Market Change

To test this result further, the reliability and validity of the scale and its dimensions were additionally evaluated with confirmatory factor analysis applied to the three latent constructs. AMOS 4.0 was selected for the analysis. Parameter estimation was done with the ML-method (Kline, 1998). The values of the overall fit measures (GFI=0.972; AGFI=0.935) indicate a very good fit of the analysed model to the empirical data. In a second step, the fit of the local structures of the model was also examined. Indicator reliability, factor reliability and explained variance of the three latent variables all imply a good model fit. The expected interdependence between the latent constructs (which was also a result of the exploratory factor analysis) appears clearly in the path coefficients. However, even the highest coefficient (0.62 between quality of performance and scope of offers/convenience) still indicates discriminant validity. Therefore, the three-dimensional structure of food retail store perception is well confirmed by the data.

Comparing the results of the consumer survey to Porter's basic types of competitive advantages, his suggested bipolar structure of potential competitive advantages is (again) not supported by the empirical data:

•

The factor “quality of performance” is much more comprehensive than the quality advantages as presented and explained by Porter, who focused on product quality advantages. Quality of performance also includes store atmosphere, service and orderliness.

•

More importantly, the factor “scope of offers/convenience” extracts another basic dimension of competitive differentiation advantage for retailers. To distinguish oneself from the competition, a wide range of goods and services might serve just as well as the merchandise quality.

•

Also, with price being a separate factor, it becomes clear that quality advantages and cost advantages are not the opposite ends of one dimension as Porter's assumption would suggest. We find quality and cost leadership to be independent factors, whose mutual achievement is not contradictory or characterised by a trade off.

All in all, this factor structure can be seen as evidence, from a consumer perspective, that the bipolar pattern of competitive advantage as postulated by Porter (1985) is probably too simplistic and cost and differentiation advantages are not mutually exclusive in retailing. The three-dimensional structure we have found for food retailing clearly contradicts Porter's (1985) assumption of two generic competitive strategies, if the central dimensions of perception from a consumer perspective are accepted as reflecting potential competitive advantages.