| Abstract: | To identify a few homogeneous subgroups with distinct personality profiles, the Interpersonal Style Inventory (ISI) was administered to two samples, each composed of 100 university undergraduates. Ward's hierarchical algorithm and the buildup nonhierarchical cluster analytic program was applied to measures of profile similarity among members of each sample. Of the six clusters that were isolated in the first sample, five were matched at significant levels across the two methods. In the second sample, six clusters were matched across clustering procedures at significant levels. Comparisons across samples disclosed five good matches between clusters generated by the Ward hierarchical algorithm, but only one for the nonhierarchical analytic procedures, These results indicate that most IS1 personality profiles can be classified into five or six categories for purposes of interpretation or prediction. |
