A little behind on this one, but I did want to direct people’s attention to a recent piece on the SAT in The Weekly Standard. Fascinating history and interesting little tidbits. (For example, did you know that the letters SAT no longer stand for anything?)
It also raised in my mind a question.
Objections to the SAT often center on alleged bias. One of the examples of bias most often cited (at least according to the article) is a question dating from the 60’s that asks students for an analogy to “runner is to marathon”. The correct answer is “oarsmen is to regatta”.
In this narrow context, I would agree with the test’s opponents. A prep school student would clearly have an advantage over his or her inner city peer answering this question.
Among other tenets, proponents of 21st century skills tend to view reading as a transferable skill. On the other side of the issue, people like E.D. Hirsch argue that reading, particularly for the purposes of comprehension, is based in very precise content knowledge.
Dr. Hirsch and his Core Knowledge Foundation have argued, persuasively I think, that the achievement gap is in fact a knowledge gap. As evidence he points to a study on the testing of reading comprehension that focused on a passage on baseball.
Four sets of students were divided by their reading ability and knowledge of the sport (in testspeak, that would be domain knowledge). The four groups were asked to read the passage and then answer a series of questions based on it, much as any reading comprehension test would do.
From the test results, researchers concluded, and Dr. Hirsch concurred,
that students with low reading skills were able to perform as well as the better readers when they had high domain knowledge.
In other words, reading ability contributed to comprehension less than knowing something about baseball.
This research supports, then, the lengths to which the College Board and Educational Testing Service go to scrub the SAT of racial and cultural bias, as well as the advocacy that drove them to do it.
But, conversely, opponents of the SAT who believe that bias produces discrepant test results among racial and demographic subgroups must concede that the move to a 21st Century Skills conception of reading will only increase the discrepancy. Don’t they? For if reading and language are transferable skills, divorced of content, then what difference does it make if the questions on the SAT refer to haute cuisine or Pokemon?