1970 Volume 12 Issue 4 Pages 154-162
(compatible with EndNote, Reference Manager, ProCite, RefWorks)
(compatible with BibDesk, LaTeX)
In this experiment, 42 Ss were randomly assigned to two rating scale conditions: 21 to the single scale condition (S-S) and 21 to the complex scale condition (C-S). They rated their favorableness for persons described by sets of 2, 4, 6, or 8 personality adjectives. Results were as follows.(a) The addition of the moderately polarized stimuli to the extremely polarized ones produced a less extreme response than the set of the extremely polarized alone.(b) The fitness of observed responses to predicted responses was significantly high.(c) While weights for informations given by C-S were almost equal, those by S-S increased with increased informations.(d) The set size effect was found obviously in responses to highly desirable traits given by both S-S and C-S, but did not found in those to highly undesirable traits.