Jb. Bjorner et al., DIFFERENTIAL ITEM FUNCTIONING IN THE DANISH TRANSLATION OF THE SF-36, Journal of clinical epidemiology, 51(11), 1998, pp. 1189-1202
Statistical analyses of Differential Item Functioning (DIF) can be use
d for rigorous translation evaluations. DIF techniques test whether ea
ch item functions in the same way, irrespective of the' country, langu
age, or culture of the respondents. For a given level of health, the s
core on any item should be independent of nationality. This requiremen
t can be tested through contingency-table methods, which are efficient
for analyzing all types of items. We investigated DIF in the Danish t
ranslation of the SF-36 Health Survey, using two general population sa
mples (USA, n = 1,506; Denmark, n = 3,950). DIF was identified for 12
out of 35 items. These results agreed with independent ratings of tran
slation quality, but the statistical techniques were more sensitive. W
hen included in scales, the items exhibiting DIF had only a little imp
act on conclusions about cross-national differences in health in the g
eneral population. However, if used as single items, the DIF items cou
ld seriously bias results from cross-national comparisons. Also, the D
IF items might have larger impact on cross-national comparison of grou
ps with poorer health status. We conclude that analysis of DIF is usef
ul for evaluating questionnaire translations. (C) 1998 Elsevier Scienc
e Inc.