After sets of mutually equally discriminable random shapes, all generated from a single prototype, had been identified, the members of the several sets were immediately recognized as differing in associativeness or meaningfulness. The meaningfulness (m) of each shape was determined through an application of the production method. The computed values of m did not, in many cases, coincide with the meaningfulness of the shapes as judged by several trained observers. Satisfactory indices of the heterogeneity (and, conversely, the homogeneity) of the verbal responses to each of the shapes seemed impossible to obtain. Thereupon, the degree of appropriateness of each verbal response (word or short phrase) for describing its corresponding shape was determined through an interval scaling procedure. The mean of 22 scale values-descriptive appropriateness values-for each shape was taken to be the connotative strength (cs) of the shape. The Pearson r for the m and cs values was an insignificant .09. The tentative, but fairly firm, conclusion was that values of cs were more clearly indicative, than were values of m, of what the shapes signified when seen by groups of untrained observers.
Proceedings of the Iowa Academy of Science
©1965 Iowa Academy of Science, Inc.
Lewis, Don and Boehnert, Joanna B.
"Assessing the Connotative Strengths of Random Shapes,"
Proceedings of the Iowa Academy of Science, 72(1), 378-389.
Available at: https://scholarworks.uni.edu/pias/vol72/iss1/55