Since Palda's pioneering work investigating the dynamic relationship between sales and advertising, the marketing literature has contained many articles on the topic of sales response model building. Until recently, most of these articles have reported the construction of econometric models based on time series data. Recent applications of multivariate time series extensions of the work by Box and Jenkins have shown the usefulness of this methodology in building sales response models. The author discusses the distinctions between the econometric and time series approaches and, through a multivariate time series analysis, explores the competitive environment of an industry in which advertising is the main source of competition.
Testing for Additive Outliers in Seasonally Integrated Time Series
