Abstract
Over the past several years, item response theory (IRT) has had a dramatic impact on the field of educational measurement. At the theoretical level and for educational researchers, it has shown itself to have certain advantages over classical theory both in terms of the research questions it can address and in terms of the applications it can support. For these benefits to be realized by the measurement practitioner, however, two major difficulties must be confronted: (a) the enormous complexity of IRT theory and procedures and (b) the lack of robustness of IRT procedures to violation of assumptions.