环比增长率测算季节调整模型的评价和选择  

Evaluate and Select Seasonal Adjustment Model to Calculate Chain Growth Rate

作　　者：桂文林[1] 韩兆洲[1] 

机构地区：[1]暨南大学经济学院,广东广州510632

出　　处：《统计与信息论坛》2012年第10期17-24,共8页Journal of Statistics and Information

基　　金：国家哲学社会科学基金项目<新型工业化指标体系及测度方法研究>(10BJY050);广东省哲学社会科学基金项目<广东省宏观经济运行和预警研究>(09E-04)

摘　　要：季节调整使子年度数据可比,有利于环比增长率测算和经济监测。国际上季节调整模型众多,模型选择是季节调整的首要任务。以国际常用的X-12-ARIMA和TRAMO/SEATS模型的选择为目标、以两模型理论差异分析为基础、以中国2001—2010年的月度CPI数据为样本,通过谱分析方法检验剩余季节性、幂等、平滑间距和修正历史等方法检验模型稳定性、通过Friedman和Kruskal-Wallis等非参数方法检验季节稳定性,得出模型之间更具体的差异,为满足实践需要进行模型选择提供科学依据。The seasonal adjustment make sub-annual data comparable,is conducive to calculation of chain growth rate and economic monitoring.There are many seasonal adjustment models,and the choice of model is the primary task before seasonal adjustment.It is the goal to choice the international X-12-ARIMA and TRAMO/SEATS model,and select 2001-2010 China＇s monthly CPI data for sample.On the basis of theoretical differences of the two models,the spectral analysis is used to test the remaining season,Idempotent,Sliding Spans,Revision Histories and other methods are used to test the stability of the model,Friedman and Kruskal-Wallis non-parametric methods are used to test season stability.The more specific difference between models is got.The scientific basis is provided for model selection to meet the practical needs.

关 键 词：环比增长率 季节调整 质量评估 居民消费价格指数 

分 类 号：C811[社会学—统计学]