Impacts of Perception on the Incentive to Strategic Bias in SP Responses: a Case Study in Users' Valuation of Rail Passenger Rolling Stock.
H Lu, Jacobs Consultancy, UK; A Fowkes, M Wardman, ITS, University of Leeds, UK
This paper attempts to link individuals' perception and their SP choice making process to better understand this process. It is found that individuals' perceptions have significant impacts on the valuations found and on model calibration precision.
Impacts of perception on the incentive to strategic bias in SP responses: A case study in users' valuation of rail passenger rolling stock.
Hui Lu, Anthony Fowkes, Mark Wardman
Stated Preference (SP) methods have been used extensively in transport research and elsewhere both for demand forecasting purposes and to value the importance attached to different product features and travel attributes. Alongside the broader acceptance and wider application of SP methods, some practitioners have argued for greater openness in discussing what they see as significant concerns surrounding SP. A particular problem is the upward biasing of Willingness To Pay in situations where respondents have little reason to expect that they would actually have to pay.
This paper, based on work undertaken for a PhD thesis, looks at the case of improved rolling stock (railway coaches) in a city in northern England. An SP experiment was used with the two alternatives being two stock types, illustrated and reasonably well known in the area. It is generally accepted that there is a big difference in quality of experience between travel in the two stock types. It is less clear if short distance travellers are willing to pay extra to enjoy the superior rolling stock. We reported some results from the SP experiment at ETC 2006.
On the questionnaire, after the SP experiment, questions on respondents' perceptions of the experiment were introduced to probe the decision process of respondents and explore the influence of perceptions on the SP responses. Respondents' perceptions of the difficulty of choice making, familiarity with experiment alternatives, and their perceptions of the realism of attribute change were sought in order to better understand individuals' responses. Importantly, respondents were also asked if they thought it likely that fares would actually rise if the better quality rolling stock was introduced to their route.
This application formed a good basis for the exploration of the effect of such perceptions due to its size and rigorous design. Target sample size was determined after careful simulation work looked into how 'important' differences in valuations could be spotted with a reasonable statistical confidence. Responses were obtained from 1222 inbound morning peak hour travellers into central Manchester, yielding 10998 preference observations. Two possible methods of influencing bias in the SP responses were considered: (a) making the experiment more complex in the hope of masking its purpose, and (b) by adding a written warning, known in the environment and politics literature as 'Cheap Talk'. It is therefore possible to look at the interaction of the responses regarding perceptions with the effectiveness of these two methods. This paper attempts to link individuals' perception and their choice making process to better understand this process.
A standard MNL and Heteroskedastic Multinomial Logit model (HMNL) estimation were used to analyse the data. Individuals' perceptions were incorporated into the models to identify their influences on their Willingness To Pay for the improved rolling stock and also individuals' consistency in choice making.
The paper finds that adding a Cheap Talk script amends the valuation of improved rolling stock in the case study in the expected direction; however, bias may well remain. Adding more attributes to the SP experiment, i.e. increasing task complexity, does not show a significant impact on the estimation bias, but contributes to a higher error variance in responses.
Significant effects of respondents' perception on their valuation of new rolling stock have been found. Individuals' perceptions have significant impacts on the valuations found and on model calibration precision. Individuals' familiarity with alternatives in the experiment increased the valuation of the improved rolling stock and increased the estimation precision. Individuals' perceptions of potential price increase have an impact on the valuation of improved rolling stock and on the estimation precision. The more likely it was that respondents perceived the potential price increase, the fewer preferences were given to the improved rolling stock and respondents were observed to be more consistent in their choice making. These findings are of great importance to the theory of SP design and to SP design practice.
Association for European Transport