Análise e avaliação da trajetória de estudantes de graduação baseadas em modelagem por cadeias de Markov

Abstract

Dropout is present in higher education institutions worldwide. Disorders during the undergraduate process can cause high levels of dropout, which makes the phenomenon a problem for institutions and for society, especially when financed with public resources. The objective of this work is to present an investigation with the possibility of intervention on the academic flow of undergraduate students, through a stochastic model based in absorbing Markov chain. Based on the proposition, simulations of the evolutionary dynamics were carried out to estimate the impacts of some interventions on the analyzed phenomenon, especially on the dropout. The existing models in the literature are based on natural hierarchies of an undergraduate course, such as hours and years successfully completed. This imposes a fixed path of states to be necessarily fulfilled before the graduation, which restricts the possibilities of intervention on the environment. Thus, an approach was proposed to analyze and evaluate the trajectory of undergraduate students, which included an absorbing, homogeneous, first-order Markov chain, with discrete time, in which the set of states is defined by the academic situation and the number of students cumulative failures. In a case study, using data from the Federal Rural University of Pernambuco, the investigation process identified the estimated impact of some changes in the system. Through preliminary changes in the model’s transition probabilities, it was realized that actions on freshmen, before the first failure, would have a greater influence on the number of graduates. When the process of modeling and selecting scenarios was implemented, incremental and achievable trajectories were identified with the power to reduce up to 57% of the proportion between graduated and dropout. And, when the experiments were carried out considering the area of knowledge of each undergraduate course, the average reduction was 73.77% in the proportion dropout over graduates.