Technology and abstraction: Complex skills development through video games [Tecnologia e abstração: Desenvolvimento de habilidades complexas por meio de videogames]

Abstract

This study aims to propose supervised machine learning models to predict the abstraction ability in students asan early warning mechanism through both the technology use and the video game use. The methodology usedwas mixed with a prescriptive and predictive design; 118 tests were carried out by Chilean pedagogy students.For analysis, several variables were correlated: age, modality, academic semester and six predictive models.The results show three relevant findings; First, regarding the relation abstraction/ age, in the Satisfactory andNo abstraction, the distribution is homogeneous in every age. Second, the relationship abstraction/modalityshows a 50% pattern for all the abstraction categories. Third, the relation abstraction/academic semestershows that most students from the seventh semester have no capacity for abstraction. The study concludesthe abstraction level is low, showing that 61,1% of the students do not have a higher cognitive level. Twoout of six of the supervised machine learning models are suggested to predict early warning, decision tree andrandom forest since they have a 100% accuracy. Therefore, through the use of technology and the video gameit is possible to ensure higher cognitive level development, by the use of several planned strategies in covid-19times concentrated in the first two tears of pedagogy training.