The comments by Cohen, Fishbein, and Ahtola about the Bass and Talarzyk article [1] and the Sheth and Talarzyk article [4] are useful in clarifying distinctions between the concepts and measures implied by the Fishbein and Rosenberg models and the concepts and measures used in the two studies. These distinctions are real, as convincingly argued by Cohen, Fishbein, and Ahtola. However, the Fishbein and Rosenberg models were not developed for the purpose of studying brand preference or indeed for the purpose of studying relative attitudes for different objects, but rather for studying the attitudes of different people for the same object. This difference in purpose has an important bearing on the method of analysis to be used in comparing alternative models and