Big Data in Education: Detection of ICT Factors Associated with School Effectiveness with Data Mining Techniques

The search for non-contextual process factors associated with school effectiveness has become a line of research with very broad dissemination, primarily from the ubiquity of research that includes analysis of secondary data from large-scale evaluations. With the aim of detecting ICT factors related to school effectiveness from the Spanish sample of the PISA 2015 evaluation, this work applies statistical techniques of data mining, specifically decision trees, for the optimization of the process. The results demonstrate that, while the availability and excessive use of ICTs, both in the educational environment and in the home, are factors associated with low effectiveness, other more personal characteristics of the pupils, such as their perception of self-efficacy in the management of ICT, or their own interest in the use of technologies, can be considered as factors in the protection of school effectiveness. We conclude by discussing the relationship between school effectiveness and academic performance through the analysis of previous studies, focusing on the common elements detected within both perspectives.