The New Old SAT

Once again, the SAT has morphed.

And once again, it's not clear the changes are entirely for the better.

Saying its college admission exams do not focus enough on the important academic skills, the College Board announced on Wednesday a fundamental rethinking of the SAT, ending the longstanding penalty for guessing wrong, cutting obscure vocabulary words and making the essay optional.

Let's go through all the changes, one at a time.

Removing the wrong-answer penalty doesn't have much to do with the content of the test itself -- but it does have to do with the distribution of scores.  I can think of two effects.  First, a penalty makes people less likely to make a wild guess.  This, in turn, means that test-takers are less clustered in the distribution than they might otherwise be, as weaker students who aren't sure of the answers lose their opportunity to "fake" into the higher brackets.  Removing the penalty means that there will be more faking -- more right answers that aren't completely indicative of the test taker's mental prowess, and that will likely lead to more dense clustering of the resulting distribution (or an expanded distribution that is less differentiated at the extremes).  Second, removing the penalty will cause grumpy old people to say, "Back in my day, when you got an answer wrong, you LOST points."  And said grumpy old people might have a point.

"Cutting obscure vocabulary words" really can be translated as "reducing the amount of vocabulary a student needs to learn."  However you want to cut it, this is just straight-up dumbing down of the test -- making it more difficult to differentiate at the high extremes (although perhaps easier to differentiate at the lower extremes).

Making the essay optional is blessed relief.  If you don't have time to read the damn thing (and graders of standardized tests do not), the student shouldn't have to take the time to write it.  The essay was always something of a joke.  It's a merciful death, long overdue.

But that's not all the changes:

[T]he scoring will revert to the old 1,600-point scale — from 2,400 — with top scores of 800 on math and 800 on what will now be called “evidence-based reading and writing.”

So now the real old geezers like me, and all the youngin's... we can compare SAT scores.  It's just the milennials who are going to be the odd people out with their odd, hyper-inflated scores.  (Of course, the score is utterly meaningless.  I've long thought that the only thing that matters is the percentages.  But maybe I'm just competitive.)

I'm not sure what "evidence-based reading and writing" is.  The only "evidence" available to people taking a standardized tests are the things that they believe to be true, and the things that the test presents to them as true.  One would imagine that any sort of reading and writing outside of creative composition and poetry is "evidence based" in some vague sense.  It probably means something like "deploying facts in an argument" and "understanding how facts are deployed in an argument."  In other words, it -- meaning the reading and writing that takes place while taking the test -- is probably "about" evidence or "involves" evidence, rather than being "based on" evidence.

The overall push seems to be to align the SAT "more closely" to schoolwork -- a sort of summative assessment of the secondary school years, I suppose. 

That strikes me as only as strong a foundation for a test as the relationship between schoolwork and college readiness in the first place -- which can be highly questionable.  Good college students aren't really made in the classroom, I think.  They are made in the hours spent pursuing intellectual hobbies, reading for pleasure, rehearsing music and drama, and having engaged and active discussion with family, friends, and teachers about subjects other than what one thinks about the what happened on the Vampire Diaries last night.