An empirical analysis of one, two, and three parametric logistic models of item response theory and differential item functioning in dichotomous and ordinal tests

Abstract

The purpose of this study was to determine the three parametric logistic IRT methods in dichotomous and ordinal test items due to differential item functioning using statistical DIF detection methods of SIBTEST, GMH, and  LDFA. The study adopted instrumentation research design. The sample consisted of an intact class of 457 Part 3 undergraduate students who registered for EDU 304 (Tests and Measurement). Two research instruments were used to collect data. Data collected were analysed using three statistical DIF detection methods. The result  showed that there was a significant difference in the three parameters logistic models in dichotomous and ordinal test items due to differential item functioning. Hence, there existed a high degree of correspondence among the three parameter models in identifying psychometric properties of dichotomous and ordinal test items. Thus, the three parameters logistic models should be used in test development processes to enhance validity of tests.

Keywords:Item response theory (IRT), item parameters, dichotomous test, ordinal test, and differential item functioning (DIF)