Determinants of Cyclist Frequency Choice Behaviour Using Multinomial Logit Model: A Case Study of Central Delhi
摘要
The prioritization of motorized transport in Indian urban planning has contributed to increased congestion, pollution and inequitable access. Non-Motorized Vehicles (NMVs), especially bicycles, offer a low-emission and inclusive mobility alternative but are constrained by infrastructural limitations and user perceptions. This study investigates the determinants of cycling frequency and infrastructure effectiveness in Delhi’s Zone D using a mixed-methods approach comprising infrastructure audits, user perception surveys and Multinomial Logit (MNL) modelling. Model diagnostics indicate good fit (Cox and Snell R2 = 0.537; Nagelkerke R2 = 0.591; McFadden R2 = 0.322). Key predictors of frequent cycling include personal safety (β = 1.96), secure parking (β = 1.91) and ease of access to cycle tracks. Factors such as pavement quality and gradient comfort also positively influenced cycling, while negative associations were observed with intersection design and signal phasing suggesting either design deficiencies or gaps in user awareness. These results highlight that cycling behaviour is influenced not only by infrastructure availability but also by user-specific perceptions shaped by experience and context. Findings underscore the need for integrated strategies that combine high-quality, continuous cycling infrastructure with user-centred design and education. Green spaces, safe crossings and end-of-trip facilities significantly enhance usage frequency. The study promotes for mainstreaming NMVs within urban mobility frameworks through context-sensitive planning, improved design standards and policy support. These measures are critical for fostering equitable, sustainable and health-promoting transport systems in rapidly urbanizing Indian cities.