Application of Confidence Intervals in the Multiple Choice Evaluation System

Autores UPV
Año
CONGRESO Application of Confidence Intervals in the Multiple Choice Evaluation System

Abstract

Abstract Exams should not be considered only as a tool for the evaluation of the knowledge acquired by the students during the teaching-learning process. Exams should be used as a feedback tool, so the lecturer can detect which subjects have not developed properly, or not enough, or should be explained in another way. In open-response exams it is easy to detect the deficiencies produced in the learning process, since students develop the knowledge acquired. The analysis of the writing allows the teacher to detect whether the correct answer was answered with full confidence or not. In contrast, test exams students mark only the answer which they think it is correct, so it is hard to know whether the response was made safely, with doubts or has been successful at random. With the implementation of graduate degrees it is required a continuous assessment; this implies several screening tests performed during the development of the subject. This causes an increased workload on the lecturer who often uses objective evidence or "multiple choice" tests because they are easier and faster to correct. This test are performed more quickly and therefore can be performed in the time available in a class session. The main drawback of the test exam is that the student does not have to support the response, and therefore it is impossible to identify whether the student has responded with conviction, by elimination of the other answers or randomly. This drawback means that there is no feedback to the lecturer and he cannot identify whether there have been problems in the learning process. The aim of this work is the development of a methodology which detects the percentage of confidence that the students have in the answer given in order to provide feedback to lecturers.