Teacher evaluation through observation: Application of classroom observation and student ratings to improve teaching effectiveness in classrooms

Once is not enough

Teacher evaluation is central to current policy efforts aiming to raise educational standards. In the many countries including the Netherlands England and the U.S., teacher evaluation involves classroom observation by the school principal which ideally visits his/her teachers once or twice yearly. This dissertation research suggests that visiting teachers once or twice will lead to unreliable and invalid evaluation. Once (or twice) visiting a teacher’s lesson provides observers with a “snap shot” and too few information to draw any firm conclusions concerning the teacher’s general pedagogical-didactical skill. Modest reliability requires at least four lesson visits, preferably by different observers. High reliability seems to be unachievable if using only classroom observation and most likely will require some combination of student questionnaire ratings and classroom observations. Hence, decisions which have high personal stakes (such as teacher tenure, salary and dismissal) and which require high reliability preferably should not be made based on observations made during lesson visits only.
This dissertation discusses this and other issues concerning the reliability and validity of classroom observation and student questionnaires.

For more information about the dissertation “Teacher evaluation though observation” please email Rikkert van der Lans: [email protected].