Abstract
The effect of personal construct elicitation methods on construct content and structure was addressed in two studies. In Study 1, a between-subjects design was used to compare Kelly's (1955) original Triadic Difference method of elicitation with a Dyadic Opposite method. Study 2 used a fully-crossed, mixed factorial design to compare Kelly's (1955) Triadic Difference method with Epting, Suchman, and Nickeson's (1971) Triadic Opposite method. Results showed that "difference" methods of construct elicitation produced significantly higher levels of construct differentiation, lower numbers of positive emergent construct poles, and less socially undesirable implicit construct poles than "opposite" methods (see also Caputi & Reddy, in press). Implications of these findings are discussed in relation to a critical reappraisal of repertory grid methods.