The objective of teacher evaluation systems is not only to increase accountability for teachers' performance but also to improve teachers' instruction via feedback. Unfortunately, evaluation policies have focused more on accountability than on feedback, as was evident when many districts suspended evaluation practices during COVID, appearing to forget that a strong evaluation system could help teachers navigate the switch to online and hybrid instruction.

A recent paper by Matthew Kraft from Brown University and Alvin Christian from University of Michigan puts the spotlight back on the quality of feedback teachers receive in their evaluations. The authors studied Boston Public School's major evaluation reform of 2011, which emphasized using evaluation as a tool for professional growth and development, and was accompanied by evaluator training specifically designed to improve the quality of feedback teachers received.

Surveys of Boston teachers' perceptions on the feedback they received found little to suggest they saw an improvement after the training program. The new evaluator training made almost no dent in the notably low percentage of teachers (around 25%) who reported both before and after its implementation that they receive feedback that would actually cause them to change their practice. The only improvement they reported was that the feedback given to them was more timely.

Still, useful information from digging further into these surveys could help Boston and other districts improve evaluation practices. For example, teachers perceived that the feedback was more useful and actionable when the evaluator was more experienced in general and had longer tenure at their school in particular. Interestingly, feedback was perceived to be less useful for teachers who taught subject specific classes as opposed to being subject generalists (though the data did not include the evaluator's area of expertise, nor whether it matched the teacher's endorsement areas). This suggests that districts would benefit from investing in a corps of evaluators with instructional experience who are also familiar with the particular environment of the teachers they evaluate, and could arguably support other research findings that stress the need for content matching between evaluators and teachers.

Districts may consider other cost-effective strategies, such as the one studied by Papay et. al., which showed that teachers benefited tremendously when low-performing teachers were paired up with high-performing ones in a particular skill area. Using video-based observations could alleviate some of the evaluators' time constraint problems.

Race must also be taken into consideration. Teachers perceived higher quality feedback where racial congruence existed between teacher and evaluator, and lower quality feedback when the evaluator was a person of color. Although racial matching can alleviate the bias in evaluations and could do the same for bias in the perception of feedback quality, the finding mostly speaks to the long road ahead in addressing racial bias.