TEMPORAL AGGREGATION IN-DIFFUSION MODELS OF FIRST-TIME PURCHASE - DOES CHOICE OF FREQUENCY MATTER

Consistent with work in the advertising response literature, the autho r addresses the time-interval bias present when estimating innovation models of new product growth and diffusion with discrete time-series d ata. Specifically, the author explores the theoretical and empirical i mplications of using varying data frequencies when estimating diffusio n models using both nonlinear least squares (NLLS) and ordinary least squares (OLS). Parameter estimates across five consumer durables are o btained using annual, quarterly, and monthly data. The central conclus ion is that the information gained and bias minimized by using seasona lly adjusted quarterly data results in empirical estimates that are an improvement over those obtained by using annual data. This is true fo r both the NLLS and OLS estimates. In contrast, the move from quarterl y to monthly data produces only marginal statistical improvement.