Systematic observation to measure teaching quality in different contexts: insights from science lessons, physical education lessons, and sports training using the classroom assessment scoring system

There is substantial evidence that the quality of classroom interactions is crucial for students’ academic and social development. Both subject-specific and generic systematic observation instruments are widely used to assess these interactions. A notable example of a generic observation protocol based on US classroom research is the Classroom Assessment Scoring System (CLASS). This study aims to evaluate the applicability of the CLASS framework for assessing teaching quality in various contexts in Germany, specifically in science classes, physical education classes, and sports training at the elementary school level. In total, 110 video-recorded observation cycles, each lasting 15–20 min, were double-coded by two independent observers. Assessments were conducted across 10 dimensions using a 7-point scale. Results indicate average scores for the dimensions within the Emotional Support domain in the medium to high range, within the Classroom Organization domain in the high range, and within the Instructional Support domain in the low to medium range. CLASS scores varied considerably across different settings, effectively distinguishing between the observed teachers and coaches. The Percentage-Within-One agreement values and Intra-Class-Correlations demonstrate good interrater reliability across all settings. These findings highlight the robustness of the CLASS framework and its transferability to various educational contexts in Germany. This adaptability facilitates future studies on predictive validity and enables cross-country comparisons.