Abstract
The formal complexity underlying a group of objects may be partitioned into two components, one of which is due to the total number of attributes involved (the between-attribute component) and the other of which is due to complexity within the individual attributes themselves (the within-attribute component). This notion of multidimensional structure is discussed within the framework of an additive component model of multidimensional scaling, where a configuration is considered to be composed of disjoint subspaces, each one of which reflects variation due to a specific stimulus component. A number of empirical examples are given to demonstrate the applicability of the additive component model to multidimensional scaling.
Get full access to this article
View all access options for this article.