Abstract
This is a comparative research of 15 equating methods, 4 based on Classical Testing Theory (CTT) and 11 based on Item Response Theory (IRT), using HSK real data. A relative reliable criterion was established. This research concluded: (1) the equating was not always the best choice; (2) some IRT based equating methods could be used in item bank development; (3) neither one, two or three parameter models, neither ms or mm methods, IRT parameter transfer methods could be used in HSK equating for these methods resulted in significant errors.
Keywords: | test equating; psychometrics; educational measurement; item response theory |
---|
[Chinese Version | Index | Acta Psychologica Sinica | Other Journals | Subscription form | Enquiry ]