The great teacher serves as a hero but never, ironically, as a lesson

At last, though, the research about teachers' impact has become too overwhelming to ignore. Over the past Cheap Rosetta Stone Software year, President Barack Obama and his education secretary, Arne Duncan, have started talking quite a lot about great teaching. They have shifted the conversation from school accountability the rather worn theme of No Child Left Behind, President George W. Bush's landmark educational reform to teacher accountability. And they have done it using one very effective conversational gambit: billions of dollars. Thanks to the stimulus bonanza, Duncan has lucked into a budget that is more than double what a normal education secretary gets to spend. As a result, he has been able to dedicate $4.3 billion to a program he calls Race to the Top. To be fair, that's still just a tiny fraction of the roughly $100 billion in his budget (much of which the government direct-deposits into the bank accounts of schools, whether they deserve the money or not). But especially in a year when states are projecting $16 billion in school-budget shortfalls, $4.3 billion is real money. This is the big bang of teacher-effectiveness reform, says Timothy Daly, president of the New Teacher Project, a nonprofit that helps schools recruit good teachers. It's huge. Despite the perky name, Race to the Top is a marathon and a potentially grueling one; to win, states must take a series of steps that are considered radical in the see-no-evil world of education, where teachers unions have long fought efforts to measure teacher performance based on student test scores and link the data to teacher pay. States must try to identify great teachers, figure out how they got that way, and then create more of them. This is the wave of the future. This is where we have to go to look at what's working and what's not, Duncan told me. It sounds like common sense, but it's revolutionary. Based on his students' test scores, Mr. Taylor ranks among the top 5 percent of all D.C. math teachers. He's entertaining, but he's not a born performer. He's well prepared, but he's been a teacher for only three years. He cares about his kids, but so do a lot of his underperforming peers. What's he doing differently? One outfit in America has been systematically pursuing this mystery for more than a decade tracking hundreds of thousands of kids, and analyzing why some teachers can move those kids three grade levels ahead in one year and others can't. That organization, interestingly, is not a school district. Teach for America, a nonprofit that recruits college graduates to spend two years teaching in low-income schools, began outside the educational establishment and has largely remained there. For years, it has been whittling away at its own Rosetta Stone Arabic assumptions, testing its hypotheses, and refining its hiring and training. Over time, it has built an unusual laboratory: almost half a million American children are being taught by Teach for America teachers this year, and the organization tracks test-score data, linked to each teacher, for 85 percent to 90 percent of those kids. Almost all of those students are poor and African American or Latino. And Teach for America keeps an unusual amount of data about its 7,300 teachers a pool almost twice the size of the D.C. system's teacher corps. Until now, Teach for America has kept its investigation largely to itself. But for this story, the organization allowed me access to 20 years of experimentation, studded by trial and error. The results are specific and surprising. Things that you might think would help a new teacher achieve success in a poor school like prior experience working in a low-income neighborhood don't seem to matter. Other things that may sound trifling like a teacher's extracurricular accomplishments in college tend to predict greatness. Steven Farr is a tall man with a deep, quiet voice. He is Teach for America's in-house professor, so to speak. His job is to find and study excellent teachers, and train others to get similar Rosetta Stone Software results. He takes his work very seriously, mostly because he has seen what the status quo looks like up close. Farr grew up in a family of teachers in central Texas. When he graduated from the University of Texas, in 1993, he had a philosophy degree and an acceptance letter to Yale Law School, neither of which felt quite right. So he deferred law school and joined a new, floundering outfit, Teach for America.