Putnam competition: they were like the Maple Leafs of the 60s).

Anyway, I got to know the prof fairly well as he also led a small discussion group in his home on Tuesday nights for second year students that focused on ethics and global issues. I got a copy of the 1938 Putnam exam from him and was blown away by the difficulty. I am impressed AI can score well on that!

Wait a second, was the story written by the AI? :smile:

3 Likes

I can call you Betty,
but Betty when you call me, you can call me AI

1 Like

A couple of years ago I gave it one of my geometry quizzes and it didn’t do all that well.

1 Like

Perhaps AI can learn a lot in a few years.

Yeah, probably