The New Orleans Times-Picayune has quite a different viewpoint about those infamous SAT analogy items than does the LA Times:
...the College Board's decision to drop the analogy section from the verbal portion of the test is to explicable as Adam Sandler is to tolerable.
Knowing the meanings of individual words is a useful skill, and so is being able to grapple with the relationships among different words. Because college professors routinely use analogies while teaching, and textbook authors use them in writing, it's certainly reasonable to include them in the SAT...
Of course, the loudest criticism of the analogy section came from people who don't like standardized testing at all. These opponents argue that the SAT is unfair to students from poor backgrounds and those whose first language isn't English.
Test makers should of course take pains to eliminate regional and cultural biases from the SAT and other standardized tests. But it's not at all clear that those differences won't show up in the new sections of the test -- or that it's the College Board's fault if they do...
With this last statement, the article sneaks up on the important acknowledgement that group score differences are not necessarily indicators of test bias, althought it would have been nicer if the writer had made that explicit. The LA Times writer, on the other hand, seemed pretty sure that it is the College Board's fault if there are group differences, as was evident by the writer's willingness to give plenty of ink to test critics who use the word "bias" loosely and incorrectly.
Posted by kswygert at July 29, 2003 04:11 PM