The understanding and management of station stops continues to be a key issue in the operation of urban railways. This paper reports a statistical meta-analysis of passenger alighting and boarding rates from an expansion of a real-life worldwide data set which includes 34 different variables reflecting characteristics of passenger flow, rolling stock design, infrastructure and management actions. This has enabled the authors to identify, test hypotheses about, and quantify the impact of, previously-untested variables. A stepwise regression method using the R statistical package was proposed and developed into a more tractable model with fewer variables. This process eliminated those variables shown to provide no statistical explanation (including the presence of platform edge doors). Of the remaining 18 hypothesised variables, all provided some form of statistical explanation at the 90% level (or more) in one model or another. The results will help railways and transport authorities around the world manage station stops, through timetabling and appropriate investment.

车站站点的理解和管理仍然是城市铁路运营中的一个关键问题。本文报告了对乘客下车和登机率的统计荟萃分析，该分析来自一个真实的全球数据集，其中包括 34 个不同的变量，反映了客流、机车车辆设计、基础设施和管理行动的特征。这使作者能够识别、测试有关先前未测试变量的假设并量化其影响。提出了一种使用 R 统计包的逐步回归方法，并将其发展成一个变量更少的更易于处理的模型。这个过程消除了那些没有提供统计解释的变量（包括平台边缘门的存在）。在剩下的 18 个假设变量中，在一个或另一个模型中，所有这些都提供了某种形式的统计解释，达到了 90%（或更多）的水平。结果将帮助世界各地的铁路和运输当局通过时间表和适当的投资来管理车站。