This article in Salon shows where the SAT comes from and discusses how useful it is, or isn’t. A bit long, but right on the money. I produced this graph from one of the tables in the article; it shows the extremely tight correlation between SAT scores and family income, whereas the predictive value of the SAT on students’ college grades is very, very small: only a few percentage points above being completely random, according to the article.
In the social sciences, it is quite rare for any two quantities to be so closely linked as we see here in this graph.
Of course, the problem is: what is an acceptable alternative?
College boards formerly offered (and may still offer) content exams in addition to the notorious SAT. (Thank God for it personally: I am a good test taker and it led to scholarships despite my meager school record.) Those were good exams, but sadly they never seem to have caught on.
Here is a story that speaks to the entire enterprise. I was a grader on the first computer science AP exam (long long ago!) The exam was about the language Pascal but just about everyone taking it had learned BASIC. Thus we had little to correct and almost everything we did have to grade was wrong. It was a disaster. I saw no paper that had more than 5% of the answers correct. But when the statistics for that exam were published the following year, they fit the usual pattern, demonstrating how factoring (or curving) grades can substitute for examiner error as well as student error.
LikeLike
Close to 800 colleges no longer require the SAT or ACT but instead simply go on high school records.
Then another question would be: should we exacerbate inequalities by giving the wealthiest kids the very best education money can buy, while relegating the children of the poor and middle classes to decidedly inferior educations?
LikeLike