From twelve randomly derived visual shapes, six were to be chosen, if possible, which were mutually equally discriminable. The chosen six were subsequently to be used as stimuli in a verbal-motor transfer task. The two members of all possible pairs of the twelve shapes, including "same" pairs, were presented in succession by means of an automatic slide projector, the exposure time for each member being .25 sec. "Same" or "different" judgments made by 12 subjects after the presentation of each pair, were 98% correct, making useful differentiation among the 12 shapes impossible. Difficulty of discrimination was increased by presenting the pairs tachistoscopically with an exposure time of .01 sec per member. Seventy-two subjects made same or different judgments, as before. The total number of errors in judgment, each shape compared with every other, was taken as an index of discriminability. Six shapes were then chosen which approximated the goal of mutually equal discriminability.
Proceedings of the Iowa Academy of Science
©1961 Iowa Academy of Science, Inc.
Carver, L. E. and Marshall, Marilyn E.
"The Relative Discriminability of Twelve Random Shapes,"
Proceedings of the Iowa Academy of Science, 68(1), 522-528.
Available at: https://scholarworks.uni.edu/pias/vol68/iss1/73