Could not download file: This paper is available to authorised users only.


Biopharmacy and Pharmacokinetics (BP) is a subject provided in the third year of the Pharmacy studies, according to the syllabus of the University of Valencia. Our teaching team has been responsible for one of the groups of the third course for several years. The theory exam includes short and multiple-choice (test) questions. Usually, students complain about the types of questions of the exam, considering that short or test questions are easier compared to each other.
Therefore, the objective of this work is to establish a correlation between the types of questions included in the final exam from two different academic years in BP in order to understand and identify whether level and contents of each type of questions are adequately balanced.

Individual marks of students from 2014-2015 and 2016-2017 academic years in BP were collected. Students with no qualification in one of the different exam categories were discarded in the analysis. 432 samples were included in the final dataset. Pearson coefficient was estimated for the correlation and non-linear regression including standard errors were added to the plot.

A correlation was developed including marks from test and short questions of each individual student from two academic years. Results show the high correlation between both variables (R=0.48), which indicates that the level of both types of questions are adequately balanced. In general, students with higher test marks receive higher marks in short questions. However, there is a significant number of students who obtained a high mark in one of the questions but failed in the other type of questions.

These results confirm that there is a correspondence in the contents asked by each type of questions and the understanding by the students. Thus, the type of questions has in general low relevance on the qualification that students achieve. However, as a future perspective, it would be helpful to understand and clarify why in some students this correlation is not established in order to detect any bias.