How to measure the worth of Los Angeles math teacher Kyle Hunsberger?
The teacher at Johnnie Cochran Jr. Middle School works 60-hour weeks, constantly searches for new teaching ideas and makes every minute count in class. During a fast-paced review of square roots and perfect numbers, he punctuated explanations with jokes, questioned his students to check their understanding and engaged them in group work.
His principal, Scott Schmerelson, praises him as a leader who heads the math department and started a campus program to give struggling students extra help.
Some of his students say he's the best math teacher they've ever had — a caring, funny mentor who explains well, pushes on homework and most of all believes in them.
"He always tells us nothing will stop us from learning and nothing will stop him from teaching us," said Edwin Perez, a gregarious 12-year-old, as three of his classmates nodded.
Yet, according to a key measure of teacher effectiveness used by the Los Angeles Unified School District, Hunsberger is average.
Two years ago, he said, he was rated above average. Then last year his ratings fell. He doesn't know what changed and there's nothing in his scores that will tell him.
The rating "didn't tell me anything about how I can get better at teaching [weaker] students," Hunsberger said. "The truth is, I don't know and I would love to know."
Hunsberger isn't the only instructor questioning the results of the Los Angeles school system's new approach to measuring teacher effectiveness. Academic Growth Over Time, as the district calls it, is based on students' progress on standardized test scores. The method estimates how much teachers added to — or subtracted from — their students' academic performance.
Whether it is a fair, accurate and useful assessment of educators is a heated issue in the nation's second-largest school system. L.A. Unified is under court order to use test scores in teachers' reviews by December, and officials are in negotiations with the teachers union.
United Teachers Los Angeles bitterly opposes the ratings as too unreliable for use in firing, tenure and other high-stakes decisions.
School districts in more than half the states have added students' test scores along with other factors to their teacher reviews, a direction promoted by the Obama administration.
L.A. Unified began giving teachers their scores two years ago for informational purposes only.
But it is now pushing to use it in a new teacher-evaluation system, along with classroom observations, student and parent feedback and contributions to the school. About 700 teachers and administrators from 100 schools volunteered to test the new observation portion last year.
That will give teachers like Hunsberger specific information about where to improve and how.
He questions whether his ratings were higher two years ago because he had a class of "rock star" algebra honors students, but fell last year when he had less-skilled students, many of them learning English.
"I did my best. I tried things. I worked hard," said the 30-year-old New York native who sports a neat beard and a receding hairline that he jokes about with students.
Hunsberger's questions recently deepened when he noticed a graph on the district's website that seemed to show that schools with stronger students have higher growth. It coincided with his experience that honors students were easier to push forward.
"I have to be reassured that I don't have to lobby for honors students," Hunsberger said. "I have to know that I have a shot at a good evaluation if I teach lower-performing kids."
But the rating system controls for outside factors that could influence growth, such as past test results, gender, race, income and English ability. Those controls give every teacher an equal shot at good performance ratings regardless of their students, according to Noah Bookman, the district's director of performance management.
"The important piece for people to understand about [Academic Growth Over Time] is that it allows us to level the playing field," Bookman said.
Bookman also said that the graph questioned by Hunsberger shows not that stronger students boost scores, but that good teachers produce stronger students at any level — an outcome possible for all educators, he said.
"Teachers with low, middle and high-achieving students have the same opportunity to demonstrate growth as each other," Bookman said.
Hunsberger understands the math but is not sure about the claims.
He was an early champion of the new system. The issue was personal: In 2010, Cochran's years of low test scores resulted in placement on the district's list of campuses eligible for takeover by charter schools or other groups with a credible improvement plan.
The South Los Angeles campus of 1,300 students, nearly all of them low-income African Americans and Latinos and a third who are learning English, consistently ranked in the state's lowest 10% of middle schools. Only about a quarter of students were at grade level in reading and math. The school scored in the low 600s on the Academic Performance Index, a 1,000-point achievement measure based on standardized test results that does not control for outside influences.
But down in the trenches, Hunsberger felt the picture was not that bleak. When L.A. Unified released schoolwide scores for the first time last year, the results confirmed his instincts. Although Cochran's student achievement was low, its rate of academic growth was significantly higher than the district average in English, algebra, science and social studies in 2010-11.
He and a colleague, Rustum Jacob, found "huge inconsistencies" at other schools between the state's API achievement scores and the district's scores. They urged school officials to include district ratings to identify schools for the takeover list. Last October, the district did just that.
"I very much embraced the idea that AGT [Academic Growth Over Time] represented a far better measure of a school's impact on student outcomes than API," Hunsberger said.
He became a bit of an evangelist. He agreed to test the district's new evaluation system, despite the union's urgings that teachers not participate. He joined a new group of educators, Teach Plus, whose proposed evaluation plan would count the district scores for a minimum 10% of a teacher's ratings.
But he and other instructors are still concerned — even those who embrace the idea of using objective student achievement measures in their evaluations.
Lisa Alva, a Roosevelt High School English teacher, said her score for last school year was based on 12 students and wonders how that can be valid or fair. Philip Gerlach at Markham Middle School got sterling scores but said they were skewed downward by counting students he had for just two months and don't measure his strongest suit — teaching writing.
Sujata Bhatt, another highly rated teacher who taught fifth grade at Grand View Boulevard Elementary, said the formula needs revision to account for different ranges of poverty, for instance, or English fluency.
Hunsberger's colleague, English teacher Daniel Badiak, said his below-average scores last year have pushed him to "teach to the test" more this year. Time for work on what his seventh-graders most need — basic lessons on where to put periods and question marks, for instance — is being eaten up by drilling on vocabulary that might appear on the state test, he said.
Hunsberger said he still thinks the use of the scores is "the right idea" but he intends to keep asking tough questions.
"It's got light years to improve," he said.