<p>Interesting article on the SAT test and wealth.</p>
<p>For once, I would like people who site income - correlation to perform a rigorous statistical analysis and tell exactly what the corrrelation factor is. Unless it is close to 1, talking about correlation is meanigless.</p>
<p>There is little reason to spend any time on extensive correlations between the SAT and income-levels. The adage Garbage-In Garbage-Out rings true, especially when checking on the manner that "information" is gathered.</p>
<p>
[quote]
Wayne J. Camara, vice president for research and analysis of the College Board, said that the shift in income levels shouldnt be viewed as a concern, citing several reasons. He said that because so many students dont fill out that question or do so and dont really know their parents income, **this is the piece of demographic data in which the College Board has the least faith. **He also said that inflation may be at play, since many families see increases in income. And SAT officials noted that they were giving out more fee waivers for the test than ever before.
[/quote]
</p>
<p>This is the same source that leads us to believe that those self-reported GPA are ... might be true.</p>
<p>Those who have very accurate information (from tax returns) about such associations are the colleges themselves. And except through special studies conducted by Gordon Winston at Williams and Cappy Hill at Vassar, they don't like to talk about it. They all have it, though.</p>
<p><a href="http://hypertextbook.com/eworld/sat.shtml%5B/url%5D">http://hypertextbook.com/eworld/sat.shtml</a></p>
<p>There is the same kind of correlation between the scores and the level of parental education.</p>
<p>It could be that more educated parents also earn more, but the reason for better scores might be valuing education/higher expectations in those families, not just the resources available for the test prep... At least at our HS, the best test takers are the kids who self-study.</p>
<p>The correlation factor could be quite high without proving anything about causation. What I find most interesting are the known cases of low-income families with children who gain very high scores. Also interesting are high-income families whose children don't score particularly high. Those specific cases in which the general correlation breaks down are important for testing any theory about why family income has any correlation at all, from whatever cause, with test scores.</p>
<p>Maybe the test does measure basic intelligence to some degree. That can and does cross all income levels. But Las Vegas would love to have a 30% advantage over the players in the long haul. They do well with around a 10% advantage.</p>
<p>Why would anyone care about "causations"? It's ADMISSIONS that matter. </p>
<p>Maybe I'm jaded; but the specific cases are particularly uninteresting to me. The Winston-Hill study already proved that income is already used (whether consciously or unconsciously) as a screening tool by prestige private colleges, regardless of SAT score. But no tragedies there - Penn's loss is UCLA's gain.</p>
<p>It doesn't matter whether the correlation is causative or not -- the fact is that it is there, and reliance on SATs therefore biases the college admission process toward higher income students, to the extent that the colleges rely on the SAT. That is one reason the most highly selective colleges can claim to be both "need blind" and promise to meet 100% need of all students -- they know that their criteria for admission will skew their class toward upper income students. If the median SAT is 2250, for example, then that is drawing from a higher income demographic than a college where the median SAT is 1850.</p>
<p>Where is the proof that colleges would be any less full of high-income students if they disregarded SAT scores entirely? There is a HUGE literature on why some young people go to college, and many do not, and higher-income families have higher participation in higher education under many different systems of college admission in many different times and places all over the world.</p>
<p>
[quote]
Maybe I'm jaded; but the specific cases are particularly uninteresting to me. The Winston-Hill study already proved that income is already used (whether consciously or unconsciously) as a screening tool by prestige private colleges, regardless of SAT score. But no tragedies there - Penn's loss is UCLA's gain.
[/quote]
</p>
<p>mini, I'm interested what exactly did the study say? (Googled to no avail)</p>
<p>Gordon Winston's data show that elite school admissions, which is not supposed to rely on SAT, admits far more high income students than it would be if they rely on SAT.</p>
<p>And that there were between two and three times as many low-income students in the SAT ranges they were admitting than were actually being admitted.</p>
<p>It is available on line - look up "Access to the Most Selective Private Colelges by High Ability, Low-Income Students: Are they out there?" October 2005.</p>
<p>And lest anyone resort to that old canard that "well, but they don't apply", both Amherst and Smith have proven they WILL apply, and in large numbers, if the school puts time, energy, and money into it - the same time, energy, and money they've spent for the past 100 years putting into the old boys' network.</p>
<p>Again - no tragedy. It means there are an extraordinary group of wonderful students at state schools.</p>
<p>
</p>
<p>The way some posters seem to use that CC ubiquitous and pomposity-laden line "correlation is not causation" would make Senator Kerry proud. I guess when the correlation does not fit the purpose, it can be dismissed. When it does fit the "deal du jour", it does become relevant. </p>
<p>Funny how that works!</p>
<p>Now, I am just waiting for someone to show the correlation between the extremely high number of Pell grantees in the UC system, the assumed superiority of the SAT scores of Asians, and the GPA among different SES levels. Don't they have a lot of very high scoring poor people in California? Or do they happen to report the income differently on the IRS, the FAFSA, and on a ridiculously meaningless SAT questionaire?</p>
<p>mini:
[quote]
Again - no tragedy. It means there are an extraordinary group of wonderful students at state schools.
[/quote]
</p>
<p>I guess it depends on what one calls a "tragedy". Many (I believe most) of the state schools do not cover 100% of need; thus, the high-scoring, high-need kids are better off at private colleges giving them full-ride or close to that. I know a number of families for which the state school in their own state is not affordable at all (even with the finaid offered by the schools). But if their kids have the right "stats", they have a pretty good chance to get their need covered in a strong private college.</p>
<p>That's not even considering the states with pretty weak state schools: OOS public schools are not within reach for those families, that's for sure.</p>
<p>I think there is more "need" to cover for a poor kid attending an expensive private school full of relatively wealthy kids. For many the subtle things that go with wealth are tough for somebody who cannot compete financially to accept and still feel like a real part of the school. Some can handle it and others cannot.</p>
<p>marmat, I believe the study alluded to by mini actually concluded that for really high ability students (combined SAT scores of 1520 or even at the 1420) the low income students (bottom and fourth quintile) are actually over-represented at the elite private schools. The under-represented groups are the third and second quintile.</p>
<p>Given that our society places a high value on giving truly low-income kids a break, their over-representation is not a surprise. The over-representation of the high income quintile is also not surprising since our society also places a high value on wealth, and wealth does have its perks. With a holistic admission process, this outcome shouldn't be a surprise.</p>
<p>So, the truly under-represented are those in the middle two quintiles. Why should that be a surprise?</p>
<p>Marmat, I can't speak for other states, but my kids both were offered great financial aid packages at the various UC/CSU campuses where they applied (or in my son's case, now attends). Much better than the "100% need" awards they saw from private colleges, which had a different definition of "need" than the FAFSA EFC. I'd also note that high-achievement brought them merit money which included (for my son) full-ride offers. So I still think that for the high-achievers, the financial options are going to be fine at in-state universities, especially when accounting for factors such as travel expenses and the intangibles Barron's refers to, which are very real.</p>