Files
Abstract
A distributed lagged advertising model with coefficients having random and systematic adjustments was estimated. The paths of parameter adjustments clearly identify structural changes in advertising effectiveness and price responsiveness. The model is used to show improvements over fixed models for forecasting.