Classification models in the digital competence of higher education teachers based on the DigCompEdu Framework: logistic regression and segment tree

Abstract

To promote and develop the digital competence of higher education teachers is a key aim in the 21st century. Teachers must have a leader or expert digital competence in order to prepare future school-leavers for a competent professional qualification. Therefore, the purpose of this study is to determine the predictor variables encouraging high digital competence, using two statistical classification techniques: multiple logistic regression and classification trees. The analysis of teachers’ digital competence was carried out in each of the areas of knowledge in which the teachers are assigned, as well as overall. For data collection, a non-experimental ex post facto design was used. A total of 1,104 higher education teachers from Andalusia (Spain) completed the DigCompEdu Check-In instrument prepared by the European Commission’s Joint Research Centre. In terms of general classification, the results found that the logistic regression technique ranked teachers’ digital competence with greater probability of success (83.7%) in comparison to the segment tree (81.7%). The results found that the level of digital competence of teachers in the creation and use of digital resources varies according to the area of knowledge to which the teachers are assigned. At a general level, the development of digital competence at the leader, expert or pioneer level is related to variousfactors, such asthe time spent on creating web spaces and digital content, and the use of virtual reality, robotics, and gamification. Further research is recommended to validate these preliminary findings in each of the areas of knowledge