The MIR 2018 Exam: Psychometric Study and Comparison with the Previous Nine Years

Abstract

Background and Objectives: The aim of the present research is to study the questions used in the 2018 MIR exam (a test that allows access to specialized medical training in Spain), describe their psychometric properties, and evaluate their quality. Materials and Methods: This analysis is performed with the help of classical test theory (CTT) and item response theory (IRT). The answers given to the test questions by a total of 3868 physicians are analyzed. Results: According to CTT, the average difficulty index for all of the test questions was 0.629, which falls into the acceptable category. The average difficulty index with correction for random effects was 0.515, which corresponds to a value within the optimal range. The mean discrimination index was 0.277, which is in the good category, while the mean point biserial correlation coefficient, with a value of 0.275 fits in the regular category. The values of difficulty and discrimination calculated according to the model of two parameters of the IRT seem adequate with average values of −0.389 and 0.677. The Cronbach alpha score obtained for the overall test was 0.944. This value is considered as very good. Conclusions: A decrease was observed in the average values of discrimination in the last three calls, which may be related to the greater proportion of Spanish graduates that take the exam in the same year of finalization of their studies in Medicine.

Resultados obtenidos en la prueba MIR según baremo académico. Convocatorias de 2019 y 2020

“…Antes de la realización del presente trabajo, los autores han realizado otras investigaciones relacionadas con la prueba MIR, en la que la han estudiado desde el punto de vista psicométrico, tanto de la teoría clásica de los test 7 como de la teoría de respuesta al ítem 13 , el uso de ambas metodologías ha permitido conocer la dificultad de las preguntas propuestas en función de la asignatura a la que pertenecen 14 , así como la evolución de dicha dificultad a lo largo de las sucesivas convocatorias de la prueba 15,16 . Incluso, a partir de la experiencia acumulada en el análisis psicométrico de la prueba, se han podido formular sugerencias acerca de la manera en la que se podrían detectar preguntas susceptibles de anulación 17 .…”

Answering Multiple-Choice Questions in Which Examinees Doubt What the True Answer Is among Different Options

“…Generalizing what has been stated in the previous cases, for a total of n questions in which there is doubt among three answer options over which is correct, the total number of possible different answers to the set of said questions is equal to 3 n . Of these answers, if you want to know which ones result in a negative score, it is possible to use Formula (7). When making use of Formula (7), it must be taken into account that there are two incorrect answer options, called i 1 and i 2 .…”

“…In Spain, all Medicine graduates who want to apply for a position as a specialist doctor must pass the MIR exam [7]. Since 1978, this test has been convened annually, by the Ministries of Health and Education, and it is carried out on the same day and time in various venues, located in different cities in Spain.…”