Interpreting Smallest Space Analysis

Abstract

Two kinds of SSA (smallest space analysis) are distinguished: R-SSA, which uses as input a matrix of R-distance coefficients (distances among pairs of variables); and Q-SSA, which uses as input a matrix of Q-distance coefficients (distances among pairs of objects). While both types are common in the literature, only researchers using Q-SSA have attempted to interpret the coordinates of the space. It is shown that the coordinates can be interpreted in R-SSA, and that this interpretation yields valuable information that is otherwise lost. Also, SSA and factor analysis are compared from a facet analysis perspective. Finally, an R-SSA is compared with an R-factor analysis of the same data.