Generic versus specific scales and the ‘fidelity versus bandwidth’ issue
Let us say you want to find or develop a test to measure the quality of life (QOL) of a group of rheumatoid arthritis patients . Should you look for (or develop) a QOL instrument which is tailored to the characteristics of these RA patients, or should it be a scale which taps QOL for patients with a variety of disorders , of which RA is only one? Indeed ,should we go even further, and tailor the scale to the unique requirements of the individual patients? There are some scales (e.g.)
Guyatt ci at. 1993) which are constructed by asking the patient to list five activities which he or she feels have been most affected by the disorder. thus, Thus this item on one patient’s scale may be quite different from those on anyone else’s instrument.