为了分析在不同水平建成环境对公共交通出行选择行为的影响，基于计划行为理论和群体动力学理论从个体心理和群体两个水平提出存在主观因素参与的研究假设，考虑数据的嵌套结构特性，建立考虑建成环境的多水平结构方程模型，以潜聚合的方式将多个测量变量聚合形成高水平构念，同时从多个水平分析建成环境对居民公共交通出行选择行为的影响。利用柳州市居民出行调查数据，使用Mplus软件标定模型参数。结果表明：考虑建成环境的多水平和单水平模型均通过适配检验且多水平模型的拟合指数优于单水平模型；在不同水平变量对出行意向产生的影响存在差异。在个体水平各变量对行为意向的影响效应由大到小依次为：建成环境（0.775）、行为态度（0.704）、知觉行为控制（0.562）、主观规范（0.447）。在群体水平各变量对行为意向的影响效应从大到小依次为：行为态度（0.733）、建成环境（0.657）。可见所建立的多水平结构方程模型的测量和拟合能力良好，解释了在群体水平建成环境对公共交通出行选择行为的影响，克服了单水平模型的局限性，为优化居民出行结构，推动城市交通可持续发展提供理论支持。

In order to analyze the impact of the built environment on travel choice behavior in public transportation at different levels. On the basis of the theory of planned behavior and the theory of group dynamics, the research hypotheses of the existence of subjective factors at the individual psychological and group levels were proposed, and the multilevel structural equation model considering the built environment was established by consideration of the structural characteristics of the nested data, aggregating multiple measurement variables to shape a high level of constructs in the form of latent aggregation, while analyzing the impacts of the built environment on the residents' public transportation travel choice behaviors from multiple levels. The model parameters were calibrated with Mplus software using Liuzhou City residents' travel survey data. The results show that both the multilevel and single level models considering the built environment pass the fit test and the fit index of the multilevel model better than the singlelevel model, there are differences in the effects of the variables on the intention to travel at different levels. The impact of variables on behavioral intentions at the individual level has the following effects in descending order: built environment (0.775), behavioral attitudes (0.704), perceived behavioral control (0.562), and subjective norms (0.447). The impact of variables on behavioral intentions at the group level has the following effects in descending order: behavioral attitudes (0.733), built environment (0.657). This shows that the multilevel structural equation model constructed has sound measurement and fitting ability, explains the group heterogeneity that exists in the impact of the built environment on public transport travel choice behavior, overcomes the limitations of the single level model, and provides theoretical support for optimizing the travel structure of residents and promoting the sustainable development of urban transport.