Michelle, I'll kick this off with why you think it's necessary to use student test scores in teachers evaluations - and whether parents should have access to those evaluations.
Teacher quality is the single most important in school factor that determines student achievement growth. That means that we should be focused on teachers and their effectiveness.
My friend is a 2nd grade teacher. She has a very good, calm and authoritative demeanor: And she is clearly considered a good, competent and motivated teacher. More and more she has lower achieving, autistic, difficult and/or challenged kids assigned to her class. This is because the administrator's know she can handle them better than some other 'less patient' teachers. As a result she struggles with her test scores. She is terrified of performance based ratings for this reason. How can a good teacher be protected from poor administration- wherein a less calm and perhaps less motivated teacher gets rewarded and a good one gets punished?
Harry, this is a great question. Good value added growth models are designed to look specifically at the value that the teacher adds to student achievement, controlling for factors outside of their purview. For example, VAM models typically control for SPED status, ELL status, poverty, attendance and other characteristics that impact a student’s learning that are not within the control of the teacher. Therefore, your friend's evaluation would take into account who her students are.
Ms. Rhee,
I am generally supportive of quantitative metrics for teacher evaluation (and of you) but this particular expression seems a little suspect. What do you think the net positive results will be of NYC releasing this information? Do you think that this actual exercise is productive for the education system?
I think it's incomplete. Teacher evaluation should be based on looking at a teacher's effectiveness through multiple lenses. Though VAM scores are one piece to it, they need to be put in context of the whole picture of a teacher's effectiveness.
Good question. In New York City, they developed this methodology for only teachers in grades 4,5,6,7,8 whose students take NY state math & English tests...
But more tests are being developed for other subject areas, as well as other ways to rate teachers in subjects that don't lend themselves to standardized tests easily...
How do you factor in other variables like class size, available materials, support for discipline? I think there are a lot of factors that would make a "principals pet" teacher from the same sorority more successful than the one in a poorly lit noisy room next to the boilers.
Current evaluation systems that are completely subjective and based on the opinion of one person (the principal) are not fair to teachers. They allow for the "pets" to get better evaluations. The current move toward more objectivity through the use of VAM (taking into consideration student and school characteristics), etc ensure that teachers can be evaluated more fairly through a more objective model.
Accountability should sit at every level. Yes, teachers should be held accountable, but so should students, parents, principals and administrators. You can't have a great school without a great principal so we have to ensure we're doing everything we can to have a highly effective principal leading every school...
If you look at the data the main reason teachers decide to leave their jobs is the culture that the principal creates in the school. In order for teachers to have confidence in the reforms that are being put in place, there have to be great principals who are leading the way in those schools.
Please comment on the fact that most teacher evaluations contain zero input from parents or students. What good is an evaluation that ignores this input?
I love the idea of using parent and student input in teacher evaluations. Kids tell me all the time that they feel they can assess teacher quality really well. Research is also showing that student assessments are highly correlated to student achievement growth, which is promising!
Hi - 9th grade English teacher here - Should we look at completely revamping the current salary schedules most schools use? It seems to me that the only true measure is time on the job. I don't think all subjects demand the same amount of training or time on the job (planning, grading etc.).
I do think we need to re-conceptualize how we pay teachers. It should be based on effectiveness. I'm proud that during our time in DC we made DC teachers some of the highest paid teachers in the country! Our nation's highly effective teachers should be paid more money than professional athletes in my opinion...
Will anything concrete be done with these teacher scores ? Are they actionable items in the sense that the lowest scoring teachers will be removed, tested, or reeducated themselves, or will the information just be made available and that will be the end of the story ? What is the next step and how long will that take to implement ?
In New York, the data you see in our database will not be used in most hiring and firing decisions. Principals were encouraged to use the data when deciding whether to give some teachers tenure...
So-called "student performance" will count for up to 40% of the evaluation. And if teachers are ineffective on that part, they get an ineffective rating overall. To simplify a bit, two ineffective ratings mean teachers can be fired more easily...
Whay actual successes occurred in the DC area which support this model and further qualify you to speak on thia issue besides being in the role you had?
In DC I think we put in place a very effective teacher evaluation system that allowed us to look through multiple lenses at teacher effectiveness. It enabled us to identify the most highly effective teachers in the district and recognize and reward them accordingly. It also allowed us to identify those who were struggling and give them additional support. There was a recent piece that showed that a lot of highly effective teachers felt incentive to stay in the system because of that recognition and the new pay structure.
Because VAM scores are derived from comparing data from different teachers, the model inherently sets up a culture of competition where the goal is not necessarily student learning, but rather “doing better” than other teachers. This can lead to perverse cultures of teaching and learning where things like teaching to the test happen. Does Rhee see this as a flaw in the value-added model, and would she like to see us move towards using data that is more about the individual achievements teachers make with their students (rather than using comparative data)?
I think there should be both an individual and team component to a teacher's evaluation. That's what we did in DC so there would be incentive for a teacher to focus on the achievement in their classroom and their neighbor's classroom as well. VAM doesn't have to result in a bad culture. We can manage against that.
Ms. Rhee, I hope you have had a chance to read what Diane Ravitch has to say about publishing teacher evaluations based on test scores. Not only will teachers be unfairly stigmatized, since all admit that there are problems with student test scores accurately measuring teacher quality, but teachers will be pressured to do test prep and little else to preserve their ratings and job security. Goodbye to wonder and creativity. No other industrial nation emphasizes student testing the way that the US does. Testing for assessment is useful, but high stakes tests which punish students, schools, and teachers are not useful and will only cull from the ranks teachers who want to truly teach and not speak off a scripted and dull lesson. This is a terrible idea. After twelve years of No Child Left Behind and its over-emphasis on punitive tests, our children's education has not improved, and the achievement gap between high and low-income kids has not closed. Let's try a new plan. This one is not working.
The research is pretty clear on this. Teachers who "teach to the test" have students who don't do better on the test. In fact, it's children who have access to a broad based curriculum that do well academically. I think it's important to ensure that educators understand this dynamic so that the tests can be put into context!
A normally-rational friend of mine is a HS physics teacher and by all accounts a very good one. Yet he worries that any rating system will be highly qualitative (dependent upon administration personnel & their relationship with teachers). He is also very concerned that without tenure the most expensive teachers--the ones who have the most experience--will be the first to go in any budget tightening. Any suggestions as to how to allay his fears?
Yes, this concern is one that can be addressed. In a number of states that we've worked to get rid of seniority based layoffs, we've also made it illegal to base these decisions based on how expensive a teacher is.
The system that you set up in Washington, D.C., offers huge salary increases to the teachers who get the highest job-performance ratings -- but only if they agree to reduced job security. The result is that the district gives its best teachers the LEAST job security and its worst teachers the most job security. Why do you want the best teachers to have the least job security?
It's not that the most effective teachers had the least job security. Those with the least job security were actually the least effective teachers. The system that we set up only ensured that effective teachers (once given that rating) had to continually show that they were effective.