Comments Indexing Product Quality: Issues, Theory, and Results

(1984) strongly condemns the use of overall product-quality ran kings, such as those published by Consumer Reports, as a basis for scientific research. His primary criticism is that creating an index of quality as a weighted average of subscores is usually misleading. He claims that this averaging process is unjustified except for brands that occupy approximately the same rank position on each attribute. This article reviews Hjorth-Andersen's arguments and identifies several important related questions on the use of published quality ratings in academic re-search. Statistical theory is developed to demonstrate that Hjorth-Andersen's conclusions are limited, and to examine the precise conditions under which his conclusions hold. Empirical evidence is presented that reveals the statistical frequency in real markets of the logical conditions deduced using our theory. Our con-clusions differ significantly from those drawn by Hjorth-Andersen. A debate about the efficacy of quality indices is rel-evant not only to academic research but also to con-sumer welfare and strategic marketing planning. Con-sumer welfare and choice is affected by published qual-ity ratings (Archibald, Haulman, and Moody 1983; Beales et al. 1981; Thorelli and Thorelli 1977). Com-mercial importance is evident from the widespread dis-tribution of magazines like Consumer Reports. The consumer testing movement, started in 1927 by Chase and Schlink, has spread to agencies in more than 50 nations. These agencies are members of the Interna