答案顺序对轨道交通乘客满意度的影响

Answer Order Effects in Passenger Satisfactionfor Urban Rail Transit

为减小被调查者身份的固有影响,得到问卷答案顺序对轨道交通乘客满意度的净影响,论文对同样本追踪法提出了一种省时省力的替代法。该法对表征身份的多属性值的排列组合进行编号,然后根据编号对两组样本进行随机筛选,获得了两组离散多属性同分布样本。论文设计了答案顺序一正一反两套问卷,并在武汉市轨道线上进行了随机抽样调查。利用T检验、U检验、信息熵等方法进行分析,发现3个可重复现象:1)反向问卷轨道交通乘客满意度均值在统计意义上显著大于正向问卷满意度均值;2)答案顺序对轨道交通乘客满意度的影响在统计学意义上是一个常数0.3;3)正向问卷稳定性大于反向问卷。为提高轨道交通乘客满意度调查结果的客观性,建议同时投放正反两类问卷。

To reduce the inherent influence of respondents'identity and get the net effect of questionnaire answer order in rail transit passenger satisfaction,an alternative method was proposed to save both time and labor towards backtracking method investigating the same person.Two samples of same multiple attribute distribution could be obtained by using the alternative method that first numbering the permutation and combination of multiple attribute value showing identity,then randoming filter and testing two samples according to the number.2sets of questionnaires were designed as the increasing answer order(1-10)and the decreasing answer order(10-1).And the questionnaires were random surveyed in Wuhan rail transit line.Using t-test and information entropy and other methods to analyze the filtering samples,three repeatable phenomena were found:one is that passenger satisfaction for urban rail transit mean value of decreasing answer order questionnaire is significantly higher than those of increasing answer order questionnaire;the other is that answer order effects in passenger satisfaction for urban rail transit is statistically constant 0.Third is that stability of increasing answer order questionnaire is higher than decreasing answer order questionnaire.To improve the objectivity of passenger satisfaction for urban rail survey,two sets of questionnaires were advised to be surveyed at the same time.