| Subject: How Can You Map a School or Conference AI Distribution without All the Test Scores? |
Author: An Observer
| [ Next Thread |
Previous Thread |
Next Message |
Previous Message
]
Date Posted: 22:42:31 10/19/25 Sun
Princeton just announced that, beginning the 2028 admissions cycle, the school will require the SAT or ACT scores of undergraduate applicants.
Not sure why, having made the decision to change, you'd wait until 2028, but whatever. . . .
I believe that this leaves Columbia as the only Ivy which will continue to be test optional, and I will leave it up to our Lion posters to confirm or correct me.
Why is this important?
Well, it obviously affects how each of our eight schools select and aim to attract their students. "Test optional" sends the message that we as an institution do not feel that standardized test scores are a meaningful data point to the point where we would require them.
This is a major change in policy, kicked off by the new administration at Dartmouth (not quite so new now), which bravely went against the orthodoxy went it announced that, not only are SAT scores meaningful, they are the best predictor colleges have in terms of which students will thrive academically. That is a sea change in the slide away from test scores which began before the pandemic and then reached a crescendo during Covid.
How does this impact Ivy sports?
I don't know how the conference overall as well as individual schools can calculate a meaningful Academic Index distribution curve when obviously it's the lowest scoring applicants who are most likely not to submit their scores.
If you can't calculate a meaningful distribution curve, how do you determine where one standard deviation lands? If you can't determine where one standard deviation below the mean (really probably the median, given the asymmetric truncation of the curve at the top end) is, how do you set your floor for athletic recruits?
Logically, if Columbia is the only Ivy to remain test optional after Princeton changes its policy, Columbia should have a *HIGHER* AI threshold and its entire AI distribution curve should be shifted to the right. This would make it *HARDER* for Lion coaches to recruit their athletes, because fewer of their star athletes will qualify academically.
Now, Columbia University infamously has a long history of gaming their numbers and mis-reporting their SAT thresholds. I suspect that remaining test optional is another element of that strategy. When you're test optional, your reported scores go *UP*.
But this hurts Lion sports teams.
[
Next Thread |
Previous Thread |
Next Message |
Previous Message
] |
|