A 2012 study about teacher retention, "The Irreplaceables," found that of 90,000 teachers studied in four large, diverse urban school districts, about 20 percent are "Irreplaceables" — teachers who are so successful that they are almost impossible to replace. It's a number that happens to correlate to the upper 20 percent bracket in Atteberry's study — the teachers who start out strongest and stay strongest.
"Knowing the power of great teachers, one would expect schools to be sharply focused on keeping far more of their best teachers than their lower performers," a summary of the study said. "Instead, they retain all teachers at strikingly similar rates: and about half of all Irreplaceables leave within their first five years."
Reasons cited in the study include principals who make too little effort to retain their best teachers or remove low-performing teachers, and policies that give principals and district leaders few incentives to change their ways.
That study suggested that retaining top teachers should be a major priority for school systems. To do that, improve working conditions in schools, pay the best teachers what they are worth, and protect them during lay-offs, it said.
It's important for schools districts to implement meaningful teacher evaluation systems that are rigorous across multiple measurements, said Michael Hansen, an education researcher at the American Institutes for Research.
"If an evaluation system doesn't have any teeth, or if it's poorly implemented — just checking boxes rather than having meaningful information for decisions to be based on — that really has the potential to undermine the school system altogether," Hansen said.
Documenting poor teacher performance is a critical area of failure, Hansen said, especially in high-need schools.
"In those schools, you often have a revolving door of principals, and of leadership for interventions," he said. "If want to remove a poor performer, you need to document for two or three years. It just doesn’t happen in many school districts."
Identifying those poor teachers quickly, while their hiring is still probationary, is a simpler process, and that's one of the areas in which Atteberry's new study about early career teachers could make a difference.
She cautioned, though, that conclusions of the study she co-authored should be implemented carefully. In her study, the teachers who failed to improve belonged to a specific group — fourth- and fifth-grade teachers in New York City under a certain set of circumstances. The limits of the study don't prove inconclusively that teacher effectiveness is predictable, she said.
"I don't want to say that these teachers couldn't improve under other circumstances — had they been supported differently, had different training, been assigned to different schools. You have to be careful in using this during personnel decisions," she said.
Atteberry said the paper is the first part of a larger project that is still under way — one that will look even harder at who improves and who doesn't, and tries to figure out why.