The class of 2003's SAT Math and Verbal scores have increased dramatically over the scores of previous classes of students:
The College Board, which owns the nation's most popular college entrance exam, said Tuesday that this year's high school graduates had an average cumulative score of 1,026 points on the SAT, up six points from 2002. Both the average math (519) and verbal (507) scores were up three points from last year.
That may not sound like much, but we're talking mean shifts here. Means don't shift that easily, especially not with sample sizes in the millions. What's more, minority test takers now make up 36% of the sample, as opposed to 30% 10 years ago, which contradicts the argument that the presence of minorities was guaranteed to explain the downward drift in mean scores.
This year's average math scores are the highest the College Board could document since 1967. Scores prior to 1995 were recalculated to reflect changes made that year so that the numbers would be comparable to more recent scores. The board was unable to provide comparable scores prior to 1967. The SAT was first given in 1926.
The College Board said the higher scores were due to increased enrollment in advanced math and science courses such as physics, precalculus, calculus and chemistry.
Right on. Of course, there's some blather about how this is the result of not teaching kids "pure calculation" methods, as though a focus on the basics of calculation was the reason for the Math SAT decline in the first place. The article also notes that young women's Math scores increased "notably" - but racial test score gaps still persist (as will, presumably, the claim that the test is culturally or racially biased).
Let's look at some numbers:
* The means have indeed climbed up - but they're just above where they were in 1972. Too many years of "progressive" and ineffective teaching techniques have taken their toll, but now that accountability is in the picture, we're seeing some positive results.
* Interestingly, almost half of the SAT-takers report having a B average, and 72% are in the bottom 90% of their class.
* The score gap is on page 6. Self-reported White students have the highest Verbal average, while Asians have the highest Math. In fact, the average for Asian females on Math is higher than for White males.
* Note that the "Other" and "No Response" categories aren't doing too shabbily. While the "Other" is only 4% of the sample, the "No Response" group contains over 355,000 kids, which would be almost a quarter of the sample. The "No Response" group has the second-highest overall Verbal mean and the third-highest Math mean. Evidence that kids who refuse to play the "group identity" game are smart? Or just sick and tired of being asked about race?
* More trivia - 2% of SATs are taken under special accommodations. 13% of SAT-takers come from families with incomes less than $20K/year, and 38% of them come from families where the parents did not attend (or complete) college.
* Oh, and there's always the fun part of comparing average combined scores for varying majors. Kids who want to go to college and major in biological sciences have an average of 1096; future philosophers or theologians, 1106; the little engineers that could, 1099. The bottom four averages? Technical and vocational, home economics, education, and agriculture.
* There's an equal or higher percentage of men than women in every score category from 550 on up.
* Size of senior class is almost perfectly correlated with SAT mean scores, except for that group who go to really tiny schools. Students who come from schools where the senior class was composed of a mind-boggling 1000 students or more had the highest mean scores; students from schools where classes were 100-249, the lowest. And check out those mean scores for Independent and Religiously Affiliated schools, vs the public schools.
Posted by kswygert at August 26, 2003 01:08 PM