Interesting standardized testing data from Princeton's freshmen survey

[Anonymous]

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

US colleges spend a lot of time trying to find the best students (while rejecting mos) for a particular class and then exert a ton of effort to ensure they graduate.

A top Canadian school like U Toronto essentially admits everyone who meets a certain threshold but then is quick to fail you out if you aren’t cutting it. They bake a failure rate into their numbers because they could never have say 100% of the starting EE majors actually continue all 4 years as they don’t plan for that when you get to more advanced classes.

This and there isn’t a linear relationship between test performance and school performance above a certain threshold.

[Anonymous]

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

I don't see any evidence for there being no association.

[Anonymous]

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

I don't see any evidence for there being no association.

You can check the retention rates. They’re unchanged.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:One interesting missing stat is AP scores and whether they were submitted. Do you know why? Because notwithstanding how much DCUM slobbers over AP scores, they don’t matter for college admissions.

It’s not missing. Only 13 percent of admitted students didn’t take any AP. The vast majority took several or a lot. It matters.

Learn to read.

They weren't surveyed on what their AP scores were or whether they submitted them. They were asked how many AP CLASSES they took. It's the CLASSES that count. Not the scores.

+1, DS took 17 APs, and didn't submit a single score. He's now a junior at Princeton.

That's bizarre. I guess one of those APs wasn't AP Stats or he could have told you that your anecdote do not substitute for actual data.

Oh amazing! Please I’d love to see the data. I’m a scientists so it’s part of my job.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

US colleges spend a lot of time trying to find the best students (while rejecting mos) for a particular class and then exert a ton of effort to ensure they graduate.

A top Canadian school like U Toronto essentially admits everyone who meets a certain threshold but then is quick to fail you out if you aren’t cutting it. They bake a failure rate into their numbers because they could never have say 100% of the starting EE majors actually continue all 4 years as they don’t plan for that when you get to more advanced classes.

Oh this is interesting, and I appreciate the explanation. I much prefer the US where we have centers for teaching excellence and invest into teaching than dropping out students who are doing poorly. When I was in college, I would have international professors who would express how lucky we are, because, back home, only 10 or so people would get into the top major and top program in their entire country and the failure rate was still insanely high.

Or maybe the international programs were more demanding?

That’d be great if you had evidence for such speculation.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

I don't see any evidence for there being no association.

You can check the retention rates. They’re unchanged.

At top schools everyone graduates, the schools basically don’t flunk out anyone. What schools saw was lower grades, kids transferring to easier majors, more remedial classes needed, and the need to make existing classes easier. At the University of Texas they saw almost a full letter grade difference (.86 gpa difference) between kids let in TO (average SAT in the 1100s) vs kids who submitted (average SAT in the 1300s) even after controlling for high school grades and class rank.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

US colleges spend a lot of time trying to find the best students (while rejecting mos) for a particular class and then exert a ton of effort to ensure they graduate.

A top Canadian school like U Toronto essentially admits everyone who meets a certain threshold but then is quick to fail you out if you aren’t cutting it. They bake a failure rate into their numbers because they could never have say 100% of the starting EE majors actually continue all 4 years as they don’t plan for that when you get to more advanced classes.

This and there isn’t a linear relationship between test performance and school performance above a certain threshold.

That is the opposite of true. The relationship continues all the the way up to 1600.

And your chances of admission continue to increase with SAT score all the way up to 1600.

Wishful thinking has created a mythology around test scores that makes it easy for mid kids to think they have a shot at top schools. They don't.

[Anonymous]

Anonymous wrote:Is it just me or does the #1 ranked school have a lot of students who look weak academically?

It’s just you.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

I don't see any evidence for there being no association.

You can check the retention rates. They’re unchanged.

At top schools everyone graduates, the schools basically don’t flunk out anyone. What schools saw was lower grades, kids transferring to easier majors, more remedial classes needed, and the need to make existing classes easier. At the University of Texas they saw almost a full letter grade difference (.86 gpa difference) between kids let in TO (average SAT in the 1100s) vs kids who submitted (average SAT in the 1300s) even after controlling for high school grades and class rank.

Really? Most top schools have more and more students in more difficult majors. This seems unsubstantiated.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

I don't see any evidence for there being no association.

You can check the retention rates. They’re unchanged.

At top schools everyone graduates, the schools basically don’t flunk out anyone. What schools saw was lower grades, kids transferring to easier majors, more remedial classes needed, and the need to make existing classes easier. At the University of Texas they saw almost a full letter grade difference (.86 gpa difference) between kids let in TO (average SAT in the 1100s) vs kids who submitted (average SAT in the 1300s) even after controlling for high school grades and class rank.

UT lets in a significantly different profile than Princeton though. They also have a top 6% rule.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

US colleges spend a lot of time trying to find the best students (while rejecting mos) for a particular class and then exert a ton of effort to ensure they graduate.

A top Canadian school like U Toronto essentially admits everyone who meets a certain threshold but then is quick to fail you out if you aren’t cutting it. They bake a failure rate into their numbers because they could never have say 100% of the starting EE majors actually continue all 4 years as they don’t plan for that when you get to more advanced classes.

This and there isn’t a linear relationship between test performance and school performance above a certain threshold.

I don't understand the point. PP said that kids don't dropout of Princeton, not that they all are achieving 4.0s. "not dropping out" is a far cray from comparing school performance among the class

Taking again the U Toronto comparison...I know my kid who applied for CS had to have scored a 5 on Calc BC to even apply for the program. It's safe to say if the only kids admitted scored a 5 on Calc BC and whatever may be the Canadian equivalent (as juniors or earlier), they are probably fairly strong students from the start. Now, there are still many kids that score a 5 and select a major and find they don't really like the major or don't put in the effort to do well, and they fairly quickly realize they need to switch majors or are kicked out of the major.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone, without an agenda, explain why it seems that lower sat scores seem to have no association with dropout or retention rates at these schools? If these tests are such good predictors of success, why wouldn’t. Complete 100 point drop in score not change your outcomes.

US colleges spend a lot of time trying to find the best students (while rejecting mos) for a particular class and then exert a ton of effort to ensure they graduate.

A top Canadian school like U Toronto essentially admits everyone who meets a certain threshold but then is quick to fail you out if you aren’t cutting it. They bake a failure rate into their numbers because they could never have say 100% of the starting EE majors actually continue all 4 years as they don’t plan for that when you get to more advanced classes.

This and there isn’t a linear relationship between test performance and school performance above a certain threshold.

I don't understand the point. PP said that kids don't dropout of Princeton, not that they all are achieving 4.0s. "not dropping out" is a far cray from comparing school performance among the class

Taking again the U Toronto comparison...I know my kid who applied for CS had to have scored a 5 on Calc BC to even apply for the program. It's safe to say if the only kids admitted scored a 5 on Calc BC and whatever may be the Canadian equivalent (as juniors or earlier), they are probably fairly strong students from the start. Now, there are still many kids that score a 5 and select a major and find they don't really like the major or don't put in the effort to do well, and they fairly quickly realize they need to switch majors or are kicked out of the major.

The common DCUM belief is that students with moderately better test scores (1450 vs 1500) have radically different GPAs. Even though fellowship attainment and graduate school outcomes have remained pretty consistent for colleges.

[Anonymous]

Anonymous wrote:

Anonymous wrote:One interesting missing stat is AP scores and whether they were submitted. Do you know why? Because notwithstanding how much DCUM slobbers over AP scores, they don’t matter for college admissions.

It’s not missing. Only 13 percent of admitted students didn’t take any AP. The vast majority took several or a lot. It matters.

I'm guessing the international student admits didn't take APs. Plus some kids from the low income/high poverty schools etc

[Anonymous]

Anonymous wrote:Is it just me or does the #1 ranked school have a lot of students who look weak academically?

Princeton alum here and I'm appalled by how many kids reported barely studying n high school. Lots of kids did less than 10 hours of homework per week.

[Anonymous]

Anonymous wrote:

Anonymous wrote:Is it just me or does the #1 ranked school have a lot of students who look weak academically?

Princeton alum here and I'm appalled by how many kids reported barely studying n high school. Lots of kids did less than 10 hours of homework per week.

A lot of school districts are reducing homework. There’s no proof that homework actually improves learning. It’s just good discipline.