Spatial models in transport: a review and assessment of methodological issues
Spatial models in transport: a review and assessment of methodological issues
复制标题
DOI：
--
复制
发表时间：
2012
期刊：
影响因子：
--
通讯作者：
Chao Wang;M. Quddus;T. Ryley;M. Enoch;Lisa Davison
中科院分区：
文献类型：
--
作者：
Chao Wang;M. Quddus;T. Ryley;M. Enoch;Lisa Davison
使用空间数据的回归模型广泛用于交通应用。尽管事实证明它们在识别和理解影响道路交通事故和旅行需求等现象的促成因素方面很有价值，但仍有一系列方法论问题限制了它们的实用性。本文探讨了空间模型中的一些问题。可修改的面积单位问题（MAUP），发生在统计推断和解释来自区域改变由于区域边界的变化。来自模拟和经验数据的证据证实了MAUP的存在。其他方法的问题进行检查，包括生态谬误，空间依赖性，以及匹配个人的意见，以正确的空间单位的问题。为了避免这些问题，建议采取一些战略，以综合的方式解决这些问题。最后，本文对全文进行了总结，并提出了进一步研究的方向.
Regression models that use spatial data are widely used in transport applications. Although they have proved to be valuable in identifying and understanding the contributory factors that influence phenomena like road traffic accidents and travel demand, there are a series of methodological issues that limit their usefulness. This paper examines some of these issues in spatial models. The Modifiable Areal Unit Problem (MAUP), occurs where the statistical inference and interpretation derived from the zones alters due to changes in zone boundaries. Evidence from both the simulated and empirical data confirms the existence of MAUP. Other methodological issues are examined, including ecological fallacy, spatial dependency, and the problem of matching individual observations to the correct spatial units. In order to avoid these problems, some strategies are recommended for addressing these issues in an integrated way. Finally, conclusions and further research avenues are discussed at the end of this paper.