Triple Sense-Making of Findings from Marketing Experiments Using the Dominant Variable Based-Logic, Case-Based Logic, and Isomorphic Modeling

Abstract

The study describes the complementary benefits of model-building and data analysis using algorithm and statistical modeling methods in the context of unobtrusive marketing field experiments and in transforming findings into isomorphic management models. Relevant for marketing performance measurement, case-based configural analysis is a relatively new paradigm in crafting and testing theory. Statistical testing of hypotheses to learn net effects of individual terms in multiple regression analysis is the current dominant logic. Isomorphic modeling might best communicate what executives should decide using the findings from algorithm and statistical models. We test these propositions using data from an unobtrusive field experiment in a retailing context that includes two levels of expertise, four price-points, and presence versus absence of a friend ("pal" condition) during the customer-salesperson interactions (n=240 store customers). The analyses support the conclusion that all three approaches to modeling provide useful complementary information substantially above the use of one alone and that transforming findings from such models into isomorphic management models is possible.