Seasonal adjustment and trend-cycle estimation.

The article offers information on various ways for revision of raw data in order to correct known non-sampling errors including replacement of imputed data, correction of previous data, and estimation of new birth. As mentioned, data from the Monthly Retail Trade Survey (MRTS) is classified into five components which include trend cycle, seasonality, trading-day effect, Ester holiday effect and irregular component. It also discusses the reconciliation process to restore additivity.