Predictive models of pedagogical competence in Vocational Secondary School teachers through educational data mining

Abstract

Research researches the use of predictive methods through Educational Data Mining, using decision trees to determine the development needs in pedagogical skills of Technical-Professional High Schools (TPHS) and contribute to research that strengthens teacher training. The research was undertaken from concurrent nested design of several three-stage levels. This article considers the quantitative stage, following the phases required by data mining: Selection, exploration, cleansing, transformation, data mining, evaluation and interpretation of results, dissemination and use of models. The results of the work are classified into two parts. In the first, latent variables were characterized through load values with confirmatory analysis. The second part focuses on finding a prediction model that establishes a strong relationship between predicted pedagogical competition and latent variables defined by social analysis, professional action, and situation diagnosis. Findings are illustrated using social data.