关键词：PLOS，MOEs，PLOS，MOE，Fuzzy-Likert，TOPSIS，RIDIT，GRA，排名技术
摘 要：关于行人服务水平(PLOS)的评估，有大量的文献可用，这些文献使用不同的方法和不同的有效性度量(MOEs)或属性来描述PLOS模型。近年来，人们对在PLOS模型中捕捉行人交通流特征、建成步行环境和用户感知三种不同结构的共识越来越多。现有的PLOS文献一直在捕捉这些广泛的结构，但不是以一种组合的方式。本文探讨了负责开发这种公共科学图书馆的教育部，并以模糊-李克特(FL)量表记录专家意见调查。然后利用三种已建立的评级数据技术——TOPSIS、RIDIT are GRA对教育部进行排名，可进一步用于开发上述PLOS模型。从这些排名中可以看出，在专家们首选的前10个MOE中，有9个属于广义的设计构造类别（构建步行环境）和用户感知，只有一个属于广义的流动特征构造。这一结果强化了这样一个事实，即公共图书馆必须使用所有三个广泛的结构来创建，而不是像迄今为止所做的那样单独或成对创建。这一结果强化了这样一个事实，即PLOS必须使用所有三个广泛的结构来创建，而不是像迄今为止所做的那样单独或成对创建。本研究还探讨了使用FL量表与李克特量表作为反应测量工具的有效性，发现FL量表在测量顺序反应方面比李克特量表准确13.08%。
Abstract: A wide range of literature is available about assessing Pedestrian Level of Service (PLOS), which use different approaches and different Measures of Effectiveness (MOEs) — or attributes — to characterise the PLOS models. In recent years, there has been a growing consensus of capturing three different constructs in the PLOS model — flow characteristics of the pedestrian traffic, the built walking environment and the user's perception. Existing PLOS literature has been capturing these broad constructs, but not in a combined fashion. This paper explores the MOEs responsible for developing such a PLOS and records expert opinion surveys on a Fuzzy-Likert (FL) scale. Three established rating data techniques —TOPSIS, RIDIT are GRA are then utilised to get a ranking of the MOEs that could be further used to develop the said PLOS model. It is seen from these rankings that of the top 10 MOEs preferred by the experts, nine belong to the broad construct categories of design (built walking environment) and the user's perception, and only one belongs to the broad construct of flow characteristics. This result reinforces the fact that the PLOS has to be created using all the three broad constructs and not separately — or in pairs — as had been done so far. This study also deals with the effectiveness of using an FL scale compared to a Likert scale as a response measurement tool and found that an FL scale is 13.08% more accurate than a Likert scale in measuring ordinal responses.
Key: PLOS, MOEs, Fuzzy-Likert, TOPSIS, RIDIT, GRA, Ranking Techniques
全文: DOI