are colleges racist?

Canuckguy · July 4, 2011, 6:19pm

I think I found the answer to what I was looking for. In order to convert the new scores to the old scores, one needs to subtract 70 points from the verbal and 20-30 points from the math in the middle range of the curve.

[`Recentered</a>’ Scores Just Another Step Toward Mediocrity - Chicago Tribune](<a href=“`RECENTERED' SCORES JUST ANOTHER STEP TOWARD MEDIOCRITY”>`RECENTERED' SCORES JUST ANOTHER STEP TOWARD MEDIOCRITY)

Even more interesting, the research shows that while the verbal score has been little changed since 1980, the math score has been steadily increasing:

[LESSONS</a> - LESSONS - Sums vs. Summarizing - SAT’s Math-Verbal Gap - NYTimes.com](<a href=“http://www.nytimes.com/2002/08/28/nyregion/lessons-sums-vs-summarizing-sat-s-math-verbal-gap.html]LESSONS”>http://www.nytimes.com/2002/08/28/nyregion/lessons-sums-vs-summarizing-sat-s-math-verbal-gap.html)

I think it should be obvious who the winners and losers are from these changes. Another intentional unintentional walk?

texaspg · July 4, 2011, 6:27pm

That story is too old. Come over to parents of 2012 thread and check out all the horror stories of people upset about their kids math scores in SAT and how much better their other scores are (some are off by 200 points - less than their reading scores).

They did make the average 500 points. But people who are good in Math do well as long as they make few mistakes while those that are above average become roadkill.

bovertine · July 4, 2011, 7:23pm

All that follows here assumes I understand your point CG. It’s entirely possible I’m not following you correctly.

There have been many more high math scorers than CR scorers for a long time. Collegeboard always weighted verbal stuff more. In my day, the NMSF index was 2V+M

Everyone knows about recentering of the test - my earlier comment was not about the score level, but rather the differential between math and verbal scores. And frankly, nobody on this thread or probably many people on this website really care about a 10 point differential at the average score level. What people are concerned with here are scores **at the top levels **. Because nobody is getting into Harvard with a 500 math SAT.

I was really hoping this interminable thread would go away, but it just won’t. In an effort to keep from bumping it, I posted these reports in another thread, but instead that thread disappeared into the pile. These are SAT reports from 1973 and 2009 and data tables from 2010:

<a href=“http://www.eric.ed.gov/PDFS/ED099731.pdf[/url]”>http://www.eric.ed.gov/PDFS/ED099731.pdf</a>
<a href=“http://professionals.collegeboard.com/profdownload/cbs-2009-national-TOTAL-GROUP.pdf[/url]”>Higher Education Professionals | College Board;
[SAT</a> Reasoning Data Tables - For Interpreting SAT Scores](<a href=“http://professionals.collegeboard.com/testing/sat-reasoning/scores/sat-data-tables]SAT”>SAT Suite of Assessments – Reports | College Board)

What I believe these reports show, is that at the 700+ level (the actual score is irrelevant - pick a high score, or compare today’s 750 to 1973s 700 - the results are similar), the differential between scores has decreased.
In 1973 around 15K scored above 700 on verbal, and around 35K scored above 700 on math.

In 2009 around 73K scored above 700 in verbal, and around 97K above 700 in math.
If you look at the 2010 data tables for above 750, you get around 23K in CR and 44 K in math

If you care to recheck my stats, it’s possible I made a mistake. But if these are anywhere near correct I don’t see some diabolical plot in place since 1973.

I agree they have dumbed down the entire test significantly, at least IMO. But I think it is the verbal section which has really been simplified, with the elimination of analogies and antonyms over the years.

But I do think they might benefit from a few harder math questions, because as it is the test can become more an issue of how meticulous the student is (as texaspg points out).

texaspg · July 4, 2011, 7:31pm

Bovertine - I am sure your numbers are accurate. I am curious about percentiles though. Have the numbers taking the test gone up proportionally?

It sounds like recentering pretty much ensures 500 is the mean which should ensure 50% above the 500 score.

sewhappy · July 4, 2011, 7:37pm

A little off topic but some interesting posts on the SAT verbal v math today and 40 years ago. What I’m gleaning is that it is and always has been more rare to score an 800 in CR than in math, even though the frequency of math 800s is greater today. The the actual average math score is about the same. So the average in math is the same but there is clustering now at the high end.

My gut instinct has always been that if you can only max one of the sections, best to max the CR – unless you want to go to MIT or Catech.

bovertine · July 4, 2011, 7:44pm

I don’t know. I think the number of examinees is up to about 2 mill this year, and was about 1.5 mil in '73. And I’m not sure if they’re all included in the reports.

But that’s irelevant to the point I’m making (probably rather badly). I’m really only talking about the ratio between the number of students with very high verbal scores to those with similar math scores. On any administration I think pretty much everyone takes both, so the percentile ratio would be the same. If you get what I’m saying.

BTW, I’m not certain my numbers are accurate. In fact, they probably aren’t. But I think they are within range, although I’m not 100% sure about that really.

bovertine · July 4, 2011, 7:47pm

I think you are right sewhappy.

If I understand what you mean by clustering, I think there has always been some degree of clustering at the high end. I don’t think most kids scoring 800 in math are taxed to their intellectual limits by SAT I math. Especially since you basically need 100% to score 800 these days. But I think that was probably always true to some extent. I also think an 800 in math may represent a wide range of ability in math - I think this is what somebody called a long tail. But I also think that was true when Bill Gates and Eliot Spitzer got their 800s.

bovertine · July 4, 2011, 8:05pm

FWIW- THey don’t say in the 1973 report how many kids got an 800 in math, but they give the mean and SD so if you assume a normal dist you can figure out the z statistic and come up with around 0.7%. It looks like in 2010 it was around 0.77%.

Of course, not sure what that means. Probably neither here nor there. And possibly wrong too, of course. :)

sewhappy · July 4, 2011, 8:11pm

Fun to noodle over though if you’re waiting for the fireworks to start.

I suppose it’s more difficult to come up with a math section that will really tease out the distribution for the whole continuum of math ability out there, in comparison to the verbal test.

Dad_of_3 · July 4, 2011, 8:17pm

The most predictable thing that happens when you just pump in money is you create a higher expense without anything else necessarily changing significantly. And this is not anything to do with K-12. Same would hold in healthcare, or defense, or anything else; people will get higher salaries, better benefits, earlier and more generous retirement, less work, and a stronger lobby to prevent anything changing dramatically.

IndianParent · July 4, 2011, 8:33pm

Use the AMC 12 exam.

epiphany · July 4, 2011, 8:36pm

Spot on correct. Specifically, admissions officers have told me that they value the 800 in CR more than the 800 in Math. Others on CC have been similarly informed, as was the faculty at my D’s high school.

epiphany · July 4, 2011, 8:51pm

Indeed, and proven, in many a school system that has done so. Higher salaries to admins, more duplicative “programs,” more waste, no more effective teaching.

Here are the schools/school types that are working well, currently, in terms of the output, the product – i.e., the student!

~certain public schools in certain wealthy areas (usually suburbs)
~rigorous (long day, enforced discipline) charter schools in some urban areas
~certain privates, especially those with a history of a challenging curriculum and dedicated, effective teachers who love their work
~certain homeschool models in which the facilitating parent is available to pace, assess, as well as provide, instruct, and challenge.
~certain tutor-model formats which resemble homeschooling, but take place in a site, facilitated by teachers with subject specialties.

Overall, the public school format in middle-class locations has uneven results. Some are fantastic (as the ‘public vs. private’ thread of a few weeks ago discussed), but there is so much inconsistency with the public school model that I now favor national standards. (Doesn’t necessarily mean national curriculum, of course.)

ducks while waiting for reaction

coolweather · July 4, 2011, 10:37pm

About duplicate functionality and oversupply of school administrators: I wonder why there are so many school districts in each city. Besides those administrators there are also school board officials elected every 4 years. And when a new school is built, the names of those people are engraved in front of the school. What do they do to deserve? I have the impression that school board officials in my school district don’t know much about education and they are just dirty old men.

IndianParent · July 4, 2011, 11:24pm

So what is the solution? Racial quota?

IndianParent · July 4, 2011, 11:43pm

I think USA would benefit from a national curriculum AND a national standard.

texaspg · July 5, 2011, 12:01am

There are a lot of schools that already teach a single curriculum. However, the schools have a lot of uneven performance for the same AP exams used by all of them.

Canuckguy · July 5, 2011, 12:14am

I am aware of a study, by James Lech, on this Math-Verbal achievement gap:

[Math-Verbal</a> Achievement Gap | Ask.com Encyclopedia](<a href=“http://www.ask.com/wiki/Math-Verbal_Achievement_Gap]Math-Verbal”>http://www.ask.com/wiki/Math-Verbal_Achievement_Gap)

While I appreciate parents agonizing over their childrens performance in the Math portion, Anecdotal evidence can hardly stand up to empirical research as presented. Are we sure that these students are not performing to their true potential?

I think sewhappy senses where I am coming from. To put it as simply as I can, re-centering gives an additional 100 points, V+Q, to students in the average range, but it also makes students at the very top harder to distinguish from one another. This has to be done for political reasons.

Furthermore, if an 800 in math may represent a wide range of ability in math, why dont they re-design the SAT to tease it out? If it is and always has been more rare to score an 800 in CR than in math, why dont they make the test so it is equally rare to score an 800 in math?

If Richard Rothstein, who said:

“Nobody really knows why we seem to make more progress in math than reading. But one likely cause is that students learn math mostly in school, while literacy also comes from habits at home. Even if reading instruction improves, scores would suffer if students did less out-of-school reading or had a less literate home environment.”

is correct, then who are the beneficiaries of the current system, and who are hurt by it?

Now can you see why I call this an intentional unintentional walk?

epiphany · July 5, 2011, 2:46am

Once again another non-responsive “response” from IP. The discussion was the improvement of the public schools. (Hint: all races, none more “preferred” than others.) The previous poster warned (correctly) of the diversion of funds and the questionable outcome of merely more cash, an objection which is validated by the data. (No improvement in certain districts/States despite increasingly greater funding, and this can be demonstrated over time.)

The discussion was limited to targeted improvement in K-12 schools, not to college admissions. Your question is, I repeat, non-responsive. Further, as I mentioned, I don’t make college admission policy. I suspect no one on this thread actually makes policy.

However, your angry complaints can be addressed to the Boards and governing administrations of the elites you abundantly condemn (but probably want your daughter desperately to attend). ;)

bovertine · July 5, 2011, 9:11am

Once again, your study is about the average level, not the top performers. At the top levels the verbal and math achievement gap is narrowing, as is pointed out by looking at actual SAT test reports. I posted some. This is not anecdotal evidence.

Nothing like jumping on another horse when the first one comes up lame.

Recentering has made it harder to distinguish top students in both math and verbal. If anything, it has affected the verbal scores more at the top - what was once a 730 verbal is now an 800 verbal. It is not anecdotal ,it is apparent from the tables:
[SAT</a> I Individual Score Equivalents](<a href=“http://professionals.collegeboard.com/data-reports-research/sat/equivalence-tables/sat-score]SAT”>Higher Education Professionals | College Board)

But if your newly revised point is that the current SAT has more bunching at the top in both math and verbal, that is true. THere is now and has always been more bunching at the top in math. That is why there exist a plethora of other vehicles for assessing math ability at the top end - AP exams, advanced courses, AMC, AIME, special math programs, etc.

Plus, I’ll point out, despite “dumbing down” and recentering, it is still extremely rare for a student to attain a perfect SAT score. A few hundred at most.

I have no idea what your answer to this question is. I suspect those most hurt by it are poor immigrants, probably mainly those from Mexico and Central America, and to a certain extent from Asia. However, I suspect if you asked most of the Asian parents on this site if they would like to have their child’s admisisons based solely on their verbal score as opposed to some other arbitrary criteria they would jump at the idea.

Nobody needs to monkey around with test scoring to give somebody an intentional walk.

are colleges racist?

CONNECT WITH US