Abstract
Spurious comopnents of economical time series often cause the biased estimates of regression parameters. This paper deals with a revised estiamtion method of the parameter of Henshaw's regression model for spurious seasonal time series, which is based on Abraham-Box's Bayesian regression model. The weights for spurious observation are adjusted to obtain the insensitivity of the parameters to the spurious or additional errors. The characteristics of the posterior distribution of parameters, the estimates of weights, and the sensitivity of the regression parameter are illustrated by analyzing a set of economical time series.
