A comparison was made between the QWB score obtained for the 132 scenarios obtained from this population (rheumatoid arthritis sufferers) with the scores that would be relevant if Bush's weights were used, and the two data groups were reported to have a correlation of r=0.937. However results for scenarios which compared arthritis type scenarios only were not given. This leaves open the question as to the accuracy of estimates made by the general public vis a vis patients of specific conditions. It would not be surprising if, as was suggested here, arthritics did not differ much from the general population when considering non arthritis type condition which both are equally uninformed about.