A Generalized Random Regret Minimization Model

Abstract

This paper presents, discusses and tests a generalized Random Regret Minimization (G-RRM) model. The G-RRM model is created by replacing a fixed constant in the attribute-specific regret functions of the RRM model, by a regret-weight variable. Depending on the value of the regretweights, the G-RRM model generates predictions that equal those of, respectively, the canonical linear-in-parameters Random Utility Maximization (RUM) model, the conventional Random Regret Minimization (RRM) model, and hybrid RUM-RRM specifications. When the regretweight variable is written as a binary logit function, the G-RRM model can be estimated on choice data using conventional software packages. As an empirical proof of concept, the G-RRM model is estimated on a stated route choice dataset, and its outcomes are compared with RUM and RRM counterparts.

The eyes have it: Using eye tracking to inform information processing strategies in multi‐attributes choices

“…Previous studies have shown that ANA (or more generally attributes attention) is linked to eye movements during the choice (Balcombe et al, 2015;Krucien et al, 2017;Spinks & Mortimer, 2015). By examining how transitions are linked to choices, we can test the prediction of random regret minimisation (RRM; Boeri, Longo, Grisolía, Hutchinson, & Kee, 2013;Chorus, Arentze, & Timmermans, 2008;Chorus, 2010;Chorus, 2012;de Bekker-Grob & Chorus, 2013) that multi-attribute information is processed on an attribute basis. In comparison, random utility maximisation (RUM) does not impose a particular type of information processing.…”

Identifying resident preferences for bus-based and rail-based investments as a complementary buy in perspective to inform project planning prioritisation

“…The regret for any considered alternative j, denoted RegðjÞ; is the sum of all binary regrets of choosing alternative j over the non-chosen alternatives j 0 e J. The random regret minimisation model, proposed in Chorus et al (2008) and subsequently refined by Chorus (2010), has been shown to be able to accommodate the compromise effect.…”

“…Where an attribute performs well relative to other alternatives, an improvement generates only a small decrease in regret; whereas the same magnitude of improvement generates a larger decrease in regret if the attribute was performing relatively poorly to begin with. The RRM is just as parsimonious as the standard RUM model, and they both have an advantage over other models of contextual effects, which typically require the estimation of additional parameters (Chorus, 2010;Hensher et al, 2015b).…”

Modeling experienced accessibility for utility-maximizers and regret-minimizers

“…Recent studies (Chorus, 2010; submitted for publication) have highlighted the promising empirical performance of RRM (also when compared to equally parsimonious RUM-models) in the context of car-type choices, mode/route choices, parking choices and shopping location-choices. As discussed more in-depth in Chorus (2010), the main difference between RRM and its utilitarian (and equally parsimonious) counterpart -RUM's linear-additive MNL-model -lies in the fact that the RRM-based MNL-model does not exhibit the IIA-property (which states that the choice-probability ratio of any two alternatives is unaffected by the presence and performance of a third alternative), even when errors are i.i.d. That is, the ratio of choice probabilities of any two alternatives i and j depends on the performance of these alternatives relative to one another as well as relative to each of the other alternatives in the set.…”

“…In their words ''anticipating regret is a powerful predictor of future choices''. Second, and this is a more pragmatic reason for adopting a regret-based approach, a generic regret-based discrete choicemodeling approach for the analysis of risky as well as riskless choice 4 has recently been developed and successfully applied in a variety of travel choice-contexts (Chorus, 2010). This Random Regret Minimization-approach has important formal similarities with conventional utility-based discrete choice-approaches (such as the MNL-model (McFadden, 1974)) and as such can be relatively easily combined with these conventional approaches to form integrative accessibility-measures.…”