The API numbers are out, and guess what? They tell you (once again) which students go to which schools.
What's that? You thought that they were supposed to tell you how well the students were being educated by their schools? You mean you bought that line? Let me tell you how the system really works:
Performance on standardized tests (including the CAT-6 and SAT) can be predicted very reliably by a few factors... none of which is the school the student attends.
Regardless of which schools students attend, the most reliable factor is the parents' level of education. More highly educated parents have kids who get better test scores. I wonder if that has anything to do with the fact that the average student spends less than 1,500 hours a year under a teacher's supervision in a classroom, and more than 3, 500 hours a year under their parent's supervision outside of school. Who do you think has a greater impact on how they spend their time, especially in middle school and high school where those 1,500 hours are split up between 5-7 different teachers?
Another factor that affects test scores more than which school a student attends is socio-economic status. Poor students do worse than rich students, no matter where they go to school. Not suprisingly, there's a lot of overlap between parents socio-economic status and level of education. An interesting exception which proves the rule is the fact that children of teachers tend to do better than other students whose parents are in the same income bracket. If anything, this might be used as evidence that teachers are under-paid...
Another unsurprising trend is that minority students do worse than white students. What might be surprising is that this trend has more to do with socio-economic factors than race. It just so happens that more minorities are poor. Rich minority students with highly educated parents do nearly as well as their white counterparts, and the same trend can be seen in the poor performance of poor white children of poorly educated parents and their minority counterparts.
So, what do API scores tell us? They tell us the level of education of the parents of the students at that school, they tell us about the socio-economic status of the students at that school, and to a limited extent, they can give us an idea of the likelihood of minorities being over- or under-represented at the school. What they can't tell us is how well that school is educating students.
What to do? How about de-segregating schools? We've been trying to do it for 40 years, but schools are still segregated. It's less a question of race, however, than of socio-economics. We need to integrate the schools with rich kids and poor kids together. We need highly educated parents' kids in school with less educated parents' kids. That's the only way that test scores will be useful in the ways we try to use them.
I expect this will happen soon... as soon as the Devil ice-skates to work and farmers need airplanes to herd their swine. Until then, by all means let's punish schools for being willing to educate the poor and needy. I mean, what's the use of being rich and well-educated if your kids don't get preferential treatment?
- "Wait a minute. Are you telling me that we're so far behind the other students that we're going to catch up with them by going SLOWER than them?" (I'm kind of cheating, this is from a TV show and I couldn't find the exact quote...)
Monday, April 11, 2005
What standardized test scores really tell us
Labels:
education,
inequity,
standardized tests,
statistics
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment