d="sp0015">Mixed methodology was used, including qualitative and quantitative phases. The psychometric properties of the MEDUC-RX32 survey were performed by factor analysis (validity), Cronbach alpha coefficient, and G coefficient (reliability). The residents assessed their teachers and simultaneously voted for the “best teacher,” which was used as a gold standard for the receiver operating characteristic curves construction comparing their votes with the global score.
d="sp0020">A total of 28 residents answered 164 surveys. The global score was 6.23 ± 0.8 (scale from 1 to 7). The factor analysis showed six domains of the resident's perception: (1) tutorial teaching, feedback, and independent learning; (2) communication and teamwork; (3) learning objectives; (4) respectful behavior; (5) radiological report; and (6) teaching and care support. The tutor's strengths were related with respectful behavior and teamwork. The instrument is highly reliable with a Cronbach alpha of 0.937 and a G coefficient of 0.831 (with a minimum of 8 residents). The MEDUC-RX32 instrument has a sensitivity of 91.7% and specificity of 83.3% to identify tutors as best teachers with at least one vote with an area under the receiver operating characteristic curve of 0.931 with a cutoff of 5.94.
d="sp0025">The MEDC-RX32 instrument is a multidimensional, valid, and highly reliable method to evaluate radiology teachers, identifying teachers with excellence in tutorial teaching in a postgraduate radiology program.