International Journal of Arts and Humanities (IJAH)

This quantitative study examined the relationship between specific teacher-child interaction behaviors and language modeling quality in Pre-K classrooms using the Classroom Assessment Scoring System (CLASS) framework. Data were collected from structured observations in 11 Pre- K classrooms in Southeast Virginia during spring and summer 2024. The Language Modeling dimension was assessed through five key indicators: frequent conversations, open-ended questions, repetition and extension, self and parallel talk, and use of advanced language. Descriptive statistics, bivariate correlation, and regression analyses were employed to examine relationships between observed teacher behaviors and overall language modeling scores. Results revealed weak and statistically non-significant correlations across all variables, with no individual teacher behavior serving as a strong predictor of language modeling quality. These findings suggest that effective language modeling is a complex, multidimensional process that cannot be captured through analysis of isolated behaviors (Pianta et al., 2008; Hamre et al., 2013).