I wrote a post addressing this on a different thread, and will repost my comments here:
7 Likes
@Mwfan1921 recently posted on another thread that MIT is reinstating their standardized test requirements. Hope more do this 
6 Likes
Well the California Cal State system followed the UC’s lead and are now permanently test blind.
2 Likes
CSU’s report: https://www.calstate.edu/apply/Documents/CSU-First-time-Freshman-Standardized-Exams-and-Admissions-Recommendations.pdf
What factors each CSU campus used recently for 2021-2022 frosh admission: https://www.calstate.edu/attend/counselor-resources/Documents/2021%20-%202022%20CSU%20Campus%20Admission%20Factors%20Summary.pdf
2 Likes
CSU’s report references this study: Predicting College Success | Policy Analysis for California Education
1 Like
DS spent quite a bit of time on the TCU app. Worked over his essays a lot, attended virtual events, emailed regional admissions officer. He has loved TCU for a long while and it was a blow to not get in! DS has been working on his list for at least 6-8 months, and decided on his schools over and over again. He didn’t submit his scores, 1450 because he thought it would hurt him at some schools.
The Kurlaender and Cohen 2019 study (referenced by CSU’s report) notes the following correlations to first year college GPA with full controls:
| Factor | CSU | UC |
|---|---|---|
| HSGPA | .48 | .57 |
| SAT | .38 | .58 |
| SBAC | .38 | .54 |
| HSGPA + SAT | .50 | .63 |
| HSGPA + SBAC | .49 | .61 |
| HSGPA + SAT + SBAC | .50 | .63 |
A few notes:
- All factors predict first year college GPA better at UC than CSU.
- CSU gets more predictive power from HSGPA than SAT or SBAC; adding SAT or SBAC to HSGPA gives only minimal additional predictive power.
- UC gets approximately similar predictive power from HSGPA and SAT, with some added predictive power from using both. SBAC has somewhat lesser predictive power at UC.
- The UC finding is at odds with studies from the 1990s-2000s that found predictive power more similar to CSUs now. Perhaps during the last few decades, the compression of HSGPAs at the top of the scale (due to both UC emphasis on HSGPA over SAT based on those previous studies, and high school grade inflation) has reduced the predictive power of HSGPA at the top end (where UC mostly admits from), but not at ranges below the top end (where CSU admits many students from).
Not shown in table above, but elsewhere in the report:
- At CSU, all factors predict first year college GPA better for White and Asian students than Black and Latino students.
- At UC, all factors predict first year college GPA best for Asian and Black students, then White students, then Latino students.
- At both CSU and UC, all factors predict first year college GPA better for not SED students than for SED students (SED = socioeconomic disadvantaged).
1 Like
Unless a college requires all first-year students take similar courses, the correlations are likely understated. Students with lower HSGPA and/or SAT scores are likely to take easier classes in their first year in college, resulting in higher college GPAs than if they had taken harder classes with other students with higher HSGPA and SAT scores.
I’m assuming this is in reference to the situation being talked about - I haven’t read that thread itself.
I definitely feel for your son and send my hugs, but it would have been a mistake to not send a 1450 SAT score to TCU. Their 25-75 range is 1150-1340 according to their website. By not submitting a higher score than their 75%, it implies that he got lower unfortunately.
I have no idea if his guidance counselor can call the school and do anything about it now assuming it is his #1 choice, or if the guidance counselor gave bad advice (happens sadly) or what the “going forward” plan is - does he like any of his acceptances?
I thought it was important to add my thoughts for anyone reading and contemplating future applications. In my experience an ACT/SAT above a college’s 75% can only help, not hurt - even considering yield protection - esp for a student showing a college their interest.
8 Likes
Yes, it was bad advice. We are very new to this process and he had someone proofread his essays but we were not sure about everything. He does like his choices and he can only attend one school. The schools who said yes to him he is carefully reviewing everything, finaid, programs, etc. I know he’ll thrive wherever he goes. He worked his tail off these 4 years and what we were told to expect and what has happened are two completely different realities especially as an underrepresented minority student. That aside, he didn’t rest on that alone and took the hardest, most rigorous courseload and played two sports on top of that (football and baseball). I am glad he chose to keep playing and doing what he loved. He also played concert band since grade 5 (Euphonium).
7 Likes
Thanks for posting this! I just glanced at the study quickly, but I imagine that the data are from when test scores were required for admission. If so, the analysis involved an artificially narrow range of test scores. Would be interesting to see correlations between test scores and college performance now that test scores are not considered, but that would be a difficult analysis…
The study was from 2019.
Many students who recently started at CSU or UC had no SAT or ACT scores at all, because of test cancellations during COVID-19. So even if CSU or UC got all of those who came in under test-blind admissions to send scores, many would have no scores to send.
He sounds like an awesome kid who will do well wherever he ends up. Some college is going to be very fortunate to have him. Kudos to him for his accomplishments!
I wish all guidance counselors were good at their job, but like any other profession, there’s great, good, average, bad, and ugh unfortunately.
Best wishes for y’all going forward.
3 Likes
Rejections are always hard, particularly when you’ve put a lot of your heart into it. And as I indicated above, to me that was the biggest surprise out of your son’s results. But he’s been doing great things in high school and I feel confident he will continue to do great things all through his life. Whichever college he decides to attend will be lucky to have him.
1 Like
We just got an email stating that he is in the USC Viterbi Pathways program. Not sure what exactly this means. Can someone help?
For more information, I would head over to the USC forum.
3 Likes
Thank you!
Everything I’ve read is that gpa is a far better predictor for college success than the testing and more fair between economic levels which is why so many AOs have wanted test-optional for so long. I think test-optional is here to stay especially after the CA schools have gone test-blind.
That is probably generally true, based on recent CSU studies and older UC studies.
However, if a college is in a situation where HSGPAs are compressed at the top of the scale (e.g. due to HS grade inflation and increasing selectivity in terms of HSGPA, as has been true at UC), HSGPA may no longer have a predictive advantage over SAT/ACT, as suggested in recent UC studies.
It is also possible that if a college is highly selective with a bias toward SAT/ACT (opposite of UCs), it may have compressed its students’ SAT/ACT at the top, reducing its predictive value relative to HSGPA. Also, a highly selective college where both HSGPA and SAT/ACT are compressed at the top may not find too much predictive value for either among its enrolled students.
In cases where compression at the top occurs, the measure may be mostly used in a negative sense, in that a “low”* HSGPA or SAT/ACT score can keep you out, but high ones will not (by themselves) get you in.
*In a relative sense compared to what admitted students typically have.
2 Likes
2.2 Million in the class of 2020 took the SAT. 1450 is 96th percentile of those. That’s 88,000
1.78 million took the ACT. 1450 is a 33, per the joint concordance, which is 97th percentile. That’s 53,000. I’m confident that’s over 120,000 individuals combined.
5% of 3.2 million graduates is 160,000.
I’m pretty certain there’s a very high correlation between the 120,000 and 160,000. More than 80,000.
1 Like