Re-evaluating a vision-related quality of life questionnaire with item response theory (IRT) and differential item functioning (DIF) analyses

Table 2 Items with DIF between subgroups of relevant variables

	Item content	Subgroups	α	β ₁	β ₂	β ₃	β ₄	β ₅	G²	df	p	ES_Δmax	θ
Basic aspects
7.	Eyes getting tired	Low vision	2.50	-0.93	-0.65	-0.10	0.32	2.13	15.7	5	0.008	-0.80	-0.80
		Moderate vision loss	2.50	-1.37	-1.23	-0.54	0.10	1.86
Mobility
3.	Seeing steps or curbs*	Low vision	2.38	-0.97	-0.61	-0.01	0.21	1.56	14.2	1	< 0.001	-0.74	0.20
		Moderate vision loss	5.76	-0.96	-0.71	-0.26	-0.02	1.68
Adjustment
1.	Vision in general	Self	1.99	-1.98	-1.44	-0.14	0.38	2.33	19.4	5	0.002	-0.70	-2.00
		Assisted	1.99	-2.47	-2.30	-0.56	-0.09	2.33
1.	Vision in general	Male	2.08	-2.05	-1.44	0.04	0.40	1.95	17.9	5	0.003	0.62	2.60
		Female	2.08	-1.83	-1.57	-0.27	0.24	3.37
12.	Unhappy situation in life^†	Male	3.27	-0.74	-0.31	0.56	1.03	3.06	15.4	5	0.009	1.00	5.40
		Female	3.27	-0.39	-0.27	0.51	0.88	7.57
Reading and fine work
19.	Reading labels	AMD	3.77	-1.34	-1.21	-0.94	-0.46	0.68	17.8	5	0.003	0.24	-0.80
		Other	3.77	-1.48	-1.16	-0.63	-0.51	0.68
24.	Using tools	Male	1.83	-1.22	-0.92	-0.38	-0.15	1.13	35.2	5	< 0.001	-1.22	-1.00
		Female	1.83	-1.80	-1.58	-1.39	-0.94	1.05

Person distribution standard normal; α = item discrimination parameter; β = item threshold parameter; G² = Chi-square statistic; df = degrees of freedom; ES_Δmax = maximum difference in expected scores, θ = person parameter (disability level is presented at which the maximum difference in expected scores was found); AMD = age-related macular degeneration.
* Non-uniform DIF; † After deleting item 1, DIF for item 12 was not statistically significant.

ISSN: 1471-2288