The Pre-K Classroom Assessment Scoring System (CLASS) is commonly used to measure three domains of teacher–child interactional quality in preschool classrooms (Emotional Support, Classroom Organization, and Instructional Support). However, there is considerable debate regarding the validity of its three-domain factor structure and the applicability of CLASS across different countries and educational settings. This study contributes to this debate with an investigation of the factor structure of the CLASS in Singapore, comparing the factor structure for preschool classrooms serving typically developing children (n = 184) and early intervention classrooms serving children who require early intervention support (n = 182). Bifactor measurement models were used to explore the dimensionality of the CLASS in these settings. A best-fitting bifactor model with one general (Responsive Teaching) and two specific factors (Proactive Management and Routines, Cognitive Facilitation) was a good fit for the overall data. Comparison of preschool and early intervention classrooms indicated partial metric invariance, suggesting that while the factor structure was equivalent across the different settings, the loadings on some dimensions differed between preschool and early intervention classrooms. Future research could examine how different aspects of interactions in different settings contribute to children’s learning.

在学前班教室评分制（CLASS）通常用来衡量在幼儿园教室师幼互动质量（情感支持，课堂组织，指导和培训）的三个领域。然而，关于其三域因素结构的有效性以及 CLASS 在不同国家和教育环境中的适用性，存在相当多的争论。这项研究有助于在新加坡之类的因素结构的调查这场辩论，比较为服务正常发育儿童，学龄前教室的因素结构（ñ  = 184），并服务于谁需要早期干预支持儿童早期干预教室（ñ = 182）。双因子测量模型用于探索这些设置中 CLASS 的维度。具有一个通用因素（响应式教学）和两个特定因素（主动管理和常规、认知促进）的最佳拟合双因素模型非常适合整体数据。学前班和早期干预班级的比较表明部分度量不变，这表明虽然因子结构在不同环境中是等效的，但学前班和早期干预班级在某些维度上的负荷不同。未来的研究可以研究不同环境中互动的不同方面如何促进儿童的学习。