摘要
因为要质疑用简单平均法测算的消费者信心指数算法的合理性,所以研究消费者信心指数的二级构成指数间的相互关系,研究发现,各二级指数间相互影响,二级指数间有信息重复。这会导致简单平均算法计算的信心指数放大了某些因素的影响,会产生系统性偏差,导致信心指数的失真。因此,用因子分析方法对消费者信心指数进行修正,研究发现修正后的信心指数与宏观经济变量的相关程度更高。
This paper questioned the calculation method of the Consumer Confidence Index(CCI). Based on the study of composition of CCI,the author found sub-indexes of CCI have mutual influence and each of them may contain duplicate information. Thus, the calculation method of CCI is an improper one, which will magnified the impact of certain factors. So, there will be systematic bias resulting in distortion. At the same time , this paper presents a new method to amend the CCI by means of factor analysis, and the author found the amended CCI has a higher degree of correlation with macroeconomic variables than original CCI does.
出处
《统计与信息论坛》
CSSCI
2013年第8期21-25,共5页
Journal of Statistics and Information
基金
首都经济贸易大学学科建设项目<学科与研究生教育重点学科:统计学国家级重点学科培养>(00591262122501)
关键词
消费者信心指数
时序模型
因子分析
:Consumer Confidence Index
time-series model
factor analysis