摘要
不同背景下因调查量表和评价标准的影响而导致的顾客评价的变化分为三类:量表变化、基准变化和结构变化。合理的CSI计算方法需要准确地反映结构变化,不反映量表变化和基准变化。本文采用蒙特卡罗模拟方法比较了两种计算方法在三种变化下的CSI结果。模拟结果表明:通用方法不会反映量表变化,能灵敏地反映基准变化,同时准确地反映结构变化;改进方法不会反映量表变化和基准变化,而是能够准确地反映出结构变化。因此,改进方法优于通用方法。同时,当不同品牌的CSI分值较接近时,CSI的排序不稳定;而当不同品牌CSI分值差异较大时,排序具有稳定性。
In terms of the effect on the investigation scale and appraisal standard,the changes of customer evaluation were classified as three kinds：the scale change,the benchmark change and the structural change.If the computing method is appropriate,CSI should be unchanged in the case of scale change and benchmark change,but changed in the case of structural change.This paper compared CSI under three changing circumstances by using two different methods of calculation.According to the Monte Carlo simulation results,the revised computing method is preferred to the general one.Because by using the revised computing method,CSI keep the same under the scale change and benchmark change, and CSI changed in the same direction of the structural change.The Monte Carlo Simulation also showed that the order of CSI was stable when the difference of CSI was large,but the order of CSI was unstable when the value of CSI was close.
出处
《数理统计与管理》
CSSCI
北大核心
2010年第4期743-753,共11页
Journal of Applied Statistics and Management
基金
2009-2010年度中国保监会部级研究课题"我国保险行业顾客满意度提升研究"的资助
上海市重点学科建设项目(项目编号:B803)的资助
2010年度上海财经大学基本科研业务费项目的资助
关键词
顾客满意度指数
顾客评价
蒙特卡罗模拟
customer satisfaction index
customer evaluation
Monte Carlo simulation
