MULTISIB - A PROCEDURE TO INVESTIGATE DIF WHEN A TEST IS INTENTIONALLY 2-DIMENSIONAL

MULTISIB is proposed as a statistical test for assessing differential item functioning (DIF) of intentionally two-dimensional test data, suc h as a mathematics test designed to measure algebra and geometry. MULT ISIB is based on the multidimensional model of DIF as presented in She aly & Stout (1993), and is a direct extension of SIBTEST, its unidimen sional counterpart. For an intentionally two-dimensional test, DIF is appropriately modeled to result from secondary dimensional influence f rom other than the two intended dimensions. Simulation studies were us ed to assess the performance of MULTISIB to detect DIF in intentionall y two-dimensional tests. These results indicate that MULTISIB exhibite d reasonably good adherence to the nominal level of significance and g ood power. Moreover, for each DIF model the average amount of DIF esti mated over the 100 simulations of the model by MULTISIB was close to t he true value, confirming its relative lack of statistical estimation bias in assessing true DIF. In addition, the simulation studies suppor ted the importance of using the regression correction to adjust the sc ores on the studied item due to impact and the importance of matching examinees on two subtest scores instead of the total test score.