3 ms·

> pay for results. > The bonus is per student. Something something about Goodhart's Law and cobras... But seriously, think about the second order effects of

> pay for results. > The bonus is per student. Something something about Goodhart's Law and cobras... But seriously, think about the second order effects of a powerful incentive like this. At best, you will likely turn schools into test prep centers where average and advanced students are further ignored (an effect NCLB is already criticized for). Worse is that students able to pass the assessments will realize they are a valuable asset worth $1000 per class by simply failing their pretest and passing their posttest and will figure out ways to trade this with teachers for a cut of the profit or for a grade. Given that trading these test scores is purely positive sum for both students and teachers and that both students and teachers are poor, I think you will see a highly efficient market emerge. And why would schools or states care? They are all making more money!

WalterBright 4y ago
And the current system's have an incentive for success? All of your criticisms of the bonus pay proposal can be easily addressed, as the other ones in this thread. For example, test the kids once a year at the end of the academic year. There goes the incentive to fail the pretest. The advanced students can go into the gifted track, which can be operated differently. > And why would schools or states care? They don't care now. Proof: Seattle got rid of their gifted programs.
- ramblenode 4y ago
  You should check out some teacher forums and see what teachers are complaining about to get an idea of what would help. Classes are too big, teachers are pressured to pass unprepared students to the next level, underperforming students are hard to motivate, there isn't enough time to prepare lessons and grade, etc. There are many areas that could be improved with funding, but you should start at the known problems. Think about this from the perspective of behavioral economics. Teaching is a profession with a high degree of intrinsic motivation given how stressful it is and how little it pays. Paying for performance on some test primarily rewards people who will game the test because that has a better cost/benefit ratio than heroically catching up kids who are performing several grade levels behind and have a difficult home life. Pay for performance tends to damage morale and increases attrition of intrinsically motivated workers (the kind you really want). This is what happens in controlled psychology experiments and it's what happened with NCLB.
  - WalterBright 4y ago
    > game the test How are you going to game this test: 1234 +5678 ----- ????? when the tests can each be unique with randomly generated 4 digit numbers? The way to "game" it is to learn how to add! > it's what happened with NCLB Once more, NCLB did *NOT* pay teachers for performance.
    - ramblenode 4y ago
      > How are you going to game this test: Plenty of ways. Teacher takes test for student, teacher re-scores test after student is finsihed, student is placed in an IEP that gives them unlimited time and "reading assistance" for test, failing students are placed in small classes and the class size of average students is expanded, failing students are moved from normal schoolwork to test prep, test questions are leaked/sold (we're talking about a multi-billion dollar industry here, right?). I didn't make up any of these. This is already happening, and it has nothing to do with positive/negative incentive. It's about test-based education and the natural consequence of trying to control a complex system using a simple metric. It's why education theorists are loathed by teachers--because they swoop in with grand theories of how to fix everything without understanding the system and what is and isn't broken. It's like an outsider thinking a reasonable way to pay a coder is by how many lines of code they write and how many bugs they fix. Or a civil engineer by how many bridges and roads they produce. Because outsiders don't understand the complexity of the system they think the solution must be simple. Since you seem invested in this idea I suggest you go get some input from teachers and see what they say. My guess is you will learn that the problems with public education is multifactorial and that there isn't a simple fix.

WalterBright 4y ago

And the current system's have an incentive for success? All of your criticisms of the bonus pay proposal can be easily addressed, as the other ones in this thread. For example, test the kids once a year at the end of the academic year. There goes the incentive to fail the pretest. The advanced students can go into the gifted track, which can be operated differently. > And why would schools or states care? They don't care now. Proof: Seattle got rid of their gifted programs.

ramblenode 4y ago
You should check out some teacher forums and see what teachers are complaining about to get an idea of what would help. Classes are too big, teachers are pressured to pass unprepared students to the next level, underperforming students are hard to motivate, there isn't enough time to prepare lessons and grade, etc. There are many areas that could be improved with funding, but you should start at the known problems. Think about this from the perspective of behavioral economics. Teaching is a profession with a high degree of intrinsic motivation given how stressful it is and how little it pays. Paying for performance on some test primarily rewards people who will game the test because that has a better cost/benefit ratio than heroically catching up kids who are performing several grade levels behind and have a difficult home life. Pay for performance tends to damage morale and increases attrition of intrinsically motivated workers (the kind you really want). This is what happens in controlled psychology experiments and it's what happened with NCLB.
- WalterBright 4y ago
  > game the test How are you going to game this test: 1234 +5678 ----- ????? when the tests can each be unique with randomly generated 4 digit numbers? The way to "game" it is to learn how to add! > it's what happened with NCLB Once more, NCLB did *NOT* pay teachers for performance.
  - ramblenode 4y ago
    > How are you going to game this test: Plenty of ways. Teacher takes test for student, teacher re-scores test after student is finsihed, student is placed in an IEP that gives them unlimited time and "reading assistance" for test, failing students are placed in small classes and the class size of average students is expanded, failing students are moved from normal schoolwork to test prep, test questions are leaked/sold (we're talking about a multi-billion dollar industry here, right?). I didn't make up any of these. This is already happening, and it has nothing to do with positive/negative incentive. It's about test-based education and the natural consequence of trying to control a complex system using a simple metric. It's why education theorists are loathed by teachers--because they swoop in with grand theories of how to fix everything without understanding the system and what is and isn't broken. It's like an outsider thinking a reasonable way to pay a coder is by how many lines of code they write and how many bugs they fix. Or a civil engineer by how many bridges and roads they produce. Because outsiders don't understand the complexity of the system they think the solution must be simple. Since you seem invested in this idea I suggest you go get some input from teachers and see what they say. My guess is you will learn that the problems with public education is multifactorial and that there isn't a simple fix.