Last update:
09-Mar-2009

OBJECTIVE Multiple choice examinations are often used to assess student achievement. Question's content is chosen mainly by the tutor according to his/her judgment and experience. This project aimed to develop a program for evaluation of multiple choice questions used for testing medical students.

METHOD Software was developed which enabled the creation of a question base into which all multiple choice questions used to examine medical students were transferred. For each question the difficulty and discrimination indices were calculated. The discrimination index represents the discrete faculty of each question that is whether it has a high rate of success within the students with high grades. Evaluation was made of 95 multiple choice questions used in a population of 176±76 (72-441) students; a special scanner and special software were used.

RESULTS Questions evaluated were correctly answered in a percentage (success rate) of 53±22%. Of all the questions, 15/95 were considered to be of recommended difficulty (16%) while 13/95 (14%) were considered to be of "high difficulty-not acceptable" and 21/95 (22%) of "high facility-not acceptable". Of most excellent discrimination were 38/95 questions (40%) while 16/95 (17%) were considered to be of low discrimination.

CONCLUSIONS Choosing multiple choice questions according to the tutor's judgment only, is not sufficient to create an objective evaluation system of students' knowledge. The use of specifically designed software is an easy and safe way of improving the quality of multiple choice questions.

Key words: Evaluation, Medical education, Multiple choice questions.