The repertory grid method (RGM) was initially devised to determine how individuals construe their worlds. There have been many attempts to extend this principle to look at relationships amongst objects, as seen by groups of individuals. However, to derive the implied consensus representation of the objects, poses conceptual and mathematical problems because of the idiosyncratic nature of constructs elicited using RGM. Generalized Procrustes analysis (GPA) offers a solution since there is no assumption of common interpretation of constructs. It is an implicit but apparently untested assumption of GPA, that mathematical and perceptual structure are analogous. To test this, combined RGM/GPA was applied to color. Using the method of triads, 24 subjects independently described constructs to characterize 26 Munsell color chips differing in Hue, Value and Chroma. Subjects then rated the colors on their own construct scales. The first six dimensions of the consequent color space were interpreted as brightness, greenness, lightness, purple/blueness, redness and yellowness. This differs from expectation only in that the colors were recovered on uni‐polar dimensions rather than red‐green and blue‐yellow continua. However, since the main perceptual attributes of color were recovered the analogy assumption of GPA is supported, as is application to repertory grids.
No takes yet. Share an insight, caveat, or question.
Gains et al. (1990) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: