Leaders in math education in Silicon Valley are concerned that school boards and the public will dismiss the importance of teacher training as a result the dismal results from a much publicized study whose findings were released earlier this month.

“It is ludicrous to conclude that professional development is not important,” said David Foster, executive director of the Noyce Foundation-backed Silicon Valley Math Initiative.

Data is overwhelming showing that effective teaching produces student achievement, he said.  “I am worried that parents will see professional development as expensive and not worth it.”

Certainly, the Middle School Mathematics Professional Development Impact Study won’t discourage budget cutters from eliminating what’s left of school districts’ training money. The study of  middle school math teachers found that teachers who were taught math content and pedagogy and then were coached over the course of a year showed barely any additional math knowledge than teachers who didn’t get the 67 hours of training. And their students didn’t perform any better on standardized tests than other students.

Large-scale, randomized controlled studies are expensive, one reason why this study, funded by the federal government’s Institute of Education Sciences, drew so much attention. There’s not much data like it. The training on rational numbers – fractions, percentages, proportions –  was led by two large training providers, America’s Choice and Pearson Achievement Solutions. It was longer than the training many districts offer, with a three-day summer institute, follow-up seminars during the year and some one-on-one coaching.

But the fact that the program was year-long could be one reason that student scores were not immediately affected, noted Julie Dunkle, program manager for Intel Math, an intensive, 80-hour middle school math training in which Intel has invested heavily. If there were positive effects, they’d more likely show up the second year, after teachers had reflected on and incorporated the knowledge and practices, she said. (The Impact Study is actually a two-year study, so there will be follow-up and a final report.)

Certainly the failure to improve teachers’ knowledge was a problem. The teachers answered 50 percent of questions in a pre-test exam and only 54 percent afterward. Since many middle school math teachers didn’t major in math in college, perhaps the half-wrong answers, depending on the difficulty of the test, while alarming, wasn’t all that surprising.***

That’s why Intel Math, which I’ve written about, focuses on providing in-depth knowledge of math concepts and less on pedagogy. And teachers get an intensive, five-day dose in the summer ­– important to break down teachers’ mental blocks and mindsets. The program helps teachers look at alternative ways students can approach problem solving.

Post-testing of Intel Math teachers in California showed between a 16 and 21 percentage point gain in content knowledge in 2009, according to a WestEd study. But there hasn’t been research into subsequent student test scores, Dunkle said, for lack of study money.

What the federal study failed to do was look at the quality of the training ­– whether the instructors connected with teachers and whether teachers were engaged.

Foster went further, criticizing math training offered by publishers, the most common form of  professional development that teachers receive in California and elsewhere. While hardly  impartial –  the Silicon Valley Math Initiative is a more expensive competitor – Foster said publishers’ primary goal is to sell books, not do the uncomfortable work of challenging teachers’ methods and assumptions.

Foster cited impressive student gains ­– 26 percentage point difference in STAR test results, compared with a control group – from a three-year professional development program in San Mateo County that followed students from sixth grade through eighth-grade algebra. But it was a small-scale study compared with the federal project.

Stanford School of Education Professor Hilda Borko, whose expertise is in middle school professional development, confirmed that smaller-scale studies have confirmed better results than the Math Impact Study found. Findings of larger studies will come out in the next few years that will help define effective training. It’s important learn from but not to overinterpret the findings of one study, she said.

*** Another study, also released this month, revealed that aspiring American math teachers, nearing the end of teacher training, performed about average, compared with peers in other nations, on a test designed to test their knowledge. Future elementary teachers were on a par with  those in Germany, Norway and Russia but below high-achieving Singapore and Taiwan. Future American middle school math teachers fared worse, according to an EdWeek account. The study supports the argument for recruiting teachers with stronger math skills or hiring math specialists in elementary and middle schools.

By John Fensterwald on April 30th, 2010