It has been determined that students learn by doing and through
experience. Faced with the appearance of tools such as artificial
intelligence, teachers have the great responsibility of turning the
classroom into a space for the construction of analysis that promote
students’ critical thinking skills and involvement in real contexts. The
present proposed is a quasi-experimental study. Two large groups of
undergraduate students are considered who through a stratified
random probability are considered as a control group and another
group as treatment one. The treatment group will be those who perform roles and the academic debate; while the control group will be those who carry out a traditional academic debate. The estimation of the results, carry a certain qualitative part through surveys and observation sheets, while the quantitative part is performed by estimating a difference-in-differences model. It is intended to demonstrate that students who play roles within an academic debate have a broader domain of the subject of study, while presenting much broader arguments by linking their theoretical knowledge with a reality. On the other hand, those who still exercise a classic academic debate are the ones who work satisfactorily, but their arguments are limited and poorly supported. This work will contribute to teachers, academic directors and policy makers involved in the educational field in developing countries to reformulate study times, dedication of subject contents to experiential learning.