Multidimensional scaling of ranking data for different age groups

Collecting data for multidimensional scaling (MDS) by pair comparisons is common practice in research. The,ranking task as an alternative promises different practical advantages. However, its validity is uncertain. In this study, the validity of nonmetric MDS of ranking data is examined. Twelve kindergarten children, high-scholars and students judged the similarity of seven colors using rankings. The validity of the MDS-results was judged by comparing them with arrangement of colors derived from the Munsell system, which had been validated in former studies. The MDS of the ranking data for high-scholars and students, but not for kindergarten children proved to be valid. Possible causes for the later will be discussed.