What explains Macau students’ achievement? An integrative perspective using a machine learning approach (¿Cuál es la explicación del rendimiento de los estudiantes macaenses? Una perspectiva integradora mediante la adopción del enfoque del aprendizaje automático)

Abstract

Although Macau students have consistently been recognized as top performers in international assessments, little research has been conducted to explore the various factors that are associated with their achievement. This paper aimed to identify factors that could best predict Macau students' reading achievement using PISA 2018 data provided by 2,979 15-year-old students. An integrative theoretical model that considered the critical roles of demographic, personal and social-contextual factors was used to understand the relative importance of 41 different factors in predicting reading achievement. A machine learning approach, specifically Random Forest Algorithm, was used to analyse the data. Results indicated that variables classified under personal factors (e.g., metacognitive strategies, reading enjoyment and perceived difficulty) were the most important predictors of Macau students' achievement. A supplementary analysis using Hierarchical Linear Modelling confirmed the findings from the machine learning approach. Implications of the findings were discussed. RESUMENAunque los informes internacionales sitúan asiduamente a los estudiantes macaenses entre los primeros puestos de las tablas clasificatorias, apenas se han realizado estudios exploratorios sobre los diversos factores asociados con su rendimiento. En este artículo se identifican los factores que podrían predecir el rendimiento lector de los estudiantes macaenses utilizando datos del informe PISA 2018 sobre 2979 estudiantes de 15 años de este país. En el estudio se utiliza un modelo teórico integrado en el que se tienen en cuenta los roles clave de factores demográficos, personales y socio contextuales para descifrar la importancia relativa de 41 factores distintos en la predicción de la competencia lectora. Para analizar los datos se aplicó un enfoque basado en el aprendizaje automático, concretamente el algoritmo Random Forest (conjuntos de árboles o ARTICLE HISTORY

Predicting Factors that Affect East Asian Students’ Reading Proficiency in PISA

“…Using data responses from PISA in the year 2018, the researchers [17] compare a series of machine learning algorithms to determine the relevant algorithm that can accurately predict the reading achievement of Macau students. The machine learning algorithms that are selected for analysis include multiple tree-based ensemble machine learning such as Random Forest (RF), Gradient and Extreme Gradient Boosting, Extra Tree, and TreeBag.…”

Exploring school factors of academic achievement in Macau: The application of educational data mining in Progress in International Reading Literacy Study (PIRLS) 2016 assessment

“…They observed the strongest correlations between reading self-efficacy and home learning materials, as well as the school atmosphere. Similarly, Wang et al [6] conducted research on Macau students' reading proficiency using PISA 2018 data and an integrative theoretical model. They discovered that personal factors, such as metacognitive strategies, enjoyment of reading, and perceived difficulty, were the most significant predictors of high reading performance among Macau students.…”