69 SS DESCRIBED APPROXIMATELY 10 DIFFERENT PERSONS THEY KNEW BY SELECTING PERSONALITY TRAIT NAMES FROM A LIST SUPPLIED BY E. A MEASURE OF TRAIT COOCCURRENCE FOR EACH PAIR WAS DERIVED AND USED IN KRUSKAL'S MULTIDIMENSIONAL SCALING PROGRAM. THE FIT FOR THE 2- AND 3-DIMENSIONAL CONFIGURATIONS FELL IN THE "FAIR-TO-GOOD" RANGE WITH ADDITIONAL DIMENSIONALITY ADDING LITTLE TO THE GOODNESS OF FIT. THE 2 CONFIGURATIONS WERE INTERPRETED BY FINDING AXES IN THE CONFIGURATIONS WHICH CORRESPONDED TO INDEPENDENTLY MEASURED PSYCHOLOGICAL PROPERTIES OF THE TRAITS AND LOCATED BY MULTIPLE-REGRESSION TECHNIQUES. 2 CONNOTATIVE PROPERTIES, GOOD-BAD AND HARD-SOFT, PROVIDED A SATISFACTORY INTERPRETATION OF THE 2-DIMENSIONAL CONFIGURATION; 2 DENOTATIVE PROPERTIES, SOCIAL DESIRABILITY AND INTELLECTUAL DESIRABILITY, PROVIDED AN ALTERNATIVE INTERPRETATION OF THIS CONFIGURATION. THE SAME CONNOTATIVE AND DENOTATIVE PROPERTIES WERE FOUND IN THE 3-DIMENSIONAL CONFIGURATION; THERE WAS ALSO EVIDENCE OF A 3RD CONNOTATIVE PROPERTY, ACTIVE-PASSIVE. IMPLICATIONS FOR IMPRESSION FORMATION ARE DISCUSSED. (25 REF.) (PsycINFO Database Record (c) 2016 APA, all rights reserved)