Analysis of questionnaires for Virtual Learning Environments based on Item Response Theory

Resumo

This paper presents a model for the design and creation of virtual adaptive evaluations for e-learning environments, combining Item Response Theory (IRT) along with log analysis of previous questionnaires. The proposed model allows the definition of a methodology for the ranking and categorization of questions. Such ranking provides valuable feedback to the teacher or tutor who can refine and adapt the questionnaire. Experiment results reveal that IRT parameters are sufficient for ranking and selecting questions that are more appropriate to teach specific topics. We believe that this approach should become an essential tool for the creation of questionnaires that are more concise and effective in the context of virtual courses.