An Analysis on the Corrective Formula for Multiple Choice Questions

Abstract

In multiple choice questions, an element inherent to this type of evaluation arises that, due to the same evaluation format, that is by chance, a percentage of the correct answers do not re ect the mastery of the question content being evaluated. For this reason, in many cases, a correction formula is applied that penalizes incorrect answers, leaving unanswered questions without penalty. Of course, it is necessary to mention this to the applicant prior to the start of the test. The penalty in the case of questions with k alternatives where only one is correct and all the others are incorrect, depends on the number of options and in most applications consists of subtracting a value of 1 k 1 foreach incorrect answer. This formula assumes that the student knows the correct answer, or that they ignore the question and answer randomly. This is di cult to achieve in practice, since the student usually has partial knowledge of the question theme, helping them to eliminate one or more alternatives. In this paper, an analysis is made of the value of the penalty in cases where partial knowledge on the part of the applicant is considered, which allows them to discard one or more distracters.