|Title:||Recognising the irrelevance of statewise–dominated alternatives in defining the composition of a choice set|
|Authors:||Hensher, David A.|
|Keywords:||best-worst, first best, second best, choice models, choice sets, full rank, irrelevance of statewise dominated alternatives, road pricing reform, choice experiment|
|Abstract:||This paper is motivated by the primary idea (or curiosity) that the distribution of choice probabilities associated with a set of alternatives defining a given choice set provides strong evidence on the way that agents appear to process the description of each alternative in a stated choice experiment, conditional on other contextual influences that are agent specific. The supplementary interest is in the extent to which the established probability distribution, given the ranking of a set of alternatives, is able to be the basis of establishing whether a specific decision rule (within a utility maximising setting) offers the preferred behavioural ‘explanation’ of which alternatives really matter in choice making. Examples of interest include decision rules such as the relevance, in a rank order of alternatives, of all offered alternatives, variants of best-worst, and first best-second best. The underlying theoretical context to guide the preference ruling (or candidate alternatives under a rank order) is the Axiom of Irrelevance of Statewise Dominated Alternatives (ISDA) proposed by Quiggin (1995). In this paper we use a choice experiment on road pricing reform scenarios to illustrate a way to determine, under utility maximisation and knowledge of the full rank order of offered alternatives in a choice experiment, which set of alternatives satisfies ISDA and hence is a preferred choice set to use in estimation and application of a choice model.|
|Type of Work:||Working Paper|
|Appears in Collections:||ITLS Working Papers 2014|
|ITLS-WP-14-09.pdf||1.11 MB||Adobe PDF|
Items in Sydney eScholarship Repository are protected by copyright, with all rights reserved, unless otherwise indicated.