"for the children".
From: Klein Joel I.
Sent: Monday, October 25, 2010 3:57 PM
Subject: Teacher data
As you have likely heard or read, several media outlets recently issued Freedom of Information Law (FOIL) requests to the City, requiring the Department to share the Teacher Data Reports we provide schools and teachers in grades 4 through 8 each year. These reports use a method called “value-added data” that seeks to predict student performance based on factors outside of a teacher’s control (high levels of poverty, for example), and then determines whether a given teacher’s students exceeded or fell short of these predicted examination scores (teachers may always access their reports at http://schools.nyc.gov/Teachers/TeacherDevelopment/TeacherDataToolkit/GetYourReports/default.htm). By controlling for factors beyond a teacher’s control, it is the fairest system-wide way we have to assess the real impact of teachers on student learning. And while the City’s particular value-add method is not etched in stone, this is why the State passed legislation this spring, endorsed by the teachers’ unions, committing to using value-added data for all teachers. It is also why value-added data is increasingly being used throughout the nation as part of a comprehensive system of teacher evaluation.
In the past we have provided the numeric value-added data to the press with no indication of the identity of individual teachers. I am writing to you today because media outlets, prompted by similar data being published by the Los Angeles Times, have requested the names of individual teachers, not just the statistics. As it is the City’s legal interpretation that we are legally obligated to provide the media this information, it is our intent to provide the data as requested.
In the time since we informed the UFT that we intended to comply with the FOIL request, the union has sued the City to prevent the release, and we have agreed to delay any release until at least November 24, when a court hearing will be held. So no data have yet been released. But I want to make sure that, as you read about these events in the newspapers, you understand the circumstances and you understand my view on the issue overall.
Our most important task is to ensure that every one of our students has a great teacher. It is critical, therefore, that when we have indications of a teacher’s proficiency, we use that indication to do what’s right for kids. One indication will never tell the whole story, and sometimes it is hard to discern definitive evidence from data alone—such as with a teacher who is “average” according to these numbers, for example. But where teachers have performed consistently toward the top or the bottom, year after year, these data surely tell us something very important. Namely, we need to retain and reward the great teachers, and we need to develop the low-performing teachers. And those who don’t improve quickly need to be replaced with better-performing teachers.
Secretary Arne Duncan last week said it best when he said, “I give New York credit for sharing this information with teachers so they can improve and get better.” More than anything, these data demonstrate that we need a better, more comprehensive system of evaluation than the one we have now. That’s why the State legislature and the unions supported an evaluation system that uses value-added data. Now it’s time that the DOE and UFT together build a new system that gives teachers an honest sense of how well they’re doing and how they can improve.
In the end, this is about real people. On one hand, for too long, parents have been left out of the equation, left to pray each year that the teacher greeting their children on the first day of school is truly great, but with no real knowledge of whether that is the case, and with no recourse if it’s not.
But this is also about teachers. They take on the hardest work there is, and they deserve our respect. If anyone sees these data as an opportunity to scapegoat public servants, that is a mistake. Doing what’s right for children means making hard decisions; it has nothing to do with personal attacks.
We’ve made huge strides for our kids over the last eight years. That’s because we’ve been willing to face hard facts. It’s also because we have made kids’ best interests our shared priority. My hope is that we approach this issue with both of those thoughts in mind, ensuring fair treatment for adults, but always keeping children first.
Joel I. Klein
Why teacher scores should be released
|Mike Bloomberg, Nicole Seligman (Mrs. Klein), Joel Klein, Diana Taylor|
Last week, the New York City Department of Education planned to release Teacher Data Reports, which include the names of more than 12,000 city teachers and what are known as their “value-added” scores.
The release of these data reports — which tell us which teachers are contributing the most (and the least) to their students’ achievement — raises complex issues. While they are provided every year to principals and teachers directly, they have never before been released with teacher names to the public, and the United Federation of Teachers has gone to court to block their release.
First and foremost, we believe that the public has a right to this information under the Freedom of Information Law. But we also strongly disagree with the UFT’s argument that the public isn’t smart enough to understand this information.
So what is value-added data and what can it tell us?
It starts with the idea of fairness. Statisticians look at factors that have historically affected student achievement, such as high levels of poverty, and create a picture of each child’s background that enables them to predict how well a child is likely to do. They then see whether the child — or the whole class — did better than or worse than was predicted. The point is to remove all of the factors teachers can’t control.
For example, let’s say Adam is a first year, seventh-grade math teacher. His students are predicted to score an average of 3.22 on the state math test. But instead, his students score a 3.3, meaning he added .08. Leslie, meanwhile, also is a first year, seventh-grade math teacher in the same school. Her students also have a predicted proficiency score of 3.22, but they only score a 3.17, meaning she subtracted .05. Both teachers are then compared to peers across the city who are first-year, seventh-grade math teachers, and it is determined that Adam’s value-added score is in the 85th percentile, while Leslie’s is only in the 33rd.
It’s a quantitative way to show what many of us have argued for years — not all teachers are equally effective. If one teacher is found to be consistently high performing, don’t we want that teacher collaborating with others? And, in turn, if one teacher is found to be consistently low performing, don’t we want to help that teacher improve, or move to replace him or her?
No one believes value-added data tell the whole story of a teacher. But it provides a valuable window into teacher effectiveness, which is why we have used and will continue to use the data when we determine whether to award lifetime tenure. And New York state recently passed a law, supported by the UFT, mandating the use of teacher effectiveness data in teacher evaluation systems. In New York, value-added data may comprise 25% of a teacher’s overall evaluation; in states like Colorado and Louisiana, it’s up to 50%.
We aren’t naive about the impact this release could have on our teachers, which is why we hope that no one misuses the data or views it as an opportunity to scapegoat teachers. Our teachers deserve the utmost respect.
But these are public schools and public tax dollars. As Education Secretary Arne Duncan said, “Parents and community members have the right to know how their districts, schools, principals and teachers are doing. It’s up to local communities to set the context for these courageous conversations, but silence is not an option.”
Joel Klein is the chancellor of New York City’s schools.