Avaliação do desempenho do estudante de medicina pela medida da proficiência: descrição de uma metodologia de análise dos resultados do teste de progresso por meio da teoria de resposta ao item

Abstract

Objective: To evaluate, through Item Response Theory, the results of the Progress Test according to the students' proficiency and performance levels. Methods: A crosssectional study was conducted with 1559 Brazilian medical students. The estimation of proficiency and calibration of item parameters was performed using BILOG MG 3.0 software. For the construction of the proficiency scale, items with discrimination a > 1 were selected, which when submitted to the clustering technique using the k-means methodology, formed six subsets with increasing levels of proficiency. The pedagogical interpretation of the scale was obtained by relating proficiency levels to the cognitive domains tested. The 95% confidence interval was calculated and Tukey's test was applied to analyze the difference between the proficiency means. Results: The frequency of students who took the test was 94.2%. Proficiency ranged from 470 to 580 points, with a mean of 500 points, and increased progressively in graduation, an indicator of the validity of the measures obtained. The students' adjustment curve to the test indicated a difficult test and discrimination was moderate to high in 61.5% of the items. Conclusions: The analysis of the results of the Progress Test performed by Item Response Theory proved to be a robust tool to assess the performance of medical students.