Nature of Spatial Data, Accuracy, and Validation

• Published in: Spatial Analysis Methods of Road Traffic Collisions pp. 164 - 175
• Main Authors: Loo, Becky P. Y., Anderson, Tessa Kate
• Format: Book Chapter
• Language: English
• Published: United Kingdom CRC Press 2016; Taylor & Francis Group
• Subjects: CIVIL ENGINEERING, SURVEYING & BUILDING; Psychological methodology; Collision Database; Tr Oa; Pedestrian Crossing Facilities; Spatial Data; Ch Ec; Collision Records; Collision Locations; Data; Road Names; Road Network Database; Road Collision; Road Network; Spatial Analysis; DBOARD; Collision Patterns; Collision Data; Traffic Collision Data; Da Ta; Network Phenomena; Ta Ba; Li Da; Case Study
• ISBN: 9781439874127; 1439874123
• DOI: 10.1201/b18937-13

Road collision data are subject to the same issues of any geographical data, namely uncertainties, inaccuracies, quality, scale, validation, and error. Road collision data can be fraught with challenges in terms of understanding the very nature of how it was collected, spatial resolution, accounting for inaccuracies, and ensuring data quality. This chapter reports on the nature of spatial road collision data and looks at issues of measurement, boundaries, transformation, time, coverage, scale, relevance, positional accuracy, and classifications. Inherently, road collision data are collected for administrative purposes rather than for scientific research; hence there is a need to validate and edit the data before conducting scientific analysis. Spatial concepts that are associated with road collisions are similar to those associated with many other geographical databases. However, the analysis of road collision data relies critically upon data quality and consistency in order to monitor and reduce road collisions. This chapter will outline some issues and examples on spatial data collection and use within the larger collision database.