<p>Correlations between income and test scores in and of themselves are next to meaningless. But they get reported in the popular press, and even in some less sophisticated studies, by test critics (or well meaning, but naive journalists) as evidence that the tests are biased. </p>
<p>What most researchers work on is teasing out how much of that correlation is due to the test (or access to coaching) and how much due to other factors. It’s rare to read an article in the popular press or a post on CC, that mentions, as interestedad did, that HS grades also correlate with income, as do ACT scores, SATIIs, regents-style state exams, enrollment in college prep and AP courses, course completion, and other similar factors. God . . . I’m starting to sound like an apologist for the Colllege Board.</p>
<p>ZM, since the SAT is a cornerstone of the admission system, it is important to evaluate this type of data. It is also very important for schools to review (periodically) the criteria and weights used in holistic reviews. Simply stated, should the “system” compare the SAT score from someone who lives in a Madoff-typre of dwelling to the SAT of someone who lives in Central Harlem. While each school has access to its own closed environment and the data from its applicants, it would be helpful to increase the pool of the “subjects” being “tested” for correlation studies of income, SAT, and probably race and family structure. </p>
<p>Last but not least, it might be useful to understand the rules of privacy that really protect the information required by both your application for admission and for financial aid. It is entirely probable that all that fine print at the bottom of the forms waived most of the rights you seem to cherish deeply. </p>
<p>Here’s an example of the FERPA rules posted on the website of one school:</p>
<p>Further, according to Jack: ‘YOU don’t want the truth,’ with the “you” being all those full-pay parents who have kids in highly selective private colleges, many of which are on the editorial board of the NYT or are Congresspersons themselves. Thus, no clamoring, no demanding for research.</p>
<p>But, I disagree with z-mom. If an extremely strong, statistical correlation was shown, the press would be all over it, as would the Regents at the University of California (to name one public Uni). In short order, UC would drop standardized testing as being discriminatory. And, since the state is large, the students in California comprise Collegeboard’s largest single market, which also means CB’s cash cow, which they could ill afford to lose. And, once California falls, other states will join. Voila, the end of CB as we know it (hmmm, perhaps not a bad thing.). :D</p>
<p>xiggi: I agree that FERPA does allow for research exemptions. But, of course the problema is that 50%-70% of many private college students do not qualify for finaid, and many of them don’t even apply; thus, no data at all for close to half the matriculants.</p>
<p>OTOH, since each college is its own closed system, we only have data on those that were successful in navigating the selective college gauntlet, not all or even a large % of high school seniors. We have no-little data on high testers who attend a state college or juco, for example, since they many are full pay and won’t apply for finaid.</p>
I thank you for the informative and courteous answer, but I still don’t think the information really matters in that way. I don’t think statistics are the answer because they don’t tell the whole story. What would happen if there was a strong connection between SAT score and income? Tax the wealthy? Give extra points to the poor? How would you take into account other factors like parental education, family structure and so forth?</p>
<p>I get it about FERPA, but one has the choice to apply to schools or not and understand that the school might release the information. It would be vastly different for the information to be accessed without notification or consent for some esoteric research.</p>
<p>BB, a cynic might point out that that the 50/70% is about to take a nosedive! On a more serious note, I do think that, based on a datum such as Harvard’s disclosure that 50% of its students DO receive need based aid, one might assume that the umber of families APPLYING to financial aid is higher. Also, since most studies seem to establish the highest income “bracket” at $100,000, one could assume that people who do not apply for aid are in that highest bracket.</p>
<p>Of course, but is the academics really want to conduct a study, it makes no sense (to me) to lop off the top income group at $1xx. It might be enlightening to parse the differences, if any, between the top 6% income earners to the current Prez’ favorite target group, those $250+.</p>
<p>Harvard’s statistics on families receiving aid is much more expansive than most other schools. Harvard limits family contributions to 10 % of income for families with imcomes stretching all the way to $ 180,000.</p>
<p>I agree with Bluebayou’s thrust on extending the research into income levels with finer gradations than just 100K+.</p>
<p>I think there is a WEAK correlation, as the College Board says it has calculated, but not a strong correlation. I don’t think that high family income causes high SAT scores, but rather covaries with other things that can also cause high SAT scores among low-income students. Some low-income students get very high SAT scores indeed, so I’m not at all convinced that considering SAT scores as one element among many in considering college applicants screens out able, low-income college students.</p>
<p>“What would happen if there was a strong connection between SAT score and income? Tax the wealthy? Give extra points to the poor?”</p>
<p>You would reduce or eliminate the emphasis on SAT in college admissions, and try to find admissions predictors that aren’t close proxies of parental income.</p>
<p>IMO, such predictors don’t exist. By its very nature, admissions to highly selective colleges ‘advantages the advantaged’ (which is why H, P & UVa don’t look any different today than they did under early apps).</p>
<p>But wait. What if there is also a strong correlation between income and success in a challenging college curriculum? What are you trying to predict with your predictors, anyway?</p>
<p>With applications in five figures at many colleges, with high school grading standards all over the lot, with personal statements written by parents or purchased off the web,etc., don’t expect the sat, however imperfect, to disappear anytime soon.</p>
<p>While a lot of noise is made about SAT scores goes up with family income. what is conveniently ignored from the 1995 chart is that the whites with the lowest SES (0-10K) on the average outscore blacks in the highest SES. In other words, cultural factor is far more important than SES. This is very political incorrect, and I think that is why college board stops publishing scores of ethnicity + SES data since 1995.</p>