Combining statistical and machine learning methods to explore German students’ attitudes towards ICT in PISA

In our age of big data and growing computational power, versatility in analysis is important. This study presents a flexible way to combine statistics machine learning for large-scale educational survey. The authors used statistical methods explore German students’ attitudes towards information communication technology (ICT) relation mathematical scientific literacy measured by the Programme International Student Assessment (PISA) 2015 2018. Implementations random forest (RF) algorithm were applied impute missing predict proficiency levels mathematics science. Hierarchical linear models (HLM) built relationships between ICT with focus on nested structure data. autonomy was an important variable RF models, associations this attitude scores HLM significant positive, while other negative (ICT social interaction) or non-significant competence interest). need further research discussed, benefits combining approaches are outlined.