Walking the talk: Comparing pedestrian ‘activity as imagined’ with ‘activity as done’

Gemma J. M. Read, Erin L. Stevens, Michael G. Lenné, Neville A. Stanton, Guy H. Walker, Paul M. Salmon

Research output: Contribution to journalArticlepeer-review

12 Citations (Scopus)
23 Downloads (Pure)


The safety of vulnerable road users, including pedestrians, is an important issue worldwide. In line with the shift towards systems thinking in transport safety, the aim of this study was to compare the normal performance of pedestrians as they navigate the road system with that imagined by road system managers to gain insights into how safety management can be improved for this vulnerable road user group. The Event Analysis of Systemic Teamwork framework was used to compare pedestrian activity ‘as imagined’ and ‘as done’ at signalised road intersections and railway level crossings. Data regarding ‘activity as imagined’ was derived from documentation review, and data on ‘activity as done’ was derived from a semi-naturalistic study of ten participants. It is concluded that in both environments pedestrians exhibited more diversity and variability than anticipated by system managers. Insights for improving the design of the road environment for pedestrians are provided. Further, it is argued that wider changes to the processes used in the design and management of road systems are needed.

Original languageEnglish
Pages (from-to)74-84
Number of pages11
JournalAccident Analysis and Prevention
Early online date7 Mar 2018
Publication statusPublished - Apr 2018


  • Event Analysis of Systemic Teamwork
  • Intersections
  • Pedestrian safety
  • Performance variability
  • Railway level crossings
  • Systems thinking

ASJC Scopus subject areas

  • Human Factors and Ergonomics
  • Safety, Risk, Reliability and Quality
  • Public Health, Environmental and Occupational Health
  • Law


Dive into the research topics of 'Walking the talk: Comparing pedestrian ‘activity as imagined’ with ‘activity as done’'. Together they form a unique fingerprint.

Cite this