“…These included iterative purification procedures under various conditions: (a) different matching criteria, (b) different analytic units (i.e., item-, bundle-, or testlet-level), and (c) different assumptions about tests' dimensional structures (e.g., Gierl, Bisanz, Bisanz, Boughton, & Khaliq, 2001;Mazor, Kanjee, & Clauser, 1995;Oshima, Raju & Flowers, 1997;Wainer, 1995). Furthermore, variations of the aforementioned procedures were developed to study non-uniform DIF (Li & Stout, 1996). Additionally, the literature includes numerous comparison studies that used older DIF procedures, such as STAND, SIBTEST, MH, item response theory (IRT) based, and logistic regression, (Chang, Mazzeo, & Roussos, 1996;de Ayala, Kim, Stapleton, & Dayton, 2002;Zwick, Thayer, & Mazzeo, 1997).…”