Seems that this would just cause "teaching for the test". I'm in favor of incentive plans but perhaps there are better ways to measure a teacher's performance than the results of a standardized test.
I think this would be agreat incentive plan, and I am not kidding: Fire all administrators, and allow teachers to run schools on a rotating basis. Increase teachers' pay substantially after saving all that money from administrator salaries, and extend the school day and year. Allow teachers one hour for lunch under this new plan, and give incentives like green stamps for merchandise to parents in inner city schools, so they can start helping their kids succeed. OK, no one is going to do any of that, but that is where I would begin!
You are correct. They already do teach to the test. I have my kids in private schools because the public ones in Florida really are a joke. Jeb Bush, bless him, is a great governor who really wants a better educational system for the kids of Florida, but the FCAT is failing because the teachers are only teaching to the FCAT. I know at our districted elementary school (we have many friends whose kids attend) that there are psychologists on school grounds the day of the FCAT to help kids who are feeling overwhelmed by the pressure. In ELEMENTARY school. Geez.
Also, we had 1 yes 1 high school in our county receive an A grade after FCAT testing. However, when looking at the numbers only 53% of kids could read at or above their grade level. If that gets you an A rating I'm afraid to see what a B school is.
It seems that Florida is using a scale of 100-500. So, if little Billy scored 347 in 4th grade and then scored the same or higher in 5th grade, you can tell that Billy's 5th grade teacher taught Billy a full year's worth of 5th grade skills.
If Billy was the smartest kid in 4th grade and scored a 499, but only got a 378 in 5th grade you know that his fifth grade teacher didn't teach him enough.
if Billy is the class dunce, and his test the following year went up the teacher did her job, even if he started below average and is still below average.
As long as the merit pay is based on each student's score the merit pay is beneficial. You can then track how well a teacher is imparting knowledge in a tested subject. If a whole class has there test scores drop, the teacher is no good or the curriculum is horrid. If a couple scores fall, well that can be attribute to those kids. If the whole class improves, instead of staying the same, you have got a great teacher or a great curriculum and should try to replicate it in other classrooms.
A merit pay plan based on standardized testing could work out fine, if it incorporated a regression model of student improvement that controlled for variation in each teacher's classroom student composition. This type of model was used in Dallas several years back, but I'm not sure how it worked out.
Can someone PLEASE tell me why "teaching for the test" is a bad thing?
The only way "teaching for the test" could be bad is if the teacher somehow had prior knowledge of what questions would be on the test, so that they could repetitively drill their students on those specific questions. I'm guessing they don't know the questions beforehand.
Without knowing beforehand what the questions will be, "teaching for the test" would, of necessity, require the instructor to... (wait for it)... teach the subject material thoroughly!
I think that's the whole point. :-)
Why not a test that covers K through 12 and each year the children take it at the end of the year. If the kids make progress through the year and take it again at the end, it would show whether the teacher is having success. Of course, they'd have to change it up but they would be testing for the same basic knowledge each year.