Incorporating Conceptual Clarity and Topic Coverage into Flesch-Kincaid Readability Model for University Course Content Analysis

Abstract

While the Flesch-Kincaid (FK) formula remains a standard for readability assessment, current research highlights its limitations in measuring actual text comprehension due to its focus on surface-level linguistic features. This study addresses this gap by developing the Conceptual Clarity and Topic Coverage (CCTC) enhanced Flesch Reading Ease (FRE) model, which integrates semantic coherence and syllabus alignment into the traditional readability framework. The system, implemented via a Flask-based web application, automates the evaluation of university course materials using a pipeline of Latent Dirichlet Allocation (LDA) for topic distribution and TF-IDF cosine similarity for conceptual clarity. Validation was performed using a dataset of Computer Science courseware benchmarked against expert human ratings. Statistical analysis via ANOVA and Pearson correlation demonstrated that the CCTC-enhanced FRE score achieves a significantly higher correlation with human judgment (τ=0.86, ρ<0.05) compared to traditional metrics such as the standard FRE (τ=0.86), Gunning Fog (-0.56), and SMOG (-0.77). These findings provide empirical evidence that incorporating semantic and thematic depth significantly improves the accuracy of automated readability assessments in technical academic contexts. This research offers a robust framework for educational quality assurance and data-driven curriculum design.