Abstract
A solution is provided for the problem of large numbers of variables in V-analysis and of large numbers of subjects in 0-analysis. The solution is that of estimating the Structure in the total set from the structures found in manageable samples of the total set. The samples are drawn randomly; structures of each are computed by existing BC TRY programs; finally the structures of the random samples are combined using new BC TRY programs. Examples of the complete analyses in two separate studies are given.

This publication has 3 references indexed in Scilit: