Enhancing General Reading Understanding: Measuring Foundational Reading Skills Relevance

Basic reading skills are crucial for literacy development, and digital tools, like machine learning techniques, offer promising ways to support the acquisition of those skills. This study evaluated the impact of assessing specific foundational reading skills on overall reading acquisition using trial-based direct instruction. We developed and tested three versions of the General Pre-Reading Competence Index (GR-CI) by analyzing data from 1,184 children who completed 228 pre-reading tasks in Portuguese. We found that the GR-CI k++ Filtered performed better, based on the silhouette scores from K-Means++ cluster analysis conversion into weights for each foundational reading skill. Implementing the GR-CI k++ Filtered was highly effective, accurately categorizing 93.8% of participants with distinct reading scores and enhancing the discriminative power of conventional reading assessments (status quo) by 52%. This result significantly increased the identification of variations in reading acquisition among students. These findings establish a new pathway for equitable and individualized assessment of reading skills. They also contribute to assisting educators and policymakers in developing more equitable learning strategies. We discuss the proposal of additional trial-based direct instruction parameters to enhance individualized reading assessment.