Multidimensional Scaling (ViSta-MDScal) is a data analysis and visualization technique for analyzing distance-like data (proximities, dissimilarities) and displaying a low-dimensional Euclidean space that summarizes the distance information in the data.

The data consist of distance-like information between all pairs of a set of objects. There may be several matrices of such data, and the data matrices may be asymmetric or symmetric. The data must be dissimilarities, not similarities (i.e., large numbers mean large distance). There may not be any missing data.

ViSta-MDScal locates points in a Eucliden space which has a point for every object in the data. The points are located so that those that are close together have data which are similar, and those which are far apart have data which are different.  The user must decide on the appropriate dimensionality of the Euclidean space.

The ViSta-MDScal visualization is based on the initial estimates of the best locations for the points. The fit of the points to the data can be improved by using the iterate button. The visualization provides several views of the multidimensional Euclidean space, as well as a fit plot to help decide on dimensionality.  
