Transit surveys based on on-board passenger interviews suffer from bias. Most commonly observed is the short trip bias: passengers travelling short distances are underrepresented in survey results. Biased data leads to an incorrect estimation of passenger demand can result in an inequitable allocation of revenues between transport operators.
This paper examines how the short trip bias can be mitigated during the data ex-trapolation process. Four methods are examined: A simple extrapolation by boarding counts, three iterative proportional fitting models and an additional weighting concept are tested on simulated survey data. The simulative approach enables the evaluation of the examined methods concerning their effects in reducing short trip bias. A total of eight survey situations with selected parameters variated are simulated to allow conclusions about influencing factors.
Results suggest that the most effective method is the weighting approach, followed by the iterative proportional fitting methods. Within the class of the iterative propor-tional fitting methods no significant difference is observed. Furthermore it is observed that the effectiveness of the weighting approach strongly relates to passenger numbers and selection rates.
Furthermore an overview on topic related literature is given to examine practical approaches to reduce bias in survey data.
Identifer | oai:union.ndltd.org:DRESDEN/oai:qucosa.de:bsz:14-qucosa-226060 |
Date | 04 July 2017 |
Creators | Neumann, Marcus |
Contributors | Technische Universität Dresden, Fakultät Verkehrswissenschaften 'Friedrich List', Dipl.-Verk.wirtsch. Stefanie Lösch, Prof. Dr. rer. pol. Ostap Okhrin |
Publisher | Saechsische Landesbibliothek- Staats- und Universitaetsbibliothek Dresden |
Source Sets | Hochschulschriftenserver (HSSS) der SLUB Dresden |
Language | deu |
Detected Language | English |
Type | doc-type:masterThesis |
Format | application/pdf |
Page generated in 0.0728 seconds