Detecting differential item functioning using generalized logistic regression in the context of large-scale assessments

Large-scale Assessments in Education

An IERI – International Educational Research Institute Journal

Table 2 Item difficulty and discrimination parameters used for generation of non-DIF items

Item	$\bar{b}$	b _min	b _max	$\bar{a}$	a _min	a _max
Item 1	.27	-.23	1.06	1.00	.57	1.33
Item 2	-.40	−1.58	.35	.97	.66	1.28
Item 3	.31	-.49	.85	.92	.50	1.17
Item 4	-.14	-.94	.45	1.04	.74	1.56
Item 5	.24	-.44	1.04	1.00	.73	1.19
Item 6	.08	-.63	.92	1.06	.80	1.73
Item 7	.12	−1.21	.93	.97	.70	1.63
Item 8	.06	−1.37	.97	.98	.78	1.23
Item 9	.03	-.55	.95	1.08	.55	1.76
Item 10	-.28	−1.23	.73	1.05	.92	1.18
Item 11	-.23	−2.00	1.09	.97	.65	1.23
Item 12	.09	−1.21	.68	.83	.46	1.23
Item 13	.22	-.46	.79	.82	.46	1.39
Item 14	-.04	-.62	.64	1.03	.76	1.55
Item 15	-.35	−1.38	.76	.92	.51	1.31

Note. These are the averages of item parameters across the seven blocks within a country that contains only non-DIF items.