基于量表型选择集的出行信息需求有序选择模型

研究了城市出行者日常活动和出行所需要的关键信息, 提出了一种针对量表数据的量化分析方法; 以出行者信息需求为研究对象, 设计了一项基于李克特5级量表的问卷调查; 总结了出行个体对各类出行信息需求程度的排名, 分别构建并标定了驾车线路、目的地位置和实时路况信息需求的有序选择模型; 与多项Logit模型在参数显著度、赤池信息量准则和对数似然函数值等指标上进行对比, 以验证有序选择模型的有效性, 并对影响信息需求的关键变量进行了局部效应分析。研究结果表明: 出行者在通勤出行前最需要公交运营变更信息, 而在通勤出行中最需要实时路况信息; 非通勤出行者最关注目的地位置信息, 在出发前和途中对目的地位置信息的需求概率较通勤出行前分别高出31.08%和29.25%;在职人员对信息的需求程度普遍高于学生和自由职业者, 说明时间价值高的人群更期望通过全面及时的信息获取来合理安排出行; 女性比男性对实时路况信息的需求概率高出10.23%, 说明女性对延误的风险规避意识更强; 20~40岁的人群出行信息需求最强烈, 随着年龄的增长该需求将逐渐下降; 各年龄段对实时路况信息都表现出了较高的需求, 说明人们对于有可能带来负面影响的信息更为敏感。可见, 采用有序选择模型能够精确分析量表型选择集。 

Ordered choice model of travel information demand based on scale choice set

The key information needed for urban travelers' daily activity and travel was studied, and a quantitative analysis method focusing on scale data was proposed. Setting traveler's information demand as the study subject, a survey based on Likert 5 scale questionnaire was designed. The rankings of travel individual's demand degrees for various types of travel information were summarized, and the ordered choice models for driving route, destination location and real-time traffic information demand were built and calibrated separately. The ordered choice model was compared with the multinomial Logit model in terms of parameter significance, Akaike information criterion and log likelihood function value so as to verify its effectiveness, and the partial effect analysis on key variables affecting the information demand was carried out. Research result shows that the bus operation change information is mostly required before the travelers' commuting trips, and the real-time traffic information is mostly required during commuting. Non-commuting travelers care destination location information most, and the demand probabilities for destination location information before departure and en-route are 31.08% and 29.25% higher than that before commuting, respectively. In general, workers have higher information demand degree than that of students and freelancers, indicating that people with high value of time are more expected to properly arrange their trips by comprehensive and timely information acquisition. Demand probability of real-time traffic information for female is 10.23% higher than that for male, indicating that females have stronger cognition in avoiding delay risks. The 20 s to 40 s have the strongest travel information needs that gradually decline with age increasing. All age groups show a higher demand for real-time traffic information, indicating that people are more sensitive to information with potential negative impact. Therefore, the scale choice set can be analyzed by using the ordered choice model.