Attribute ordering and attribute non-attendance in discrete choice experiments

Abstract

Existing empirical evidence shows that both contingent valuation and discrete choice experiment (DCE) methods are susceptible to various ordering effects. Despite this, very few studies have analysed attribute ordering effect in DCEs and none of them investigated its potential influence on information processing strategies, such as attribute non-attendance (ANA). This paper tests for attribute ordering effects and examines whether the order of attributes describing the alternatives affects respondents’ propensity to attend to or ignore an attribute. To this end, we use a split-sample approach, where one sample received a DCE version in which the positions of the first and last non-monetary attributes are switched across the sequence of choice tasks compared to the other sample. The results show that attribute order does not affect welfare estimates in a significant way, rejecting the notion of procedural variance. However, significant differences in welfare estimates between the two samples are detected for several choice attributes once ANA behavior is accounted for in the estimated choice models. The share of respondents who ignored the attributes differs significantly between the two attribute ordering treatments. These findings suggest that attribute ordering is not detected under the standard assumption of full attribute attendance underlying discrete choice modelling, while the opposite is true when ANA is considered in the choice models.