Some teachers to get a sneak peek of new evaluations this week

About one in five city teachers will get a sneak peek on Tuesday about how they might be rated under a new evaluation system.

That’s when the city Department of Education will be sharing the state’s “growth scores” with teachers for whom a score was generated. The scores reflect how well a teacher’s students performed on state math and reading exams last year compared to other students like them and, according to state law, must eventually constitute 25 percent of annual evaluations for teachers who work in tested grades and subjects.

In New York City, about 17 percent of teachers teach fourth or fifth grade or English or math in middle school. They will get their growth score for the 2011-2012 school year Tuesday evening in their Department of Education email, department officials said.

The department has had the information since the end of the summer, state education officials said at a briefing for reporters last month. Principals got the reports last week and are expected to use the scores to help teachers at their school improve, according to Connie Pankratz, a department spokeswoman. But teachers are supposed to get access only to their own scores.

At at least one school, last week’s data upload did not go as planned. Starting Wednesday, when teachers at a large middle school in Brooklyn, logged on to the department’s data warehouse, they could download the effectiveness reports for all of their colleagues. The reports appeared in a long list in a section of the data system, ARIS, called the “Community Space.”

That should not have happened, Pankratz said. “We are working to ensure that only administrators or other authorized staff have access to the growth scores,” she said.

The reports, which the State Education Department generated, show the number of students tested and the percentage who scored above the state average. They also show how many of the students outperformed other students like them, using disability, socioeconomic status, proficiency in English, and prior test score history to break students into smaller groups. A poor fourth-grader with special needs whose native language is English and who scored at the lowest level on his third-grade tests would have his scores compared only to other students who fell into the same categories.

Compiling their students’ adjusted performance gives teachers an “adjusted mean growth percentile” that is then converted into a  score between 0 and 20, representing what would account for 20 percent of the teachers’ annual rating under an evaluation system that conforms to the state’s new evaluation law. Teachers with a rating between 9 and 17 will fall into the “effective” category.

Those data points will be made public, in aggregate form, for each school early next year.

In addition, a chart on each teacher’s growth score report shows the performance of students with disabilities, English language learners, poor students, and students with particularly high or low test scores in the past. (The information will be displayed only for groups that contained 16 or more students.) The subgroup information will be used for school accountability, state officials said, but will not be reflected in the teacher effectiveness scores that are released publicly next year.

For the 83 percent of city teachers who do not work in tested grades or subjects, the state will calculate student growth using a different set of measures. Those measures, known as Student Learning Objectives, require districts to choose state, homemade, or third-party assessments that can be used to calculate how much students have improved over the course of a school year.