Factor Analysis of the Classroom Assessment Scoring System Replicates the Three Domain Structure and Reveals no Support for the Bifactor Model in German Preschools

The quality of early childhood education and care (ECEC) is important for children's development. One instrument that was developed to assess an aspect of ECEC quality is the Classroom Assessment Scoring System for pre-kindergarten children (CLASS Pre-K). We examined the factorial validity of the instrument using data from 177 German preschool classrooms. The three-factor <i>teaching through interaction</i> model (Hamre et al., 2013) was contrasted to a one-factor, a two-factor, and a bifactor model as proposed by Hamre et al. (2014). Our results indicated that the three-factor structure with the domains of <i>emotional support</i>, <i>classroom organization</i>, and <i>instructional support</i> fit the data best. The fit of the <i>teaching through interaction</i> model was satisfying after adding a cross-loading of the dimension language modeling on <i>emotional support</i>, and two correlated residuals. Evidence of convergent and discriminant validity are provided. In terms of factor structure and pattern score comparisons, the results were similar to previous United States and German studies. The discussion concerns the justifiability of the factor model revisions and draws directions for further research. We concluded that our study offers further evidence of the applicability of the CLASS Pre-K for the assessment of teacher-child interaction quality in the German context.