Using Discriminant Analysis to create a Dropout Prediction Thermometer for the Applied Social Sciences Field.

In Brazil, there is a vast discussion in the academic environment that seeks to understand/explain the elements that influence the dropout of higher education students. Thus, this article presents an "evasion thermometer" to indicate how close to evading the student may be or not. The discriminant analysis technique was applied, using academic data in the institution's database to suggest which elements contribute to the possible dropout of students. As a result of the study, a function was identified that discriminates which elements contribute to student dropout so that by knowing these elements in advance, the institution can take appropriate measures to promote student permanence.