On the direct scaling approach of eliciting aggregated fuzzy information: the psychophysical view

We describe an experiment in which we elicit aggregated fuzzy information, and link it to the "direct scaling" approach of S.S. Stevens (1957). We also re-analyze earlier experiments [H.-J. Zimmermann and P. Zysno, 1980] and compare the results to the proposed scaling method. Our results seem to indicate that a compensatory operator is not always necessary to model aggregated fuzzy information, and that humans internally use a simple rescaling.