Advanced Search
Three Reasons Paying for Test Scores is a Bad Idea
As you might remember, part of the compromise package last legislative session was the requirement that at least 35% of a teacher's evaluation be based on “student performance,” as most commonly measured in test scores on the Minnesota Comprehensive Assessments (MCAs). This is the beginning of a long-term effort – already playing out elsewhere in the country – to replace seniority with test scores as the determinant of a teacher's pay and retention.
There are three broad reasons why this is a bad idea. I'll go into each in more detail in later posts, but for now it's enough to be familiar with the basics:
Reasons Why Paying For Test Scores is a Bad Idea
Reason 1: It's Wrong in Theory
Many advocates of pay-for-test-scores plans think it will be an effective carrot-and-stick approach to driving teacher improvement. In fact, for cognitively complex tasks (most definitely including teaching) simplistic carrot-and-stick pay schemes actually reduce performance.
Reason 2: It's Wrong in Practice
Other areas have already tried this, and their experiences suggest this approach doesn't work. The facts support the theory, in other words.
Reason 3: It's Wrong in Principle
A major reason for replacing seniority-based policies with test-based policies is that tests are supposedly fairer than seniority. If we're trying to protect the principle of fairness, however, there are plenty of problems with docking pay or firing teachers based on simplistic systems using unreliable data that can be manipulated by those in authority.
Why does this matter in Minnesota? Because ending Last In, First Out (LIFO) policies for laying off teachers will be one of the big fights in state government this year. LIFO has its own problems, but we must ask what's supposed to replace it. Defaulting to evaluation schemes that base 35% or more (and many people would like it to be much more) of a teacher's fate on test scores is a bad idea. We should know the facts and hold our policymakers accountable for honoring them during the debate.
Posted in Education | Related Topics: K-12 education Student Assessment Q-Comp Teacher Assessment
9 Comments
February 2, 2012 at 1:45 pm
Pay for Performance is a misleading name: The mane implies that the teacher is paid for their own performance when in fact it is pay for having a class of high performing students. The research has consistently shown that while teachers may influence about 10% of a child’s test score 90% is highly correlated to income. In other words if you can predict a class’s test scores based on income with 90% accuracy. This is bad news for schools because American poverty rate is rising and school are struggling as a result. But it is good news because if poverty is reduced children’s success in school is guaranteed to rise. In the mean time we need to reward and support teachers who are called to wrok with the poorest and lowest perfroming students and create incentives for teachers to stay in the field and commit to kids over the long haul. Paying for committment, creativity and challenge!
February 2, 2012 at 12:13 pm
I hope people do some serious thinking about how to evaluate teachers. First of all if the MCA test graded everything important for students in that grade it would be awesome to use. But it is far from a complete test. Then what about all of the educational professionals that are not directly tied to a classroom who is taking the MCAs? What would their evaluation look like? Remember the MCAs don’t start until 3rd grade in this state. I also agree that homerooms are not sorted out equally by ability. I enjoy working with children who struggle academically but we rejoice for different improvements. This is not as easy as some people think. I hope teachersa are asked for some feedback before the change happens.
February 2, 2012 at 11:47 am
Chris,
Part of the concern is that there is no clarity on how decisions would be made. A statewide teacher evaluation system (which would presumably be used when making retention decisions) is still in development, but the only clear requirement is that at least 35% of that evaluation be based on “student performance” scores derived from standardized testing.
We’re not talking about directly swapping seniority for test scores, but we are moving closer to a score-based approach. That 35% is upward exposed, and we know that some people would like to see it get much bigger.
Given the ambiguity about what new standards would be used for retention, I want the public and our policymakers to be aware of the perils of relying too heavily on test scores.
February 2, 2012 at 11:22 am
My wife is a teacher. When classes were divided last fall, it turns out that 75% of her students are in the lowest quartile. Test scores from her class are not very good, when compared to all others at that grade level. Does that make her a bad teacher?
February 2, 2012 at 11:11 am
I have been under the impression that the proposals are not to REMOVE seniority, rather to supplement it. I.e. seniority today is THE deciding factor, where the proposals would supplement it with other metrics, like student achievement.
E.g. Teacher A has been here 10 years, Teacher B has been here 8 years, so A keeps their job and B doesn’t. Vs. Teacher A has been here 10 years and B has been here 8 years. Teacher B shows consistently high levels of student growth year over year among their students, and Teacher A doesn’t. However, if both had consistently high levels of student growth, A would have more seniority.
Is this an incorrect scenario?
I just think that conversation is more nuanced than you are making it out to be.
February 2, 2012 at 10:36 am
Nathaniel
Go read “Drive” by D. Pink - there is lots of information in there about how the rewards system - in the long-term - stifles creativity and innovation and instead pulls people towards getting the reward any way they can.
Before I read this book, I was using the rewards system with my grandkids and had no clue as to why - long-term - it was not working. Now I know!
February 2, 2012 at 10:36 am
Instead of giving teachers merit pay for students achievement, why not pay the students and the families? School work would be the students part time job AND might give the border-line students the incentive to do better in school.
February 2, 2012 at 9:35 am
Can you provide some citations to back up your assertions? I’d like to read more about how this hasn’t worked elsewhere, the theory behind this carrot reducing performance, etc. I need more information.
Traci says:
February 2, 2012 at 1:45 pm
Many of you have discussed the problems with using student test scores as a factor in pay and retention of teachers as a problems of “which kids we get.” A very legitimate concern. I would also like to bring attention to an issue at the high school level. In my school, 4/7 of a student’s day is spent in a course which is not tested. Currently we test reading, writing, math & science. So the question becomes, how do we then measure those that teach social studies, computer technology, agriculture, band & choir etc.? Are they not held up to that same 35% requirement? Or - even worse in my opinion as a math teacher - are these other teachers paid based on how well I teach students math? This concept actually causes me to have many sleepless nights.