A SMART PLS-SEM MODEL OF COMMUNICATIVE COMPETENCE BASED ON SPEAKING, LISTENING, READING, AND WRITING SKILLS

Abstract

This study develops and tests a structural equation model of communicative competence grounded in the four core language skills: speaking, listening, reading, and writing. Drawing on expert-based criteria for language proficiency, each skill is conceptualized as a reflective latent construct measured through performance-oriented indicators. A quantitative, cross-sectional design was employed, and data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The model specifies a hierarchical structure in which the four skills contribute to a higher-order construct of communicative competence. Results indicate that all measurement indicators achieve adequate reliability and validity, with strong outer loadings, satisfactory composite reliability, and acceptable average variance extracted values. Structural analysis reveals that all four skills have significant and positive effects on communicative competence, with writing and speaking showing the strongest contributions. The model demonstrates substantial explanatory power and predictive relevance, confirming the multidimensional nature of language proficiency. These findings support the integration of productive and receptive skills in both assessment and instruction, providing a robust framework for evaluating language competence in academic and applied contexts.