ANALYSIS OF INDONESIAN STUDENTS' READING LITERACY USING THE SMOOTHLY CLIPPED ABSOLUTE DEVIATION (SCAD) PENALTY

Abstract

Reading literacy significantly impacts a country's educational level, making it crucial to further investigate into this issue. Identifying factors that influence students' reading literacy, particularly in Indonesia, is a key area of exploration. PISA survey data, conducted every three years, is relevant for researching student proficiency. Each survey period focuses on one of three main topics: science literacy, mathematics literacy, and reading literacy. The 2018 PISA survey data is suitable for studying students’ reading literacy, as the main topic that year was reading literacy. However, PISA survey data includes many strongly correlated independent variables, potentially violating the multicollinearity assumption. To address this, regression analysis with a penalty function is used for variable selection. The SCAD (Smoothly Clipped Absolute Deviation) penalty function has proven effective in previous studies on PISA data. The model using the SCAD penalty function yielded excellent results, indicated by an Adjusted R² value of 0.967. Based on this model, three main factors influence students' reading literacy in Indonesia: learning facilities, general knowledge, and students' self-confidence.