基于结构方程模型的不同规模城市居民出行方式选择影响因素的关联性分析？？
Correlation analysis of travel mode choice for urban residents in different urban size based on structural equation model

针对目前城市居民出行方式选择的相关研究主要集中在个人属性对交通方式选择的影响方面，忽略了不同城市建成环境与出行方式的关联性分析问题。选取个人属性、交通供给属性和城市规模为外因变量，同时选取出行目的、出行时间、出行偏好、出行距离和出行方式选择为内因变量，采用结构方程模型（SEM）建立各影响因素与城市居民出行方式选择的关系结构；以3个不同规模的城市为研究对象，利用其居民出行调查数据对构建的SEM进行辨识与修正，采用卡方自由度比、适配度、调整适配度和均方残差4个指标对构建的SEM进行拟合优度检验，从而得到各因素对居民出行方式选择的正负影响及程度。结果表明：4个评价指标的结果均符合适配标准或临界值要求，表明构建的SEM较好；性别、职业和收入水平3个外因观察变量对“个人属性”具有显著的正影响；目的地是否有停车场对交通供给属性具有显著的正影响，目的地停车设施越完善、停车供给越充足，城市居民选择小汽车出行的比例越大；大城市和中等城市对城市规模有显著的正影响，大城市和中等城市的居民更偏爱公交车出行；个人属性、交通供给属性和城市规模3个外因潜在变量对出行方式有显著的正影响；出行偏好对出行方式有显著的正影响。研究结果对进一步优化居民出行方式结构、改善城市交通拥堵具有指导意义。
At present, the research on the choice of urban residents？？ travel mode mainly focused on the influence of individual attributes on the travel mode choice, ignoring the correlation analysis of urban built environment with travel mode. Three variables, including individual attributes, transportation supply attributes and urban size, were selected as exogenous variables in this paper. Five variables, namely travel purpose, travel time, travel preference, travel distance and travel mode choice, were selected as endogenous variables in this paper. The structural equation modeling (SEM) was used to develop the relationship between those influence factors and travel mode choice for urban resident. Three cities with different urban size were selected as research sample, and the developed SEM was identified and corrected using investigation data in three city. Four indicators, including the ratio of the chi？？square degree of freedom, goodness？？of？？fit index, adjusted goodness？？of？？fit index, and root mean square residual, were used to test the goodness？？of？？fit of the developed SEM. Finally, positive and negative influence and its degree of all influence factors on urban residents’ travel mode choice were obtained. The results show that the results of four indicators meet the adaptation criterion or critical value requirement, which indicates that the developed SEM is good. Gender, occupation and income level significantly positive influence on individual attributes. Whether the destination having a parking lot has significant positive impact on transportation supply attributes, which indicates that perfect parking facility and enough parking space will encourage urban resident choose car. Large and medium？？size cities have significantly positive impacts on urban size, which indicates that residents in large and medium？？size cities are likely to choose bus. Individual attribute, transportation supply attributes and urban size have significantly positive impacts on travel mode choice, and travel preference has also