I always thought that the SAT analogy section tested vocabulary, not analogical reasoning per se. That's why you can game the system, by memorizing lists of SAT words, typical analogies, etc. You can game essays, especially through empty formalism, too -- crash AP courses and prep for writing proficiency tests are already on top of this.

Now if the SAT enlarged their reading section and really tested students' ability to spot precisely when someone was trying to fool them, whether through bad analogies or other faults of logic or rhetorical sleight-of-hand -- that would be something.

I don't buy a word of Cohen's argument. Effectively it is: "Look, I read Miller's analogies when I was a kid and now I'm a moral and political animal. But the kids of the future will just be trashy simpletons who write with no substance, but wonderful style."

Couldn't the world use a bit of wonderful style? What's so wrong with that?

And I agree with Tim: I've always heard the analogies pitched as a test for vocabulary.

Put me down in the vocab camp also. If we want to chart the ability to describe logical relationships, analogies are definitely not the best course of action. I always hated SAT analogies. Most were pretty straightforward, but some were ridiculously subjective.

it's so stupid to think it!!!!!

The Miller Analogy Test seems to be one of the most ridiculous tests I have ever seen. The analogies are, in many cases absurd, and so subjective, that the test is an insult to anyone who is required to take it.