The role of perceived acceptability of alternatives in identifying and assessing choice set processing strategies in stated choice settings: The case of road pricing reform
Summary (2 min read)
1. Introduction and Conceptual Context
- Stated choice modelling focuses primarily on identifying the role of a set of attributes and attribute levels in establishing individual preference as an alternative/complementary approach to revealed preference modelling.
- Regardless of what information is used to construct a choice set response, without such knowledge it is not possible to establish, from the full set of alternatives, which subset of alternatives are processed in making a preferred choice, and indeed in establishing a rank order for a behaviourally meaningful set of alternatives.
- Asking whether an alternative is acceptable or not does not preclude the possibility that the alternative was actually considered when the final choice was made, and hence does not necessarily suggest that the alternative should be assigned a zero choice probability.
- The attribute descriptions of each elemental alternative are embedded in the CCSPS model together with explanatory variables designed to capture process heuristics such as extremeness aversion and to identify other possible ways in which the attribute levels are processed in establishing the acceptability of particular alternatives in defining a CCSPS.
2. Conceptual Positioning of the Proposed Acceptability Approach
- Hensher and Louviere (1983), in one of the very earliest choice experiments, show that an ‘ideal’ choice experiment is defined as one in which “… the basic elements of the choice process are abstracted and everything is controlled to permit unbiased estimates of choice strengths and choice probabilities.” (p.228).
- For such popular choice experiments, when the authors only know which alternative is chosen (or the full ranking of the alternatives), the choice of an alternative is by implication conditional on the full set of potential availability choice sets as defined by the statistical design.
- In total, seven candidate choice sets, each with a minimum of one acceptable alternative, can be defined and one of these candidate choice sets is considered by the respondent when they reveal their preference.
- The authors use a random parameter with error components model to take account of preference heterogeneity (through random parameters) as well as the correlated errors due to the presence of common elemental alternative across the choice sets (through error components).
- The model may approximate a covariance structures in a more accurate way than the typical nested logit model by forming complex covariance structures between alternatives.
3. Testing Choice Set Formation - The Empirical Illustration
- The authors use a data set collected in Sydney in 2012 that focussed on investigating commuters’ preferences for a number of alternative road pricing reform packages.
- After these choice responses, they were asked to indicate whether each of the three alternatives is acceptable or not acceptable.
- Given that the empirical topic is road pricing reform (involving cordon-based and distancebased charging), there is likely to be a wide range of awareness (or lack thereof) within a sampled population.
- The data includes personal income, age and gender.
4. Model Results
- To estimate the endogenous choice set model of three alternatives, the authors need to create multiple elemental alternatives (one for each candidate choice set processing strategy or CCSPS) from each alternative shown to the respondent.
- Thus, the observed utilities of elemental alternatives differ by either the CCSPS component (same initial alternative in different candidate choice sets) or the alternative component (different alternatives in the same nest).
- At the elemental alternative level, the authors find that all of the pricing and cost attributes are statistically significant, as are three of the revenue allocation plans (namely allocation to public transport, roads, and reduced income tax relative to being contributed to general government revenue and to compensate toll road operators).
- This suggests, for example, that a 10 percent increase in the range of the offered costs, ceteris paribus, will produce a 2.0 percent decrease in the probability of a respondent choosing the cordon-based charging scheme in a choice set where the acceptable alternatives are the cordon-based charge and a distance-based charge.
5. Conclusions
- The objective of this paper is to show how additional information on the perceived acceptability of an alternative, aligned with the literature on consideration sets, can be used to inform the relevance of specific choice sets that respondents find acceptable (up to a probability), given an imposed or offered set in a stated choice experiment.
- The authors use a random utility maximisation mixed logit error components model to consider the role of contextdependency, amongst other possible effects, in influencing the acceptable alternatives processed by the respondent.
- The empirical study which analyses the level of support for road pricing reform from car commuters, illustrates how researchers can take the imposed set of alternatives offered in a stated choice experiment, and use an acceptability response as a candidate choice set processing strategy to identify which set of alternatives are processed as the considered choice set.
- The authors found that accounting for the consideration sets of alternatives (akin to choice set generation) results in varying sensitivities to changes in attribute levels which differ from the findings when the choice of choice sets is not taken into account.
- What the authors have is a potentially important additional candidate criterion for segmenting markets just like they do with trip purpose, time of day, income, trip length and the like.
Did you find this useful? Give us your feedback
Citations
21 citations
13 citations
10 citations
7 citations
7 citations
Cites background or result from "The role of perceived acceptability..."
...…used by the respondents in revealing their preferences, there has been a limited amount of inquiry into the overall relevance of imposed alternatives in stated choice studies in contrast to the literature on choice set formation, especially in revealed preference studies (Hensher and Ho 2015)....
[...]
...Hensher and Ho (2015) investigate the relationship between the attribute levels associated with an alternative offered in a choice experiment and the construction of a respondent’s choice set based on the alternatives considered through the acceptability response dummy variable....
[...]
References
15,741 citations
"The role of perceived acceptability..." refers background in this paper
...This distribution is assumed to be explained by the utility version of Luce’s choice axiom (McFadden 1974)....
[...]
7,768 citations
1,682 citations
"The role of perceived acceptability..." refers result in this paper
...This is also in line with the literature on choice set formation set out in the context of revealed preference data (see Manski 1977 and Swait and Ben Akiva 1987)....
[...]
1,461 citations
1,281 citations
"The role of perceived acceptability..." refers background in this paper
...…is also normally supposed that a “betweenness inequality” holds in choice making, in which the middle alternative (for example, in a three alternative choice set) loses relatively more than an existing extreme alternative when another extreme alternative is introduced (Tversky and Simonson 1993)....
[...]
...…is the application of context-dependent heuristics such as extremeness aversion and compromise (proposed by Simonson and Tversky (1992) and Tversky and Simonson (1993)), that take into account the variations in attribute levels across a set of alternatives predefined in a choice…...
[...]
Related Papers (5)
Frequently Asked Questions (11)
Q2. What is the main purpose of stated choice modelling?
Stated choice modelling focuses primarily on identifying the role of a set of attributes and attribute levels in establishing individual preference as an alternative/complementary approach to revealed preference modelling.
Q3. What is the evidence that sanitising the full choice set?
The evidence suggests that the attribute dimensionality across the choice sets and within choice sets (i.e., the context-dependent effects) has a statistically significant role in sanitising the full choice set offered.
Q4. What is the effect of a 1 percent increase in personal income on the probability of choosing an?
In the choice set in which CB and DB are acceptable this becomes a positive effect, with a 1 percent increase in personal income resulting in a 0.80 and 0.87 increase in the probability of choosing the CB and DB alternatives respectively.
Q5. What is the informative way of presenting the findings?
A behaviourally more informative way of presenting the findings is through a set of direct elasticities which indicate the impact of a unit change in the level of an attribute on the probability of choosing a particular alternative conditional on a CCSPS and the probability of choosing a CCSPS (i.e., ∂lnPin/∂lnxikn).
Q6. What does the evidence suggest that there are segmentation effects?
If as a consequence of road pricing reform, for example, the authors find that the behavioural responses identified through direct elasticities for specific attributes vary across subsets of alternatives that define the choice making set, then this suggests that there are segmentation effects (or heterogeneous responses conditioned on relevant choice sets) that should be recognised, and the evidence herein hints that this does make a difference to the overall behavioural response (once the incidence of membership of each segment is known).
Q7. What is the way to study the choice of alternative?
The suggestion in a growing number of studies, cited above, that a respondent’s perception of the acceptability of an alternative might be a way forward to narrow down the subset of alternatives that define the domain in which the probability of choosing an alternative is maximised, has merit and is worthy of further consideration.
Q8. How do the authors examine the possible processing rules of alternative sets?
The authors examine these possible processing rules through innovative ways of introducing attributes into the utility expression of each ‘alternative’ as defined by a choice set, such as the attribute range across the alternatives in a choice set.
Q9. What is the effect of a full set of elasticities on behavioural responses to changing attributes?
A suite of elasticities offer behaviourally rich evidence on the sensitivity of behavioural responses to changing attribute levels when account is taken of the probability of subsets of alternatives being chosen in eliciting preferences for specific alternatives, given the acceptability or otherwise on each alternative in the full set of fixed offered alternatives common in stated choice experiments.
Q10. What is the way to include the acceptability of an alternative in the time of modelling?
Another way of including the acceptability of an alternative at the time of modelling is to assign a zero probability to alternatives that have been deemed to be out of the acceptable consideration set (Gilbride and Allenby 2004, Horowitz and Louviere 1995).
Q11. What is the probability of a respondent choosing a CCSPS?
In the choice set in which all alternatives are acceptable, the authors see a 1.9 percent increase in the probability of a respondent choosing the cordon-based charging scheme; however this increases to 5.3 percent where the only acceptable alternative is the cordon-based charge (CCSPS3).