The Non-Parallel Slope In The Data Of Consumer Confidence Survey And The Modification Of The Index

The Consumer Confidence Index(CCI)is a measure of changes in consumers’psychological perceptions of consumption,which not only comprehensively reflects and quantifies consumers’ evaluation of the current economic situation and their subjective feelings about economic prospects,income levels,income expectations,and consumer psychological status,but is also a leading indicator of trends of economy and consumption.Most of the current research on CCI mainly focuses on the correlation between index changes and related macro indicators,but there is very little analysis of the raw data of the survey.This article uses the raw data of the "Haixi Finance,Tourism and Consumer Confidence Index" to conduct an empirical analysis of factors that affect the consumer confidence index from the perspective of a single interviewee.Since the data surveyed by the Likert scale is the ordering index,this paper establishes a ordinal logistic model.This paper conducts the Brant parallel slope hypothesis test on the established cumulative logistic model,which confirms that the original survey data of the Hercynian Index has the problem of non-parallel slopes.Therefore,the stereotype logistic model that takes into account the level information of the response variable and the non-parallel slope of the slope is adopted.In the study,the data are generally modeled in stages,ignoring the correlation between the period data.In fact,the effect of arguments on scoring should be only small or even unchanged in the adjacent period.In this paper,we first model the raw data of each sub-index by stages,and then perform Chow test on a period-by-period basis to find when structural mutations occur,and then merge the data in the periods without mutations to re-model.At last,based on the evidence of slope non-parallel of raw data,we propose to amend CCI,assigning nonequidistant scores to different options and assigning different weights to each sub-index;we judge the economic prosperity according to the actual situation of China’s economic,and establish a loss function,using genetic algorithm to find the optimal weight and score,to obtain CCI that fit economic better.