Academic Performance, Socioeconomic Level and Digital Divide: A Big Data Analysis of the Saber 11 Tests. (2008-2024)

Abstract

This article analyzes the relationship between socioeconomic conditions, access to Information and Communication Technologies (ICTs), and the academic performance of Colombian students on the Saber 11 standardized tests during the period 2008–2024. The study is based on the premise that the digital divide is a determining factor in educational inequality, especially after the COVID-19 pandemic, which accelerated the transition to learning environments mediated by digital technologies. To this end, a quantitative, longitudinal, and descriptive-correlational approach was employed, using a database of more than nine million records cleaned from the Colombian Institute for the Evaluation of Education (ICFES). The methodology included data extraction, transformation, and loading (ETL) processes, statistical analysis, Pearson correlations, and machine learning techniques, particularly Random Forest, MissForest, and the K-Means algorithm. The results show a significant increase in internet access, rising from 22.2% in 2008 to 73.4% in 2024. However, significant differences persist between students in public and private institutions, both in access to and intensity of use of technology. Cluster analysis identified three student profiles. The group with the best academic performance exhibited higher socioeconomic levels and almost universal access to the internet and computers. In contrast, the lowest-performing group showed marked digital exclusion and unfavorable socioeconomic conditions. Furthermore, correlations showed positive associations between technological access, socioeconomic status, and performance in mathematics, critical reading, and English. It is concluded that the digital divide transcends access to technology and is closely linked to socioeconomic factors that influence the educational use of ICTs and learning opportunities.