Seasonality in Regression: An Application of Smoothness Priors
This article argues that conventional approaches to the treatment of seasonality in econometric investigation are often inappropriate. A more appropriate technique is to allow all regression coefficients to vary with the season, but to constrain them to do so in a smooth fashion. A Bayesian method of estimating smoothly varying seasonal coefficients is developed, based on Shiller's (1973) approach to estimating distributed lags. In a sampling experiment, this technique outperforms ordinary least squares by a substantial margin. An application of this technique to the estimation of the demand for soft drinks is also presented.
|Date of creation:||1977|
|Date of revision:|
|Publication status:||Published in Journal of the American Statistical Association, 73, 1978|
|Contact details of provider:|| Postal: |
Phone: (613) 533-2250
Fax: (613) 533-6668
Web page: http://qed.econ.queensu.ca/
More information through EDIRC
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
- Michael C. Lovell, 1963. "Seasonal Adjustment of Economic Time Series and Multiple Regression," Cowles Foundation Discussion Papers 151, Cowles Foundation for Research in Economics, Yale University.
When requesting a correction, please mention this item's handle: RePEc:qed:wpaper:257. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Mark Babcock)
If references are entirely missing, you can add them using this form.