Amending the Incentive for Strategic Bias in Stated Preference Studies: Case Study in Users' Valuation of Rolling Stock

Document Type

Journal Article

Publication Date

2008

Subject Area

infrastructure - rolling stock, mode - rail, planning - surveys, ridership - forecasting

Keywords

Valuation, Travel surveys, Stated preferences, Scenarios, Rolling stock, Railroad commuter service, Questionnaires, Projections, Passenger trains, Manchester (England), Forecasting, Commuter rail, Case studies, Bias (Statistics)

Abstract

Stated preference (SP) methods have been used extensively in transport research and elsewhere, both to forecast demand and to value the importance attached to different product features and travel attributes. Before introducing new or refurbished rolling stock as part of their franchises, British Train Operating Companies (TOCs) have often carried out SP surveys to investigate passenger preferences and to test whether the improvement of the fleet and services is enough to recover the cost through increased fares and patronage and thus to evaluate the costs and benefits of this investment. The rolling stock value derived from SP results is used not only for welfare appraisal and pricing but also for rail demand forecasting. Alongside the broader acceptance and wider application of SP methods, some practitioners have argued for greater openness in discussing what they have seen as significant concerns surrounding SP. The issue of the strategic biasing of SP responses was specifically addressed by examining the incentive compatibility of SP responses related to task complexity and the presence of specific antibias warnings in the questionnaire. Methods to amend incentive to bias were suggested. Adding cheap talk script reduces the overestimation of valuation of improved rolling stock in the case study; however, bias may remain. Adding more attributes to the SP experiment does not show a significant impact on the estimation bias, but contributes to a higher error variance in responses.

Share

COinS