An analysis of the performance of the hypothesis testing content for the population average of health students who participated in the active teaching methodology called Problem Based Learning adapted to the teaching context and supported by the Conceptual Fields Theory is presented. This performance was compared with that of students who took traditional classes through an instrument created for this purpose that sought to replicate typical questions of this content. The instrument was applied to 99 students from a higher education institution in southern Brazil, by the Moodle environment, in person and using Excel and Bioestat software. Scores were analyzed by Exploratory Data Analysis, the Classical Test Theory and Hypothesis Tests and were higher for students who did not participate in the PBL, but the differences were not significant at α = 5%. Possible causes are the better adaptation of the instrument to the traditional education system, contextual limitations to the application of the active methodology such as low workload and resistance to new methodologies. It is concluded that the difference in scores is not large enough to discourage the effort to adapt active methodologies to biostatistics, given its advantages other than yield.