My thesis explores Item Response Theory (IRT) as a tool for analysing and scoring tests by measuring latent abilities. It provides a theoretical overview of key IRT models, including one-parameter (1PL), two-parameter (2PL), three-parameter (3PL), and various polytomous and multidimensional models. Using the RStudio environment, I applied these models to actual university exam data with the ‘mirt’ (Multidimensional Item Response Theory) package, analysing item parameters, standard errors, and factor scores indicative of examinees' true ability, and the amount of information provided by displaying the appropriate diagrams. In addition, the graphs show the relationship of the latent variable with the students' score and the expected scores in detail for each question. Therefore, a multifaceted visual and conceptual understanding of the theory and its application is offered through practical analysis on real data.
Παρουσίαση Πτυχιακής Εργασίας-Μαρία Ιωάννα Κατσιγιάννη(1059194).pptx