ERIC Number: ED045732
Record Type: RIE
Publication Date: 1970-Jun
Reference Count: 0
Independence of Dimensions in Multidimensional Scaling.
Models for multidimensional scaling use metric spaces with additive difference metrics. Two important properties of additive difference metrics are decomposability and intradimensional subtractivity. A prediction was derived from these properties and tested experimentally. Eleven non-psychology students were used as subjects. Rectangles varying in area and shape served as stimuli. Two types of dissimilarity judgments were obtained: rating scale judgment and pair comparison. In the rating scale experiment two stimuli at a time were shown and the students were asked to judge the overall difference on a scale running from 0 for "no difference" to 8 for "very great difference." In the pair comparisons, slides showed two pairs and the subjects had to judge which of the two pairs showed the larger overall difference. The decomposability and intradimensional subtractivity assumptions of the model were violated by most students due to an apparent interaction between area and shape. The experiment failed to confirm the model and it is suggested that the model be rejected or modified. (Author/LR)
Publication Type: N/A
Education Level: N/A
Authoring Institution: Technische Hochschule Darmstadt (West Germany).