Introduction.

In Spain, in order to practise any medical speciality, it is necessary to have qualified as a specialist. Access to training as a specialist is gained through the MIR test. This test is mandatory to qualify for a training position in any medical specialty. This article carries out a psychometric study of the 2020 and 2021 MIR tests (with 175 and 200 questions, respectively) through both the classical test theory and the item response theory.

Material and methods.

The database used in this study contains the responses from a total of 4,810 individuals who took the exam in 2020 and 3,888 who did so in 2021. The information available was analysed using indicators such as Cronbach's alpha, the formula number 21 of Kuder and Richardson, the difficulty index, the difficulty index corrected for random effects, the discrimination index and the point biserial correlation index. The difficulty and discrimination indices were also used, according to item response theory.

Results.

The MIR tests of 2020 and 2021 present high reliability, with only slight variations from one year to another in the behaviour of the different subjects and question blocks in terms of the indices used in the present study.

Conclusions.

The results obtained in the present study allow us to affirm that the MIR exam is objective, of appropriate difficulty and discrimination, as well as stable from one year to another.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )