The Washington Post’s The Answer Sheet blog discusses the shortcomings of evaluating teachers using student test scores. It cites a recent American Institute for Research report on the New York State teacher evaluation model that found when the percentage of a classroom's students with disabilities or in poverty increases, the average teacher evaluation score decreases.
Add to this a new survey of 500 New York principals regarding the “student growth score” evaluation model. Seventy-three percent of principals surveyed felt that the “ineffective” label assigned to some of their teachers was either a "not very accurate" or an "inaccurate" reflection of that teacher based on their observations and the performance of that teacher’s students.
In the survey, some principals stated that next year they would reassign certain teachers to less needy students so they could protect excellent teachers from an ineffective rating. Others expressed concerns that excellent teachers would choose to leave for schools or districts with less needy students to avoid the risk of an ineffective label. This suggests that new teacher evaluation systems may be discouraging teachers from taking the most challenging assignments and thus decreasing the chances that needy students will get the best possible teachers.