Measuring People’s Perceptions of Different Features of a New Type of Electric Vehicle Using Stated Preference
Peter Davidson, Peter Davidson Consultancy, Helen Porter, Peter Davidson Consultancy, Rob Culley, Peter Davidson Consultancy
Stated Preference surveys were undertaken in three small focus groups. Different estimation model techniques including Mixed Logit and Latent Class were applied, and the results used to develop a larger scale survey.
We used a Stated Preference exercise to understand and measure the features that potential users wanted for a new type of electric vehicle. This paper documents the research exercise, results and the findings.
A traveller’s decision to use an electric vehicle may be related to design features such as the number of seats or the inclusion of storage space for shopping or luggage, or their perception of the vehicle as well as the more conventional attributes of journey time and cost. All these features were explored using Stated Preference Surveys. This was undertaken initially as part of focus groups at three European locations, Lyon, Barcelona and Glasgow, at an early stage in the project to understand how different features of the vehicle are valued, in different countries and by different sorts of people including the mobility impaired and hence influence the design of the final vehicle.
Different estimation models have been used to analyse the data, and we show the results of applying Multinomial, Mixed Logit and Latent Class techniques which allow valuable insights to be revealed. The advantages and disadvantages of each are discussed. The samples from each group were small and we show how conclusions can be drawn from relatively small samples. The results showed marked differences in perception between different groups, with some groups very receptive to the new vehicle and others less so, underlining the importance of understanding the perceptions of different sorts of people from different backgrounds.
The results were used to design a second and more detailed Stated Preference survey undertaken on-line. The paper then goes on to describe how the results were used to do this including consideration of the relative merits of different design methodologies including orthogonal designs which can result in a large number of choice alternatives being required, and efficient designs where a limited set of choice alternatives was selected using statistical methods to optimise the games.
Association for European Transport