Design of a Strategic-Tactical Stated-Choice Survey Methodology Using a Constructed Avatar

Abstract

This paper presents findings from a small-sample qualitative study on people's activity travel behavior in the presence or absence of carsharing. A carsharing service provides its subscribers with short-term access to a fleet of shared cars. In previous research, subscribers have reported distinctive travel patterns, such as more car usage by subscribers than by non–car owners but less than by car owners. Reflexive techniques were used in which interviewees adapted a week of their recent activity travel behavior in response to stimuli. Findings from this study informed the design of a stated-choice survey that addressed three principal forms of complexity: (a) strategic-tactical choice situations, (b) situations in which respondents might select multiple interacting options in a single choice situation, and (c) situations in which sufficient knowledge of the individual survey respondent to tailor such a complex choice situation could not feasibly be gathered during a single interview. In the proposed design, the respondent indicates a strategic choice of which methods of travel to make available for use given a set of representative out-of-home activities. Accessibility to each activity by various means of travel is generated by using empirical distributions from Britain's National Travel Survey data sets to maximize plausibility of the information presented to respondents. An avatar (a virtual character for purposes of the survey) is constructed for each respondent on the basis of a small set of self-reported demographic characteristics. The use of multiday activity travel diaries would ideally involve multiple points of contact with each respondent at substantial cost. Therefore, an alternative method involving a single interview per respondent was sought.