Thursday, Aug. 13, 2009 | San Diego Unified is crafting a new, more rigorous way of judging principals that would include whether student test scores, dropout rates and attendance had improved on their watch, a change that could prove controversial with their new union.
Superintendent Terry Grier said the new evaluation has "tremendous potential," and praised it for linking student success to the fates of the principals who lead their schools. The new evaluation also includes dozens of other criteria beyond testing and would happen more frequently than in the past.
It will be the first major issue for the nascent principals union to hash out in negotiations. Tying test scores to the fates of employees is contentious among educators, some of whom fear that tests can be an incomplete and sometimes misleading measure of school progress, depending on how the data are crunched. The Obama Administration is now locked in a grueling debate with teachers unions over whether schools should evaluate and reward teachers based on test scores.
Evaluating principals using test scores may be an easier step for schools to take, and many have already taken it. Experts say that even when evaluations are mum on the subject of test scores, low scores may implicitly be part of the calculus when principals are let go or transferred to other schools. Principals' work rules are often looser than teachers' contracts, making it simpler for school districts to change them without a fight. Even principals who are unionized have signaled that they are open to the idea.
Before San Diego schools can hold their principals to the new standards, the plan must pass muster with the newly formed principals union. That union has yet to formally discuss the proposal, but its leaders are already uneasy with some of the new criteria and the sheer scope of the new evaluation.
"How the heck are you going to be held accountable for all of those things at the same time? And who will want to go to a lower performing school?" asked Jeannie Steeg, executive director of the Administrators Association. Yet she added that compared to the brief and vague evaluation used in the past, "Anything's an improvement."
San Diego Unified started reshaping its principal evaluations last year with the help of an outside consultant after an internal report found that it was delinquent in evaluating employees, sometimes not doing so for years. The old form for judging principals was only four pages long and included only nine criteria in three categories that could be rated "unsatisfactory," "requires improvement" or "meets standards."
More than two dozen principals piloted the new evaluation this school year, putting themselves through the new wringer. It holds principals to dozens of standards, ranging from how savvy they are about curriculum to whether they make good use of their budgets. It also judges them on how well they know their staff, how accessible they are to parents and how they use new research. And it includes a wider range of ratings, from "not demonstrated" to "distinguished," along with the criteria for achieving them.
It wasn't until after the evaluation was first piloted that Grier decided to add test scores into the mix. He said the school board urged him to do so. The idea is that principals should have to show that children are doing better in school thanks to their leadership and be held accountable if children are floundering. Principals will then have a stronger incentive to show that children are making real progress. Jean Williams, vice president of the nonprofit that helped develop the new evaluation, said the evaluation was tested again after the new criteria were added and was praised by principals.
"Why would you not want how well kids do in school to be tied to how well adults do? Including the superintendent? Including the board of education?" Grier asked.
Though the principals union has yet to take a stance on the evaluation, Steeg said including test scores is "a great concern" for employees. When an area superintendent tried to demote some principals based on their test scores two years ago, her group argued that California law actually bars schools from using standardized test scores to evaluate both teachers and principals. If test scores are used in a ham-handed way, principals may be penalized for changes beyond their control, she said.
For instance, if more English learners poured into a school one year, its English scores might seem to drop that year, even if the children were learning and making huge gains while at the school. If a private school closed and the same school was flooded with wealthier children, who tend to do better on standardized tests, its scores might seem to leap, even if those students stagnated at the school. And in neighborhoods where families move frequently, children may have spent little time at a school before being tested, making it unclear whether their scores are a good reflection of that school.
"It's hard to use that information in a fair and effective manner," said Thomas Toch, executive director of the Association of Independent Schools of Greater Washington.
Reformers who want to use test scores to evaluate teachers and principals say those problems can be avoided by following the same children over time to see how much progress they make. That would measure how much the students at a school actually learned, not how they compare to different children who attended the school the previous year. Statisticians say it gives schools a better measurement of how much the school or the individual teacher is impacting their test scores.
It is a kind of data crunching known as "value added," and it is often tied to merit pay. Toch said that is the only sensible way to use tests to measure employees. Other critics remain unconvinced that it solves the problems and question whether number crunchers can really screen out all outside factors.
But the new evaluation at San Diego Unified uses test scores in a way that makes "value added" advocates wince. Many of the testing goals in the new evaluation are based on changes in scores across a whole school, not changes in individual children over time. Principals would be judged partly on whether they met testing goals under No Child Left Behind, which would leave them vulnerable to losing points if they suddenly gained a large number of children with disabilities or English learners. Even proponents of linking data to principals say taking that tack could be dicey, giving principals a perverse incentive to game the system.
"You would just find the kids that are really smart and bribe them to get to your school," said attorney Tyler Cramer, once the chairman of the now-dissolved Business Roundtable for Education. "We would look like the most rockin' school you've ever seen. That's the fallacy of this."
Grier said that it's best to measure both raw scores and "value added" data, such as the goals that principals have set for growth at their schools. If students are severely behind, he said, it might be great that a school can show improvement in their scores, but if they still end the year far behind where they should be, the school needs to do more. He added that while principals will be evaluated yearly, they will not be punished for temporary blips in their scores. Nor are tests the sole measure of a principal.
The idea still worries Katrina Peterson, a parent at Birney Elementary. Test scores have dropped at the University Heights school over the past two years, putting it in the lower echelons among schools with similar demographics across California. Peterson would be alarmed by the drop, but she has also noticed a growing share of Birney students with disabilities, which makes her doubt that the scores reflect a decline at the school.
"It's odd that we, as parents, would say, 'We want to keep this principal despite the fact that the scores dropped,'" Peterson said. She herself is a teacher in a suburban school with stellar scores. "Yet you walk onto that campus — the kids are happy. The teachers are working hard and keeping kids engaged in ways that haven't been happening at Birney before. We value that."
Administrators Association Director Bruce McGirr also worried that principals may be held to task for factors they can't change. Principals do not have complete control over which teachers work in their schools, for instance.
Deputy Superintendent Chuck Morris argued that good principals are able to improve mediocre teachers and sway genuinely bad ones — who Morris believes are rare — to find work elsewhere. That's part of their job, he said.
"It's not about teachers not wanting to perform," Morris said. "It's not knowing how to perform."
His hope is that the new evaluation will give principals a better sense of how they're supposed to perform too, and a clearer way to guide them when their performance falls short.
Please contact Emily Alpert directly at firstname.lastname@example.org with your thoughts, ideas, personal stories or tips. Or set the tone of the debate with a letter to the editor.