IRT-based evaluation of learning material in e-learning systems
Αξιολόγηση πολυμεσικού μαθησιακού υλικού με εφαρμογή της θεωρίας απόκρισης ερωτήματος (IRT) σε συστήματα ηλεκτρονικής μάθησης
The scope of this thesis is to provide a methodological and architecturalframework which embeds in a Learning Management System (LMS) a user-friendlyanalysis tool based in Item Response Theory (IRT) in order to maintain a high level ofquality in the self-assessment tests and help test developers decide whether a test itemcan be reused as is, should be revised before reuse or should be taken out of the activeitem pool. Additionally, it introduces a comprehensible way to present IRT analysisresults without delving into unnecessary details. Instead of memorizing numerouscommands and scenarios from technical manuals, test developers can easily detectproblematic questions from the familiar user interface of a LMS. Finally, this thesisuses introduces four quality indices as a means to evaluate the accuracy of the IRTanalysis and provide test developers with a success rate of detecting potentiallyflawed items, thus proving that when used appropriately by a LMS, IRT can improvethe assessment quality, increase the efficiency of the testing process, and provide indepthdescriptions of item and test properties.