A Bayesian resampling approach to estimate the difference in effect sizes in consumer social responses to CSR initiatives versus corporate abilities

Abstract

AbstractWe expand previous analyses and advance our understanding of the difference in the impact on consumer purchasing behaviour between pursuing corporate social responsibility (CSR) initiatives and improving corporate abilities. To this aim, a Bayesian bootstrapping simulation is applied to selected consumer samples for 123 homogeneous choice‐based conjoint studies. We develop a simulation‐based approach to estimate the empirical distributions of effect sizes under two Bayesian bootstrap resampling schemes. This approach permits us to evaluate the results for two predefined classifications: Foote‐Cone‐Belding (FCB) and gender. The results indicate that females exhibit higher concern for CSR initiatives. Furthermore, we found that managers can exploit the classification of their product in the FCB grid and obtain more efficient consumer responses by implementing strategies focused on certain attributes. This is the first application of the FCB grid to identify differences in consumer responses across different products with different levels of rational consideration and involvement.