The item response theory aims to estimate the skills of its respondents, based on tests composed of questions of different levels of difficulty. There are several models of item response theory, which have been proposed for various types of problems, such as binary or probabilistic responses and even applications in the machine learning area. In order to evaluate certain models, it is necessary to obtain sets of answers, which is usually very expensive. This project aimed to create a response database to train item response theory models, the creation of a web page to facilitate access to data and software, which allows for easy manipulation of downloaded data.
keywords: Item response theory, Psychometrics, Machine Learning.
Leave a comment