<p>
[quote]
I'll peruse the ed research literature to see what it says.
[/quote]
</p>
<p>Just keep in mind, anything you find would likely be the tip of the iceburg. </p>
<p>We haven't published the work we've done in this area, and I'm sure we're not alone. A lot of it is used internally and administratively. I've seen some interesting things presented at forums and conferences, too, which also doesn't always make it into the literature.</p>
<p>marite, much of the work correlating SAT scores to first year performance was done by the College Board, so you can bet that, while the research was, in and of itself valid, much was left unpublished that did not support the use of the SAT.</p>
<p>Most of the CB research seems to have used one or two factor research designs, and a lot of useful information gets smoothed out by their looking across a wide swath of institutions. It would be much more interesting to see research of the nature done by Avery, using multifactoral designs and controlling for college type etc.</p>
<p>Newmassdad: I agree. I just wonder how adcoms would try to figure whether they had made the right decision? On GPA? At what point in a student's career? It's known that students flounder in the first semester but usually find their feet by the end of freshman year. What about different grading practices across different departments? What about students who move out of their comfort zone and get lower grades?</p>
<p>I'd be interested to learn what criteria colleges can use to gauge whether they made the right admission decisions. Perhaps Hoedown could elaborate a bit on the type of data that has been presented in non-public forums?</p>
<p>There would be some information that colleges have from the committee that sits to decide academic probabtion or dismissal from the school. If there was a general uptick in these numbers than is typical to that school (particularly amg. freshmen, say), a college might investigate further each case of academic probabtion/dismissal to see if there are any similarities to be drawn.</p>
<p>Reasons for poor performance could be varied, though - like some who had drinking/addiction problems; gambling problems, just not showing up for class, etc. that can envelop students who may have had outstanding high school records.</p>
<p>Anyway, though academic dismissal data would be kept in college stats, I don't know that much of the reasons for the dismissal could be published...I'd think it would be part of student records being confidential.</p>
<p>Irishbird:</p>
<p>Would probation/dismissal be so widespread as to suggest a pattern of poor admission decisions? Could adcoms learn that students with a certain profile, from a certain educational background, have incurred probation or dismissal at a higher rate than others, or are these cases so rare as not to provide solid ground for drawing conclusions?</p>
<p>As well, about the various factors you cite for poor performance, could adcoms glean that they might exist from the application folders they receive?</p>
<p>I don't go to NACAC or AACRAO conferences anymore; they'd cover this in a lot more depth.</p>
<p>UCB is in the midst of a study that was presented at a recent conference (didn't go to the session, but I have the summary). They started with the standard study, examining traditional high school academic measures and their correlation to outcomes (GPA freshman year, GPA at graduation, leadership on campus, engagement with classes, pursuing graduate degrees, etc). Then they went further to see if the "softer" admissions variables (like drive, leadership in high school, overcoming obstacles) added to the prediction of outcomes. I think what they essentially found is that these factors doesn't greatly affect the traditional measures of "success in college" like graduation or GPA, but did help predict who seemed happiest and most engaged while in college.</p>
<p>Hoedown:</p>
<p>Thanks. could you elaborate a bit? At what stage of a student career are these measurements taken? I ask because I once went to a panel of tenured profs talking to parents of prospective students. I don't know what their brief was, but they all said that they did very poorly in their first year in college. Obviously, they subsequently did well enough to go on to grad school and to rise in their profession.</p>
<p>
[quote]
but did help predict who seemed happiest and most engaged while in college.
[/quote]
Hmmm. You can have very happy B students who are engaged in all sorts of ECs as opposed to less happy A students who study all the time and thus have less time for socializing.</p>
<p>Here's the link to a recent UC study on the respective predictive reliability of GPA, SAT1 and SAT2 scores:
<a href="http://www.ucop.edu/news/sat/research.html%5B/url%5D">http://www.ucop.edu/news/sat/research.html</a></p>
<p>The interesting thing to me is the dramatic decline in predictive force of the GPA from 1996 to 1999. What's up with that? The other thing is this: the study uses freshman GPA as the "test" of success. This is just anecdotal, but 40 years ago I was what would have been a 3.0 GPA - 1500 SAT kind of student. I almost flunked out of college as a freshman, so the relative predictive value of GPA vs. SAT would hold true. On the other hand, once I righted that boat I ended up doing well, getting into a top 10 law school, etc. So the next question: is freshman GPA really a good test of collegiate "success"? But not to stray too far from the topic, I think that validating admissions decisions is most likely done on a periodic study basis by most schools, like the UC study I link to.</p>
<p>From the summary it looks to me like they were most interested in GPA at graduation. Some of the other measures were taken from The U of California Undergraduate Experience Survey, and I don't know when that is administered. Your point, however, is well-taken--looking at freshman the freshman year is not enough. Of course, you don't want to admit students who will struggle so much freshman year they can't recover, but students do need time to adjust to college and relying solely on first-term or first-year grades will not give the admissions office a clear picture on whether they made the correct admit decision.</p>
<p>Back to the UC study--I think the engagement they were talking about included course engagement. So it wasn't just being club-happy. I really just have the two-page summary which is a little short on details. </p>
<p>My main point is that yes, colleges do look into these sorts of things, but they may not be shared so widely. Hopefully they ARE shared with enrollment managers and admissions counselors who make decisions.</p>
<p>I agree with Kluge. This was the gist of what all these tenured profs were saying at the panel I attended: "Don't worry about freshman year. Your child will turn out well." And, if we look at graduation rates of 97%, they mostly do.</p>
<p>I followed up on SSV's comment about the CSU statistics and it's interesting. Looking at high schools in northern California which have a good academic reputation what I see is that the students from those schools who enroll at CSU's tend to have lower GPA's than the state average enrolling at CSU (by about 5%), higher SAT's (by about 10%) and go on to have higher college freshman GPAs (by about 5%), again, compared to the statewide average. What that tells me is that SATs are probably a valid tool in determining which high schools grade more strictly, if nothing else.</p>
<p>This is straying a little further from the topic, but it fits in with the study kluge cites. I found this series of articles by the Omaha World-Herald (entitled "Lost in College") interesting. They commit the sin of focusing on the freshman year, but it's still pretty good reading. Alas, you have to register, but it's free.</p>
<p>They found that some of the U-N students who came from some of the more rigorous college prep schools & magnet schools struggled more than did the students from small towns and less elite public schools.</p>
<p>marite: in answer to your questions: academic dismissals/probation would probably be too small in number to have much validity in drawing conclusions. But if, after noting if there were any outside influences (i.e. drugs, alcohol, sickness, depression, gambling, not attending class) that impacted academics, you could look at the cases which didn't have any of those factors involved and then delve into the hs backgrounds to see what their preparation had been. But again, the cases to look at would probably be too few to draw firm conclusions.</p>
<p>To be able to discern from a college app whether someone will stop going to class, or start running all-night poker games in their dorm room, or become caught up in drugs/alcohol, would be asking the impossible, I think!</p>
<p>Also, much of this in-depth reviewing of student cases would probably be done by upper level administrators like Deans or the Retention Office and Enrollment Mgrs. and only be presented to the Admissions office if they felt that conclusions could be drawn that impact how they need to review applicant files.</p>
<p>More of this info. might be imparted to other Uni. offices that serve the students like Advising, Learning Resource Center, Residence Life & counseling services so that their services may be improved for helping the current students overcome the academic dificiencies (whether caused by weaker hs preparation or the various outside factors I mentioned.)</p>
<p>Irishbird:
This is pretty much what I think. In most cases, at the highly selective colleges with high rates of rentention/graduation, there just is not enough data to draw generalizable conclusions. And of course, adcoms cannot know in advance which student will have a nervous breakdown or become the life of the party and skip classes.<br>
In larger universities, there may be more cases of dropping out, probation or dismissal that would allow adcoms to find a pattern (low GPA,, low SATs, unchallenging set of courses, etc...)
I believe profs often complain about their students' lack of preparation (they don't know how to write proper English, they can't function without a calculator, etc...) but their complaints seem directed less at adcoms than at high school curricula and/or standardized tests.
In other words, I am not sure that there is a good way for adcoms to know whether they made the right decisions or not, since they can only work with the information they are given.</p>