Lexicographic answering in travel choice: Insufficient scale extensions and steep indifference curves?

Abstract

This study assesses lexicographic answering in stated choice surveys of travel alternatives. Respondents answering lexicographically in three different data sets are analysed in relation to how important they found the attributes that dominated their choices. Lexicographic answering is also regressed against covariates indicating commuting situation and socioeconomic status. A larger share of those answering lexicographically in relation to one specific attribute stated that this attribute was decisive in their choice compared to the share trading-off attribute levels in choices. Furthermore, a large majority of those answering lexicographically stated that the difference between the highest and lowest values of the attribute, according to which they chose lexicographically, was “very important”. Relevant variables explained lexicographic answering in a logistic regression analysis, e.g. that the probability of lexicographic answering with respect to travel time increases with income and travel distance. Response strategies other than neo-classical trade-off, e.g. simplification with a focus on one attribute alone, cannot be ruled out. However, the results indicate that lexicographic answering is due primarily to steep indifference curves in combination with insufficient attribute scale extension. These findings have implications for choice design and for the treatment of respondents answering lexicographically.