FITTING AN EXTENDED INDSCAL MODEL TO 3-WAY PROXIMITY DATA

The INDSCAL individual differences scaling model is extended by assumi ng dimensions specific to each stimulus or other object, as well as di mensions common to all stimuli or objects. An ''alternating maximum li kelihood'' procedure is used to seek maximum likelihood estimates of a ll parameters of this EXSCAL (Extended INDSCAL) model, including param eters of monotone splines assumed in a ''quasi-nonmetric'' approach. T he rationale for and numerical details of this approach are described and discussed, and the resulting EXSCAL method is illustrated on some data on perception of musical timbres.