In this note it is indicated on which assumptions Luce’s similarity choice model for confusion data is based and how they can be tested. Application of the tests on ten published confusion matrices reveals that Luce’s model is not universally applicable due to systematic violation of the quasi-symmetry assumption.
How to Cite:
De Soete, G., 1984. On the Validity of Luce’s Similarity Choice Model for Confusion Data. Psychologica Belgica, 24(1), pp.83–86. DOI: http://doi.org/10.5334/pb.717