Criteria for Formative and Summative Teacher Evaluation Using Peer Observations

• In Event: Mixed Methods Approaches to Assessment, Evaluation, and Accountability Effectiveness in Schools: Graduate Student Reports

Abstract

In teacher evaluation formative and summative purposes are routinely merged. In this study, we use limitedly trained peer-observers and specify different reliability and validity criteria; one for formative purposes and one for summative purposes. To explore whether these criteria are realistic we sampled 133 peer-observations of 45 teachers. Using Generalizability theory and Rasch model person fit statistics, the results show that three peer-observers are required to achieve sufficient reliability (Eρ2 ≥ .70) and sufficient validity (GNORMED < .30) for formative purposes. For summative purposes, results indicate that eight peer-observations are required for to achieve sufficient reliability Eρ2 ≥ .90, while sufficient validity (GNORMED < .10) will require minimally nine observations, but for 44.4% of the teachers evaluated more than nine.

Authors

• Rikkert van der Lans, University of Groningen

• Wim van de Grift, University of Groningen

• Klaas Van Veen, University of Groningen