01595801

Component

Stated Preference (SP) surveys are widely used in studies across many disciplines, such as marketing research, transportation, economics, and environmental studies, to collect data on choice behaviour. Inconsistency in responses to choice tasks in SP surveys is well recognized and known as one of the most important cases of violation of axioms of preference based consumer theory. It has been attributed to different causes, including respondent inattention, the complexity of choice tasks, as well as to learning and fatigue effects. In this research the authors extend the newly proposed systematic consistency test (SCT) that detects inconsistent choices in complex surveys by proposing weighting measures to evaluate the level of inconsistency of respondent choices. The SCT is based on the assumption that each respondent has a given preference structure and that the set of her/his choices should be consistent with this structure. As such, choices that are not consistent with an individual’s observed preference structure are identified as inconsistent with his/her own choices. The proposed weighting measures are based on the volume of the intersection of decision cones representing knowledge inferred from decisions made, the geometric distance between the points representing choices, and lengths of vectors representing the subtractions of the differences between attribute levels of different choice tasks. This provides the opportunity to differentiate between cases where inconsistency is more noticeable and those with less noticeable differences in decision attribute levels. In this paper, the extended SCT procedure is applied to rate inconsistencies in data from two SP data-sets. The results show that using the geometric distance measure can improve the efficiency of the test.

This paper was sponsored by TRB committee ABJ40 Standing Committee on Travel Survey Methods. Alternate title: Detecting Inconsistent Responses in SP Surveys: Weighted Systematic Consistency Test.

01584066

16-6644

English

18p

2016

Transportation Research Board 95th Annual Meeting

Digital/other

Figures; References; Tables

Behavior; Logic; Stated preferences; Weighting

Inconsistencies

Data and Information Technology; Freight Transportation; Pedestrians and Bicyclists

Transportation Research Board Annual Meeting 2016 Paper #16-6644

TRIS, TRB, ATRI

Jan 12 2016 6:53PM