考虑社交属性的合乘选择行为研究

论文总字数：48016字

摘 要

本文主要介绍了在中国合乘出行市场中，社交属性对乘客之间互相选择行为的影响。较之与其他交通方式，合乘出行兼具环保、经济、舒适、便捷等特点，除此之外，合乘出行还具有成为一种新的社交方式的潜力，值得全世界范围内的进一步推广。然而，受到安全性、私密性的影响，大多数人仍然无法接受合乘作为其日常出行的主要方式。并且由于年龄、性格、学历各方面差异的影响，共同合乘的乘客并不一定能较好进行交流社交，从而产生尴尬、无法达到将这种出行方式作为社交方式的初衷。针对上述现象，本文以各类人群合乘对象选择的偏好为突破口，结合出行目的、出行线路、节省金额等外部因素，探究影响人们选择出行方式以及合乘对象的内在原因。

本文以对“滴滴出行”的用户与潜在用户进行的SP问卷调查为数据来源。在对清洗后的数据进行描述性统计分析后，对数据进行Ordered Logit 回归分析与Ordered Probit回归分析找出关键影响因素，并对关键影响因素之间的交互效应进行分析；然后对信誉度、长相、相识度三个社交属性变量建立广义费用模型，探讨并尝试量化不同环境中这三个变量对于不同属性乘客的价值；结合两种模型联立求解，进而分析社交因素间的交互效应、其他外部因素与社交因素的交互效应。论证合乘出行的网约平台在结合了乘客社交属性之后对于乘客匹配与定价收费的设想在未来实施的可行性。

关键词：社交属性；合乘出行；Ordered Logit回归分析；关键影响因素；交互效应；广义费用模型；选择偏好

ABSTRACT

This paper mainly introduces the influence of social attributes on the choice behavior of passengers in the ridesharing market of China. Compared with other modes of transportation, ridesharing is environmentally friendly, economical, comfortable and convenient. In addition, ridesharing has the potential to become a new social way, and is worthy of worldwide further promotion. However, due to security and privacy, most people still cannot accept ridesharing as the main way of their daily travel. And because of the differences in age, personality, and academic qualifications, the passengers who carpool together may not be able to socialize well, resulting in embarrassment and inability to achieve the original intention of using this mode of travel as a social way. In view of the above-mentioned phenomena, this paper takes the ridesharing preference of various people as the breakthrough point, and combines the external factors such as the purpose of travel, travel routes, and savings to explore the intrinsic reasons that influence people's choice of travel modes and ridesharing partners.

This article uses the SP questionnaire survey for users and potential users of “Didi” as the data source. After descriptive statistical analysis of the cleaned data, the data is subjected to the Ordered Logit regression analysis and Ordered Probit regression analysis to identify the key influencing factors, and the interaction between the key influencing factors is analyzed; then the credibility, appearance, and acquaintance are put into a generalized cost model, to explore and quantify the value of these three variables in different environments for passengers with different attributes; combines the two models to solve the interaction, and then analyzes the interaction effects among social factors, other external factors and social factors. Demonstrate the feasibility of the future ridesharing network platform for the passenger matching and pricing charges after combining the passenger social attributes.

Keywords: Social attributes; ridesharing; Ordered Logit regression analysis; key influencing factors; interaction effects; generalized cost model; choice preferences

目 录

第一章 引言 3

1.1研究背景及意义 3

1.2 国内外研究现状 4

1.2.1 合乘出行研究现状 4

1.2.2 社交属性行为影响相关研究 4

1.2.3出行行为方式选择的离散选择建模 5

1.3 论文主要内容及研究安排 5

1.3.1 主要研究内容 5

1.3.2研究重难点 7

1.3.3 章节安排及技术路线 8

1.4 本章小结 10

第二章 基于社交网络的网约车平台 11

2.1 现阶段国内外主要网约车平台的服务模式 11

2.1.1滴滴出行 11

2.1.2 Uber 11

2.2基于社交网络的网约车平台设想 12

2.2.1设想背景 12

2.2.2设想描述 12

2.3本章小结 14

第三章 社交属性对合乘行为影响因素调查与初步分析 15

3.1 SP调查方案设计 15

3.1.1 潜在因素的选择 15

3.1.2 定性调查 16

3.1.3专家修改意见 17

3.1.4重要因素交互影响调研 18

3.1.5 SP调查问卷最终设计 19

3.2问卷调查实施 22

3.3数据初步分析 23

3.3.1数据的清洗与筛选 23

3.3.2 描述性统计分析 25

3.4本章小结 28

第四章 网约车合乘社交属性影响因素分析模型研究 29

4.1 Ordered Logit模型与Ordered Probit模型介绍 29

4.2社交属性的交互影响 30

4.3关于面板数据的随机效应 30

4.4 本章小结 31

第五章 模型求解、验证与结果分析 32

5.1模型检验 32

5.1.1平行线假设检验原理 32

5.1.2检验结果 33

5.2模型结果 33

5.3模型解释 38

5.3.1单变量影响分析 38

5.3.2多变量交互影响分析 39

5.4乘客选择偏好分析 41

5.4.1广义费用模型的建立 41

5.4.2 长相偏好 42

5.4.3信誉度偏好 43

5.4.4相识度偏好 44

5.5 基于社交平台的合乘匹配定价 45

5.6本章小结 45

第六章 总结与展望 46

6.1全文总结 46

6.2发展建议 47

6.3问题与不足 47

致 谢 49

参考文献 50

附录1 SP调查问卷 53

社交属性对合乘行为的影响分析调查问卷（调查表2） 53

剩余内容已隐藏，请支付后下载全文，论文总字数：48016字