Inclusion of MCQs written by radiology residents in their annual evaluation: innovative method to enhance resident's empowerment?

Aim We hypothesized that multiple-choice questions written by radiology residents (MCQ(resident)) for their weekly case presentations during radiology staff meetings could be used along with multiple-choice questions written by radiology teachers (MCQ(teacher)) for their annual evaluation. The current prospective study aimed at determining the educational characteristics of MCQ(resident) and at comparing them with those of MCQ(teacher). Methods Fifty-one radiology residents of the first to the fifth year of training took the 2017 exam that contained 58 MCQ(resident) and 63 of MCQ(teacher). The difficulty index, the discrimination power, and the distractor's quality were calculated in the two series of MCQs and were compared by using Student t test. Two radiologists classified each MCQ according to Bloom's taxonomy and frequencies of required skills of both MCQ series were compared. Results The mean +/- SD difficulty index of MCQ(resident) was statistically significantly higher than that of MCQ(teacher) (0.81 +/- 0.1 vs 0.64 +/- 0.2; p < 0.0001). The mean +/- SD discrimination index of MCQ(resident) was statistically significantly higher than that of MCQ(teacher) (0.34 +/- 0.2 vs 0.23 +/- 0.2; p = 0.0007). The mean number of non-functional distractors per MCQ(resident) was statistically significantly higher than that per MCQ(teacher) (1.36 +/- 0.9 vs 0.86 +/- 0.9; p = 0.0031). MCQ(resident) required recalling skills more frequently than MCQ(teacher) which required more advanced skills to obtain a correct answer. Conclusions Educational characteristics of MCQ(resident) differ from those of MCQ(teacher). This study highlights the characteristics to optimize the writing of MCQs by radiology residents.