Scholarlink

<p>This study presents, to the best of our knowledge, among the earliest studies to integrate multidimensional Rasch-based psychometric framework with DIF analysis, clustering, and validation, for classifying motorized LOS at uncontrolled intersections under heterogeneous Indian traffic. A total of 1,150 responses were collected from 23 intersections across 10 Indian cities via revealed preference surveys. Driver satisfaction was assessed across five multi-dimensional quality-of-service (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(QoS\)</EquationSource> </InlineEquation>) dimensions: geometric design, traffic facilities, operational efficiency, pavement condition, and comfort and safety. The Rasch model enabled quantification of latent satisfaction scores and item difficulty levels associated with each attribute. These satisfaction indices (<InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(MDS_{j}\)</EquationSource> </InlineEquation>) were then clustered using Affinity Propagation to define six motorized level of service (<InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(MLOS\)</EquationSource> </InlineEquation>) categories (A–F), representing a user-centric service classification. Model validation using an independent dataset from six intersections demonstrated strong predictive reliability. A differential item functioning (<InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(DIF\)</EquationSource> </InlineEquation>) analysis further examined perceptual variability across vehicle types and driver experience levels. Findings reveal that attributes such as presence of bus stops, inadequate turning lane flaring, and substandard pavement quality are critical dissatisfaction drivers. In contrast, features such as leg count and average travel speed exhibit higher satisfaction. This study offers a behaviourally grounded classification methodology that integrates subjective driver perceptions with objective intersection performance metrics, thereby offering a perception-aligned framework that can inform mobility and safety improvements in heterogeneous urban traffic environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A psychometric framework for classifying motorized driver experience at urban uncontrolled intersections using Rasch analysis

  • 期刊论文

  • 发表日期：

    2026年3月31日

查看全文
  • Suprabeet Datta

摘要

This study presents, to the best of our knowledge, among the earliest studies to integrate multidimensional Rasch-based psychometric framework with DIF analysis, clustering, and validation, for classifying motorized LOS at uncontrolled intersections under heterogeneous Indian traffic. A total of 1,150 responses were collected from 23 intersections across 10 Indian cities via revealed preference surveys. Driver satisfaction was assessed across five multi-dimensional quality-of-service ( \(QoS\) ) dimensions: geometric design, traffic facilities, operational efficiency, pavement condition, and comfort and safety. The Rasch model enabled quantification of latent satisfaction scores and item difficulty levels associated with each attribute. These satisfaction indices ( \(MDS_{j}\) ) were then clustered using Affinity Propagation to define six motorized level of service ( \(MLOS\) ) categories (A–F), representing a user-centric service classification. Model validation using an independent dataset from six intersections demonstrated strong predictive reliability. A differential item functioning ( \(DIF\) ) analysis further examined perceptual variability across vehicle types and driver experience levels. Findings reveal that attributes such as presence of bus stops, inadequate turning lane flaring, and substandard pavement quality are critical dissatisfaction drivers. In contrast, features such as leg count and average travel speed exhibit higher satisfaction. This study offers a behaviourally grounded classification methodology that integrates subjective driver perceptions with objective intersection performance metrics, thereby offering a perception-aligned framework that can inform mobility and safety improvements in heterogeneous urban traffic environments.