Sensitivity of Classroom Observation Systems

Abstract

AbstractClassroom-based measures of teaching effectiveness may be collected at a variety of levels including a macro (teacher-to-group) or a micro (teacher-to-student) level. This study examined correspondence between these two levels of measurement using (a) multivariate analysis of variance (MANOVA) of extreme groups and (b) whole-sample cluster analysis procedures. Mixed results were found. The MANOVA indicated limited correspondence between macro and micro measures of teaching effectiveness. Cluster analysis, however, failed to show any significant correspondence between macro and micro measures of teaching effectiveness. Instead, results suggested that macro measures tended to separate more effective teachers from all others, whereas micro measures were more useful for separating less effective teachers from all others. These outcomes are discussed in terms of their implications for the purposes of teacher observation, as well as the use to which measurement data are used.