Items do not always function equally in different groups (e.g. across genders, languages and cultures). Consequently, where patient-reported outcomes are used and different groups are compared, data should be checked for differential item functioning (DIF). An item that functions with the same constant magnitude of difference across a construct measured possesses uniform DIF. In contrast, non-uniform DIF is characterised by an uneven difference in item function across the latent variable measured. The aims of this paper are to report on the methodological aspects of DIF using Rasch analysis and to demonstrate how the mean scores in a scale can be adjusted due to uniform DIF. Different examples of DIF are reported including examples of differences between the mean scores before and after adjusting for an identified uniform DIF. In conclusion, the difference between subpopulations and, therefore, other outcomes such as economic impact could be under or overestimated if one or more items in a dimension possess DIF.
Read More: http://informahealthcare.com/doi/abs/10.3111/13696990701557048Items do not always function equallyin different groups (e.g. across genders,languages and cultures). Consequently,where patient-reported outcomes are usedand different groups are compared, datashould be checked for differential itemfunctioning (DIF). An item that functionswith the same constant magnitude ofdifference across a construct measuredpossesses uniform DIF. In contrast,non-uniform DIF is characterised by anuneven difference in item function acrossthe latent variable measured. The aimsof this paper are to report on themethodological aspects of DIF using Raschanalysis and to demonstrate how the meanscores in a scale can be adjusted due touniform DIF. Different examples of DIF arereported including examples of differencesbetween the mean scores before and afteradjusting for an identifi ed uniform DIF.In conclusion, the difference betweensubpopulations and, therefore, otheroutcomes such as economic impact couldbe under or overestimated if one or moreitems in a dimension possess DIFItems do not always function equally in different groups (e.g. across genders, languages and cultures). Consequently, where patient-reported outcomes are used and different groups are compared, data should be checked for differential item functioning (DIF). An item that functions with the same constant magnitude of difference across a construct measured possesses uniform DIF. In contrast, non-uniform DIF is characterised by an uneven difference in item function across the latent variable measured. The aims of this paper are to report on the methodological aspects of DIF using Rasch analysis and to demonstrate how the mean scores in a scale can be adjusted due to uniform DIF. Different examples of DIF are reported including examples of differences between the mean scores before and after adjusting for an identified uniform DIF. In conclusion, the difference between subpopulations and, therefore, other outcomes such as economic impact could be under or overestimated if one or more items in a dimension possess DIF.
Read More: http://informahealthcare.com/doi/full/10.3111/13696990701557048Items do not always function equally in different groups (e.g. across genders, languages and cultures). Consequently, where patient-reported outcomes are used and different groups are compared, data should be checked for differential item functioning (DIF). An item that functions with the same constant magnitude of difference across a construct measured possesses uniform DIF. In contrast, non-uniform DIF is characterised by an uneven difference in item function across the latent variable measured. The aims of this paper are to report on the methodological aspects of DIF using Rasch analysis and to demonstrate how the mean scores in a scale can be adjusted due to uniform DIF. Different examples of DIF are reported including examples of differences between the mean scores before and after adjusting for an identified uniform DIF. In conclusion, the difference between subpopulations and, therefore, other outcomes such as economic impact could be under or overestimated if one or more items in a dimension possess DIF.
Read More: http://informahealthcare.com/doi/abs/10.3111/13696990701557048Items do not always function equally in different groups (e.g. across genders, languages and cultures). Consequently, where patient-reported outcomes are used and different groups are compared, data should be checked for differential item functioning (DIF). An item that functions with the same constant magnitude of difference across a construct measured possesses uniform DIF. In contrast, non-uniform DIF is characterised by an uneven difference in item function across the latent variable measured. The aims of this paper are to report on the methodological aspects of DIF using Rasch analysis and to demonstrate how the mean scores in a scale can be adjusted due to uniform DIF. Different examples of DIF are reported including examples of differences between the mean scores before and after adjusting for an identified uniform DIF. In conclusion, the difference between subpopulations and, therefore, other outcomes such as economic impact could be under or overestimated if one or more items in a dimension possess DIF.
Read More: http://informahealthcare.com/doi/abs/10.3111/13696990701557048
