“…Bemmaor (1979) employed the binomial runs test in the study of French panel data for regular coffee, instant coffee, and margarine and found that the zero-order hypothesis could not be rejected in 84%, 71%, and 56% of the cases, respectively.3 2Wierenga (1974) reports brand switching on successive purchase occasions of 8.9%, 11.4%, and 26.3% for the food product, beer, and margarine respectively (p. 12). These switching levels are considerably lower than the brand switching activity of about 40% found with the US consumer panel data (see, for example, Herniter 1974 andMassy 1966).…”