The development of a self-administered psychophysical scaling method: range effects

Publisher Website
Google Scholar
Abstract
101 subjects engaged in a self-administered cross-modal matching task to scale descriptors of pain intensity and unpleasantness. The correlation between the two responses was used to determine reliability (internal consistency). When r .gtoreq. 0.90 was set as a criterion about 3% of the subjects were unable to calibrate standard stimuli reliably, 6% were unable to calibrate intensity descriptors and 60% were unable to calibrate unpleasantness descriptors reliably. When the criterion reliability was set at r .gtoreq. 0.85 all subjects successfully scaled the intensity descriptors. Further relaxation of the criterion to r .gtoreq. 0.80 reduced the ''failure'' rate for unpleasantness descriptors to 24%. These data compared favourably with other published results from non-chronic and chronic pain patients. The provision of a fixed reference value produced range effects in the resulting scale but do not severely attenuate the range of the scale as has been previously suggested. It is concluded that a self-administered version of the task is feasible. Reasons for the poor performance with the unpleasantness descriptors are discussed.