Multi-map comparison for group concept mapping: an approach for examining conceptual congruence through spatial correspondence

Group concept mapping, a participatory mixed-methods approach used extensively in behavioral and social research, is used to specify and generate a two-dimensional conceptual model based on input solicited from an identified group. In situations where the systematic evaluation of the multidimensional conceptualized patterns generated by different subgroups is meaningful, little guidance exists. This paper contrasts two analytical approaches, configural similarity comparison and Procrustes comparison, emphasizing the latter as a more rigorous and appropriate technique for facilitating such comparisons. As demonstrated in this study, Procrustes analysis provides a solid statistical and interpretative foundation to measuring the similarity of MDS configurations found in concept mapping output. Paired with a permutation strategy for assessing significance and examination of residual values, Procrustes analysis offers an objective means to evaluate the general concordance of multivariate patterns generated through group concept mapping. Statistical and visual techniques are also used to further explore the specific patterns of residual values generated in the Procrustes comparison. From this demonstration, a procedure for testing the correspondence between multiple two-dimensional concept maps where the same content is considered by independent groups is suggested.