Sunday, June 22, 2008


Let's see, what would possibly be the motive of allowing score choice but not modular SAT testing (allowing students to take each of the sections separately)? If we have score choice for the SAT I, students can take the SAT as many times as they want without colleges knowing how many times they took it. I can just hear those gears rolling at the College Board, especially the bean counters. "We can really get great profit margins by giving into the base instincts of students who are already totally obsessed with the SAT. No longer will they have to worry whether college will think they are going overboard with testing. Maybe we can get kids to take the test a dozen times starting in 9th grade. Think of the revenue stream!! And we can get kids to test, retest and retest more with the full test, even
if they just want to try to raise their math score. Brilliant! Who thought of this idea. We gotta get this guy a bonus!"

Bill Shain when he was Dean of Admissions of Vanderbilt was doing a Q and A with counselors when I was visiting. He was asked a fairly common question: Do you average scores of students? Bill answered with a pretty common response, i.e. that, no, they take the highest individual sub-scores of the tests the students submit. But, he continued, he said that they did see how many times a kid did test and, at a certain point, that information did have an affect on how they viewed the student. So I ask, is it a good thing that colleges will no longer have this information? Does it not say something about a student who has taken 8, 9, 10 SAT or ACT tests?

Let's also do a little math here, particularly statistics. In an individual retest, a score is only likely to go up (or down, for that matter) a certain amount and the more a student tests, the less likely that score is to go up significantly. If a student took a test three times, for instance, it is not likely that the score will go up significantly with a fourth testing. But if you have a very large sample, say eight or ten tests, there IS a good likelihood that one of those tests will be an outlier, that a particular test will fall on the high end of the test range. After all, no one argues that an SAT test is an exact measure of ability but an approximation, affected by many factors: whether the student happens to get more questions on topics they handle better, whether they guess better, whether they are more seasoned test takers, whether they got enough sleep, enough food, whether the testing situation is better (fewer distractions, less humidity). I could go on and on, but logic tells us that the bigger and bigger the sample, the more likely that there is one instance where the mode, the highest possible result, will occur.

So let's go to human nature. There are lots and lots of kids, I would venture in the hundreds of thousands or even millions, who go through the college admissions process each year, who spend a huge amount of psychic energy worrying about whether their scores are high enough to get into the college they want. I see it in my own household and in a huge proportion of the kids in my school. It may be much more prevalent in suburban schools like ours or urban schools, but even there you are talking about a very large number of kids. And what are these kids going to do to relieve this tension (and their parents pushing to relieve their own anxiety) with score choice: test again and again. For everyone wants the illusion of control, particularly when someone else is making a big decision about their life. So now on top of the money spent on SAT prep, independent counseling, essay coaching, etc., the folks at the College Board can join in on this largess.

Is this a bad idea?: absolutley. We should be going in the opposite direction and reducing the obsession with testing. I run the testing at my school (not so next year...yeah!), and we are eating up more and more of the school year with testing at the expense of teaching and learning. We have, including make-ups, 12 days of HSPA (our graduation test for 11th graders) testing, 12 days of Terra Nova Testing (for grades 9 and 10), 4 days of biology testing (end-of course and performance assessment), 18 days of AP testing (including the exception testing), one day of PSAT testing, seven weekend days of SAT testing. And the state is now coming up with a new Algebra II test they are rolling out next year. When is it too much?

How about
some ways to reduce testing:

*aligning state graduation testing with college admissions testing so that one test counted for both purposes and the total would be a couple of sittings.
*Allowing modular testing so students would not have to re-take testing they did not need, including writing a new essay again and again and again.
*College announcing they they WILL average the scores of test takers who take many tests.
*More colleges at the most selective range going test optional (Harvard, you were the maverick in getting rid of early plans and embracing the Common Application...ready to take the plunge?)
*Allowing, as many colleges do, other testing (AP, SAT II) to substitute for SAT or ACT testing.

The College Board gave lip service to the idea of modular testing with ad hoc arguments against it. They said it would hurt the integrity of the test, despite the fact the the subject tests are given modularly with apparently
no affect on their validity. They stated that it would cost more money for the test for they would still have the same test administration costs no matter how many sections were tested. But again, the third hour of Subject Test testing often has one or two kids in the room, but that hasn't changed the administration of this test. And do you really believe that students would end up paying more if they only took the tests they needed and wanted? They rejected modularity (and, lets get serious here, never really considered it) for two reasons: It would get them less total revenue and might doom the already shaky writing test (note how they touted the study on the usefulness of the writing test with the conclusion that it should not be optional because it gave a .01, yes .01 increase in the predictability of a college freshman GPA).

I just sent my College Board membership bill back with no check (for the second year in
a row) with an explanation that I saw no advantage to membership. Until the College Board really gets serious about bringing to the members issues like modularity of the SAT (or this new score choice option), why bother to be a member? I encourage others to do likewise. Like the College Board, we can make financial decisions too, which is exactly what this score choice decision is. Any gobbley gook about serving kids in nonsense. This is just an awful idea. If it is instituted, the average number of SAT sittings per student will soar. Who out there thinks that is a good thing...except for the College Board bean counters?

I got to read some some news reports on this new policy (nothing yet on the College Board web site that I could find) and need to comment on issues that have been brought up in the discussion:

1) A quote from the College Board: "Students were telling us the ability to have more control over their
scores would make the test experience more comfortable and less
stressful," said Laurence Bunin, senior vice president of the SAT. ". .
. We can do that without in any way diminishing the value and integrity
of the SAT." Students taking the test many more times without those receiving that information knowing it not affecting the value and integrity of the test? Really? Maybe I took a different stats course than you guys. Aren't you the ones telling us that the score is not indicative of an exact point but a range? Isn't more testing more likely to get one to the maximum of that range? And telling us that students want this is a bit disingenuous, isn't it? Students want lots of things. 1st graders would want to play video games all day and eat candy, but we don't give it to them. Our job as educators is to give students what they NEED, not what they want. We are the professionals and our job is to do want is best for students. Maybe the College Board doesn't see things this way.

2) This will clearly just exacerbate the achievement gap between the rich and the poor. Now the rich will have one more tool to cement their societal advantage and the expense of those who cannot.

3) The College Board spokeman stated that they will not allow students to submit separate highest subscores but only the highest single test result so students cannot "game the system." Really? You really believe that this new policy will not do more to encourage students to game the system? This fails all credibility tests.

4) A major point that was brought up is that this places the SAT in line with the policy of the ACT. I find it interesting that the College Board took one of the worst policies of the ACT as their own (and coincidentally one which will have the greatest potential to raise their revenues) rather than some other more sensible and useful parts of the test. Why not consider making the SAT more straightforward and achievement based, eliminate the penalty for guessing (talk about a stress reduction) and making the writing section optional? Oh, none of these will increase revenues for the College Board. Silly me.

Lastly, one has to realize that almost every college in the country takes the highest individual math and critical reading (and sometimes writing) from the different administrations of the test a student submits. So what advantage is this really to students? And what about the increased anxiety of deciding which tests to report? “Should I send the score with the highest math score even if it has my lowest reading score?” This policy has no value added to students and significant costs. It will result in much more unnecessary testing with colleges having less information to assess students.