Abstract

Introduction: Multiple-choice assessments are the instrument widely used in science to assess students, how-ever, the recent pandemic required adapting this type of instrument to the virtual environment. This context required evaluating the quality of the instruments through discrimination indices, internal consistency and relating it to academic performance. Objective: Evaluate the evaluation instruments used in online mode during the COVID-19 pandemic and the performance of students in health sciences. Method: The revision of the 5 instruments of the Structure and Function subject formed by 290 banks of random questions was carried out to evaluate each content in first-year students during 2020 in the school of health sciences at the Viña del Mar University. The data obtained from the virtual platform and the indices of discrimination, facility, discriminative efficiency, internal consistency and academic performance were interpreted through a report that was shared with the teachers to identify the parame-ters of quality and validity. Results: Of the total number of question banks evaluated, 70.2% of the questions presented adequate discrimination and only 5.6% should be eliminated. Contest two obtained the lowest average performance 3.9 ± 0.99, however, it presented the highest internal consistency 81%. When comparing all the instruments, a gradual im-provement in the formulation was observed, reflected in the final exam, in which the academic performance also agrees with the average of the semester 4.2 ± 0.92. Conclusions: Academic performance must be weighed in relation to the quality of the formulated instrument, in which, at a lower ease index, there is greater internal consistency, represented by the greater discriminative efficiency of the questions. The design and formulation process must take care of and examine these guidelines to safeguard quality criteria.