With all the column space devoted these days to fixing the evaluation of teachers, it's easy to lose sight of how important (perhaps even more important) it is for us to fix principal evaluation as well. Two reports out this month provide some useful guidance.

Educators 4 Excellence (E4E)--a new outfit started by two New York City teachers to put the teacher voice in teacher policy--wants to make sure that teachers' perspectives get considered when evaluating principals. Citing research that half of all teachers in New York City leave within five years and principals are the biggest factor in teachers' decisions to leave schools, E4E proposes a number of components that should comprise a robust evaluation of a school principal, including:

1. Multiple observations by trained supervisors;
2. School climate surveys;
3. Retention rates of effective teachers under the principal's leadership;
4. Student attendance;
5. School safety information; and,
6. Student growth data.

Also out this month, a Rand Corporation write-up on the little-studied subject of principal effectiveness. Using a sample of some 500 first-year principals working in six urban districts, Rand found that one out of every five new principals were removed within two years of starting their first assignment. While it might be assumed that the district was yanking underperforming principals (scores tended to be lower for those principals), such quick-trigger trades rarely turned out to be advisable. The schools which lost a principal after only one year tended to have even lower performance in the subsequent year.

So what does Rand find makes new principals more effective? Though constrained by a small sample, the researchers found no significant relationship between how principals spend their time (be it management, external relations, instructional leadership, or non-academic interactions)and their effectiveness. Instead, it was the strength of teacher capacity, collaboration, and cohesiveness, at least as estimated by the principal, that made the difference. In schools in which the principals identified these factors as important, principals were more likely to last more than a year and advance student learning.

Superintendents take note: these two reports argue for the careful inclusion of climate surveys in principal evaluations, emphasize cohesion and collaboration along with very careful placement of principals from the start, and patience to give new leaders time to effect improvements and produce results.