Generalising IRT to Discriminate Between Examinees

Research output: Contributions to collected editions/worksArticle in conference proceedingsResearchpeer-review

Authors

We present a generalisation of the IRT framework that allows to discriminate between examinees. Our model therefore introduces examinee parameters that can be optimised with Expectation Maximisation-like algorithms. We provide
empirical results on PISA data showing that our approach leads to a more appropriate grouping of PISA countries than by test scores and socio-economic indicators.
Original languageEnglish
Title of host publicationProceedings of the International Conference on Educational Data Mining
EditorsOlga Cristina Santos, Jesus Gonzalez Boticario, Cristobal Romeo, Mykola Pechenizkiy, Agathe Merceron, Piotr Mitros, Jose Maria Luna, Cristian Mihaescu, Pablo Moreno, Arnon Hershkovitz, Sebastian Ventura, Michel Desmarais
Number of pages2
Place of PublicationMadrid
PublisherNational University for Distance Education (UNED)
Publication date2015
Pages604-605
ISBN (print)978-84-606-9425-0
Publication statusPublished - 2015
Externally publishedYes
Event8th International Conference on Educational Data Mining - EDM 2015 - Madrid, Spain
Duration: 26.06.201529.06.2015
Conference number: 8
http://educationaldatamining.org/EDM2015/