Re: performance based payabfarrerMarch 10 2009, 23:30:54 UTC
while I strongly believe that there needs to be some form of performance based pay for teachers, you're right that a measure such as grades cannot be an effective basis for such pay.
In the current system, a teacher's base pay is based entirely on two factors, the number of years they've been teaching, and the number of (often meaningless) pieces of paper that they have. And, of course, by pieces of paper, I mean degrees and the like, which may or may not come from a school that's essentially a diploma mill. (I'm looking at you, university of Phoenix)
I think, and this will be nearly impossible to implement, this is a case where subjectivity needs to take play a role, rather than straight out objectivity. It's not that hard to see who among the teachers in Sharon are good teachers, who are great teachers, and who are taking full advantage of their professional status in the union, and the near impunity it gives them to do a half-assed job. If nothing else, the students could tell you, but then you run the risk that students will rate a teacher poorly simply because the teacher actually makes the class a challenge.
And therein lies another problem with basing pay on grades, for the most part, besides things like the MCAS, grading is based pretty much entirely on what the teacher presents to test the students. It would be trivial for any teacher to simply make the tests easier to ensure that their students perform well, removing the challenge and doing as much harm to education as the MCAS/No Child Left Behind programs do. (If you're wondering what damage those do, ask any teacher at the levels where MCAS testing is done, and you'll get an earful about how they have no time to teach anything but the MCAS test itself). And, of course, if you're paying based on something like how well the students do on a standardized test, then teacher compensation will be as geographically, racially, and economically biased as the tests themselves are. The very teachers who are taking on the most difficult students and schools, and are being pushed to perform the most, will be compensated the least.
And then where will our educational system be?
No, the first thing that needs to be done is to level the playing field. Rather than the education spending being left up to the towns, all of the spending should be done equally nationwide, or at least regionally. Spending, on a per student basis, should be kept at least equitable. No more small, expensive rich towns spending 2-3 times per student what inner city districts do, all kids should have a chance at a good education, not just those whose parents can afford to live in a town with good schools or can afford to send them to private schools.
Re: performance based paytoastedspleenMarch 11 2009, 19:39:32 UTC
Rather than the education spending being left up to the towns, all of the spending should be done equally nationwide, or at least regionally.
The numbers disagree with you. The cost of education is not directly related to quality. You are assuming that Boston spends less per student than a small town with a better school system like Sharon. Actually Boston spends $16466.79 per student and Sharon spends $12203.98. Only 23 districts spend more per student than Boston. You can argue that other districts spend it better, but what is better?
You can't find a way to give students a "good education" unless you can define what a "good education" is. And that seems to be impossible to define because it is so subjective. Not only that, but a good education is probably different for each student.
Re: performance based paycochise15March 12 2009, 01:46:15 UTC
I'm just guessing, but the reason Boston has to spend more per student than Sharon is because of programs like busing, which they are actually thinking of getting rid of because it costs so much, that is a legacy of discrimination aimed at a certain segment of the population over a long period of time, one that Sharon can't compare to numbers-wise, and the deeper kind of poverty Boston experiences than Sharon. When you think of the number of kids that attend Boston Public schools vs. the numbers that attend Sharon, and the fact that there is only a four thousand dollar difference in spending per year, I'd think that adjusted for the differing factors, Sharon probably spends more on their kids than places like Boston or Brockton or Fall River does.
But I do agree with you that the idea of a "good education" is based more on the individual student, more than some sweeping idea that what works for everyone works best. That's why I think smaller class sizes are really the key. They could allow teachers more room to recognize the individual needs of the students in their classes.
In the current system, a teacher's base pay is based entirely on two factors, the number of years they've been teaching, and the number of (often meaningless) pieces of paper that they have. And, of course, by pieces of paper, I mean degrees and the like, which may or may not come from a school that's essentially a diploma mill. (I'm looking at you, university of Phoenix)
I think, and this will be nearly impossible to implement, this is a case where subjectivity needs to take play a role, rather than straight out objectivity. It's not that hard to see who among the teachers in Sharon are good teachers, who are great teachers, and who are taking full advantage of their professional status in the union, and the near impunity it gives them to do a half-assed job. If nothing else, the students could tell you, but then you run the risk that students will rate a teacher poorly simply because the teacher actually makes the class a challenge.
And therein lies another problem with basing pay on grades, for the most part, besides things like the MCAS, grading is based pretty much entirely on what the teacher presents to test the students. It would be trivial for any teacher to simply make the tests easier to ensure that their students perform well, removing the challenge and doing as much harm to education as the MCAS/No Child Left Behind programs do. (If you're wondering what damage those do, ask any teacher at the levels where MCAS testing is done, and you'll get an earful about how they have no time to teach anything but the MCAS test itself). And, of course, if you're paying based on something like how well the students do on a standardized test, then teacher compensation will be as geographically, racially, and economically biased as the tests themselves are. The very teachers who are taking on the most difficult students and schools, and are being pushed to perform the most, will be compensated the least.
And then where will our educational system be?
No, the first thing that needs to be done is to level the playing field. Rather than the education spending being left up to the towns, all of the spending should be done equally nationwide, or at least regionally. Spending, on a per student basis, should be kept at least equitable. No more small, expensive rich towns spending 2-3 times per student what inner city districts do, all kids should have a chance at a good education, not just those whose parents can afford to live in a town with good schools or can afford to send them to private schools.
/end rant
Reply
Although I am less opposed to charter schools then my husband.
Reply
Reply
The numbers disagree with you. The cost of education is not directly related to quality. You are assuming that Boston spends less per student than a small town with a better school system like Sharon. Actually Boston spends $16466.79 per student and Sharon spends $12203.98. Only 23 districts spend more per student than Boston. You can argue that other districts spend it better, but what is better?
You can't find a way to give students a "good education" unless you can define what a "good education" is. And that seems to be impossible to define because it is so subjective. Not only that, but a good education is probably different for each student.
Reply
But I do agree with you that the idea of a "good education" is based more on the individual student, more than some sweeping idea that what works for everyone works best. That's why I think smaller class sizes are really the key. They could allow teachers more room to recognize the individual needs of the students in their classes.
Reply
Leave a comment