David Freedman said:
The College Board has suggested a “college readiness benchmark” that works out to roughly 500 on each portion of the SAT as a score below which students are not likely to achieve at least a B-minus average at “a four-year college” – presumably an average one. . . .
How many high-school students are capable of meeting the College Board benchmark? This is not easy to answer, because in most states, large numbers of students never take a college-entrance exam (in California, for example, at most 43 percent of high-school students sit for the SAT or the ACT). To get a general sense, though, we can look to Delaware, Idaho, Maine, and the District of Columbia, which provide the SAT for free and have SAT participation rates above 90 percent. . . . In these states in 2015, the percentage of students averaging at least 500 on the reading section ranged from 33 percent (in D.C.) to 40 percent (in Maine), with similar distributions scoring 500 or more on the math and writing sections. Considering that these data don’t include dropouts, it seems safe to say that no more than one in three American high-school students is capable of hitting the College Board’s benchmark. Quibble with the details all you want, but there’s no escaping the conclusion that most Americans aren’t smart enough to do something we are told is an essential step toward succeeding in our new, brain-centric economy – namely, get through four years of college with moderately good grades.
via http://www.theatlantic.com/magazine/archive/2016/07/the-war-on-stupid-people/485618
Similarly, ACT estimates that only about 28 percent of recent high school graduates meet its alleged ‘college-readiness benchmarks’ in all four subjects of reading, English, math, and science.
For the record, the 6-year graduation rate in 2013 for first-time, full-time undergraduate students who began at a 4-year postsecondary institution in fall 2007 was 59%. Not all of those graduates have B- averages, of course. But, nonetheless, perhaps these definitions of ‘college readiness’ from SAT and ACT are too stringent?
There’s something even more profoundly wrong about this: The SAT is a norm-referenced test (which is why you’re not supposed to compare score across years because they are different cohorts, not that the media every pays attention to this leading to all of the “SAT scores are rising/falling” articles that don’t actually mean anything). 500 is supposed to be media, with each 100 points being a standard deviation (480 = 0.2sigma below median, 530=0.3sigma above), so this is another case of trying to cram over 50% into the top half. The reason that the states that are mentioned have less than 50% scoring over 500 is the fact that they have a larger population taking the test than the rest of the nation. Not sure if it’s still true, but North Dakota used to have the highest average SAT score, because <10% of their students took it.
Perfect example of how reformers will proudly show off their ignorance of how statistics work, and how standardized tests are scored (never mind what they actually measure).
An excellent point, Bill. I completely forgot to mention the norm-referencing in my blog post. Thanks for bringing this up since it highlights the inanity of using SAT and ACT as graduation exams…