Research On Seasonal Adjustment Method Based On The Balanced Rotation Sample Survey

Most economic time series are proceed based on repeat sampling survey, which can incur the overlapping of the sample between months, season, or year. Thus, there are autocorrelational sampling errors in time series data. Traditional seasonal adjustment methods are used to do seasonal adjustment, assuming that error is a white noise, ignoring the correlativity of the error's structure. Seasonally adjusted based on the pattern of sampling survey started relatively late, and relevant researchers is a little. Until now, no relative researchers have built the general seasonal adjust model based on the successive sampling survey, considering from the data resources perspective.In order to develop seasonal adjustment accuracy, processed from successive sampling survey to study the influence of sampling error and correlativity to seasonal changes based on the balanced rotation sample survey. Then, this paper builds the general seasonal adjustment model, and uses Kalman filtering to estimate the parameters and use forecast error, error variance to evaluate the model accuracy. At last, taking 12?0 rotation scheme of China's urban household survey for an example, the seasonal adjustment model including or ignoring sampling error are established to verify the effectiveness of seasonal adjustment method including sampling error.The seasonal adjustment method based on balanced rotate survey, combining the statistics knowledge and econometrics knowledge, studied the seasonal adjustment problems of time series from survey view. This method not only makes the seasonal adjustment much fine-grained, but also expands the research range of seasonal adjustment. Besides, building general seasonal adjustment method, develops the accuracy of seasonal adjustment and also provides new ideas for the theory research of seasonal adjustment and offers the theory basic for the application research of seasonal adjustment so that it is of great value to be applied to the similar condition.