should be the only measurement, but it should be part of the analysis. Start with a baseline test of where the students are at the beginning of the year and compare it to where they end the year. Did the students improve? Was the improvement appropriate given their starting point? How does the improvement compare to the teacher's peers?
Then there should be some subjective analysis by the administrator who is in charge of the teacher. What kind of students does the teacher have? Does he or she do well or poorly when dealing with the marginal or more difficult students? Does the teacher reach the entire classroom, or is he or she effective with only part of the students? Is the teacher doing the things that you, the administrator, want the teacher to be doing? How much experience does the teacher have, and subjectively is the teacher doing things appropriately given the teacher's level of experience?