A structural equations model

Introduction Over the last decade there has been a growing interest in the market orientation concept (Day, 1992; Grönroos, 1989; Webster, 1988, 1992) and its usefulness in increasing businesses’ economic performance (Jaworski and Kholi, 1993; Narver and Slater, 1990; Reukert 1992). In most studies market orientation is defined as either the adoption of a marketing concept (see Deng and Dart, 1994; Deshpandé et al., 1993; Hooley et al., 1990; Kholi and Jaworski, 1990) or as the adoption of a high-quality marketing practice (see Kohli et al., 1993). However, in most cases it is not clear how marketing is defined, or a definition of marketing that has not been empirically validated is used. This poses a serious difficulty to any attempt to compare the various studies on the market orientation concept. The first empirically validated market orientation measure was developed by Narver and Slater (1990). They define market orientation as organizational culture and climate composed of: customer orientation, competitor orientation, and interfunctional orientation. For each of these components, they generate a set of items and obtain component scores as an unweighted sum of the items corresponding to each component. Finally, Narver and Slater report the correlations among these three components. Despite the importance of their pioneering effort, Narver and Slater’s work has been criticized on several grounds. For instance, Webster (1994, p. 223) noted that “...although they had no specific measures of company or business unit culture, they used the concept of culture to interpret some of their results”. Methodologically, their work is suspect since the assignment of items to components was made on purely theoretical grounds. No attempt was made to test empirically the match of the individual items to each of the components. With respect to this, after inspecting the content of the items, Siguaw and Diamantopoulos (1994) point out that even the theoretically driven assignment of the items they generate to components of their model is questionable. To these criticisms, we would add that these authors neglect the importance of the distributors and the environment as stakeholders in their operationalization of market orientation (Rivera, 1995). Kohli et al. (1993) have proposed and empirically validated an alternative market orientation measure. These authors delineate the construct’s domain by conducting personal interviews with managers. Then, after generating a set of Received September 1996 Revised November 1996