Abstract

Regression analysis with time series data is frequently used in marketing research. However, despite its popularity and ease of interpretation it cannot provide any information regarding the relations between marketing time series over different frequencies. This article proposes a new research tool, wavelet analysis, that when incorporated in regression analysis can provide some frequency domain insights about the effectiveness of marketing instruments over different cycles. In addition, by adopting appropriate regression-modeling techniques, wavelets can provide increased estimation and prediction accuracy of marketing causal effects.

Recommended Citation

Michis, Antonis A. (2009) "Regression Analysis of Marketing Time Series: A Wavelet Approach with Some Frequency Domain Insights," Review of Marketing Science: Vol. 7 , Article 2.
DOI: 10.2202/1546-5616.1083
Available at: http://www.bepress.com/romsjournal/vol7/iss1/art2