Wednesday, April 03, 2013

School Testing

Teachers in Georgia have gotten in trouble for faking results of school tests.  They shouldn't do that, of course, but looking at the perverse system is deserving of at least some attention.

I don't know these particular tests, but I think I am safe in assuming they are much like the other broad tests given to students in nearly all schools in the country.  When I was but a lad, we had the Iowa Tests of Basic Skills every year. They are lead-ins to the PSAT's and SAT's, and are very similar.  They are advertised and treated as achievement tests, measuring what the teacher, the school, and the pupil have accomplished. Hah.

But the SAT is essentially an IQ test.  Insofar as these tests resemble the NEDT's, NMSQT's, NECAPS and other alphabets, they are just IQ tests as well.  (AP exams are more of an achievement test.) IQ is overwhelmingly genetic, and barring the extreme negative circumstances of blows to the head and children kept in closets, some would say "It's all genetic."  In fact, Steve Hsu did say that in the video I posted recently. I'm saying it's north of 90%.

There are some ways of gaming the system within the rules.  I suspect my town, which inexplicably scores highest in the state, is raising the floor by reducing the number of kids who blow off the test and just screw with it or leave it blank after fifteen minutes. (There is some suspicion that the Finns do this on an international level as well, appealing to the national honor to get all students to do their best.  Of course, you have to be a pretty decent and intelligent country to accomplish that, so it may not be unfair. A lot more American children would just say "Yeah, right." to that.)  You can "teach to the test" with more and more specificity, and buy some points that way as well.

But in general, you can't move the dial much.  In NH, Hanover HS generally scores very high in all academic categories.  Dartmouth College.  Ditto UNH professors' children at Oyster River HS, which also scores well.  Inner cities score poorly, very rural schools often do also.  Wealthy communities score well. Yet even at that, the differences from highest to lowest, at least in NH, are not that huge.

So the tests are measuring something other than teacher or school competence, yet it's what the teachers and schools are graded on.  And if anyone knows how grading systems work and how to crack the code, it's going to be teachers.

I am putting some energy into the thought-experiment of Americans understanding that the schools cannot make children more intelligent in terms of candlepower, but might be able to do a lot in teaching the other 70% of stuff one needs for good life outcomes: culture, conscientiousness, cooperation, character.  They do some of that now, but the big payoff of community attention is around that item which the school has the least control over: student IQ.  How would things change if we accepted that?


bs king said...

My brother Dan had that epiphany at one point in Chicago. They were having them do a few "moments of meditation" per day with the kids where they basically were taught to breath and count to calm themselves down. There was some grumbling among the teachers that this was going to cut in to class time.

It hit Dan one day that for kids in a bad section of Chicago, the Kreb cycle would never be used. The ability to calm ones self down however could mean the difference between being a productive member of society and a lifetime in jail.

Texan99 said...

You can't teach candlepower, but one school can do a very different job from another school in the basic task of producing kids who, after a year of instruction, have improved their ability to make sense out of a basic written text, or to perform some routine calculations, neither of which is hard to test for. I don't have any problem tailoring the compensation of educators to reflect that test's outcome.

To the extent that schools really can't make much of a difference in the outcome for a student, it's not clear to me why they're entitled to any public money at all, or to any power to enforce attendance.

Assistant Village Idiot said...

They can affect outcome, but much more in the manner that you describe than in what they currently think. What the intelligence goes into - the specific skills, the specific items of culture, the ranking of virtues - the school has a subsidiary, but significant voice in, competing with the parents and the surrounding culture.

jaed said...

It is lost on me why so many of these tests are on absolute measure of ability or achievement rather than on degree of improvement (or otherwise). It isn't difficult to test a student's reading comprehension or ability to do arithmetic (or ability to dance or to read French - whatever is being tested for) at the start and end of the school year, and take the average difference.

IQ affects absolute scores much more than it affects what can be learned during the year, at least if the instruction is pitched at anywhere near the correct level (if it isn't, that's a whole 'nother problem). When judging schools or teachers, we don't want to know how smart their students are; we want to know how well, if at all, the teaching helped the students they have to make the most of what they've got.

So why isn't it done that way?

Assistant Village Idiot said...

Amen and amen. And because what is learned is not being measured, I would far rather the focus be on what culture is taught and values advertised.

Why? Because most people who are involved with the question have the strong agenda of "what schools can my kids get into?" "My" being defined as either progeny or students. What gets published in the local paper? Athletics, plus the honor roll and the awards, which are largely high-IQ measures. The various competitions, robotics, DI, etc also get press and I think that is better.