Travel Behaviour and Society 1 (2014) 7990

Contents lists available at ScienceDirect

Travel Behaviour and Society

journal homepage: www.elsevier.com/locate/tbs

Applications of theories and models of choice and decision-making

under conditions of uncertainty in travel behavior research
Soora Rasouli , Harry Timmermans
Eindhoven University of Technology, Urban Planning Group, Eindhoven, The Netherlands

a r t i c l e

i n f o

Article history:
Available online 27 January 2014

a b s t r a c t
The overwhelming majority of models in travel behavior research assume implicitly or explicitly that
individuals choose between alternatives under conditions of certainty. The validity of this assumption
can be questioned as in reality urban and transportation networks are in a constant state of ux. It is
important therefore that transportation researchers develop relevant approaches and models to analyze
and predict decision-making under conditions of uncertainty. Transportation researchers have tended to
adopt theories and models, originally developed in social psychology and decision sciences, and explore
their applicability to travel choice behavior. This invited review paper summarizes the most important of
these theories and models, and illustrates their application using examples of empirical research in travel
behavior research. Arguing that some basic assumptions underlying these theories may not mimic the
quintessence of day-to-day activitytravel behavior, the paper is completed by reecting on the limitations of these models and identifying avenues of future research.
2014 Hong Kong Society for Transportation Studies Published by Elsevier Ltd. All rights reserved.

Introduction
The analysis and modeling of human choice and decision-making has a long history in travel behavior research. This statement
should not come as a surprise, considering that the essence of travel demand forecasting is how people choose to participate in
activities, decide on the departure times for these activities, choose
a transport mode, decide where to execute their activities, choose a
route to arrive at the chosen destinations, etc. Moreover, modeling
consumer response to exogenous policies is essential in assessing
the impact and effectiveness of transportation policies, captured
in terms of changing attributes of the choice alternatives of
interest. Finally, understanding how choice behavior co-varies
with genders, income categories, age groups and other sociodemographics is crucial in evaluating social effects and equity of
transport investment and trafc management scenarios.
Over the last decades, the travel behavior research community
has applied a variety of theories and modeling approaches to individual and household choice and decision-making processes. Until
the mid 1970s, spatial interaction and entropy-maximizing models, based on the theory of social physics, dominated the eld
(Wilson, 1974; Batty, 1976). Later, random utility theory
(McFadden, 1974) and psychological choice theory (Luce, 1959)
led to the formulation and application of many discrete choice
Corresponding author. Tel.: +31 402474527.
E-mail address: s.rasouli@tue.nl (S. Rasouli).

models (Hensher, 1981; Ben-Akiva and Lerman, 1985). Soon the

multinomial logit model became the working horse of the profession, to be complemented by various more advanced, less stringent
models, such as the nested logit model and generalized extreme
value models, which avoided some of the rigorous assumptions
underlying the multinomial logit model. Lately, the mixed logit
model (Train, 2003), MDCV model (Bhat, 2005) and hierarchical
choice models (Walker and Ben-Akiva, 2002) have become rather
popular. In parallel, but less pertinent, researchers have also
applied rule-based models to capture decision heuristics (e.g.,
Arentze et al., 2000).
Regardless of the modeling approach and the underlying theory
of choice and decision-making, these models have in common the
assumption that decision-makers have perfect knowledge about
the attributes of their choice alternatives. At the moment of choice,
the values of these attributes are invariant, while individuals are
assumed to hold perfect and complete knowledge about these
attributes. All these models relate to choice and decision-making
under conditions of certainty.
In reality, however, the assumption that attribute values are
certain is not very realistic. When leaving home, individuals will
not be certain about their arrival time to the intended destination
as travel times exhibit inherent uctuation. When choosing public
transport, travelers may not be sure whether they will have a seat
as demand and therefore occupancy fundamentally varies on a
day-to-day basis. When choosing a place to park their car, drivers
may face unexpected queues, and parking garages may be full.

http://dx.doi.org/10.1016/j.tbs.2013.12.001
2214-367X/ 2014 Hong Kong Society for Transportation Studies Published by Elsevier Ltd. All rights reserved.

80

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

When choosing a route, drivers can never be sure about the congestion situation. Even if the chosen route is normally not congested, an accident or sudden adverse weather conditions may
cause signicant delays.
Thus, the state of the transportation system and the urban
envelope are inherently uncertain. Consequently, decision-makers
always face conditions of uncertainty when choosing departure
times, activities, destinations, transport modes, routes, etc. In that
sense, it is surprising that applications of theories and models of
decision making under conditions of uncertainty are relatively
scarce in travel behavior analysis. Moreover, the majority of studies, albeit also small in number, are concerned with uncertainty or
variability in the transportation system (Rasouli and Timmermans,
2012a, 2012b), but do not address how individuals make decisions
when facing such uncertainty and how it affects their activity
travel decisions. Considering the inherent uncertainty in the state
of the transportation system, the formulation and application of
(improved) models of decision-making under conditions of
uncertainty should be a eld of research of high priority in travel
behavior research. This paper is meant to provide readers with
some basic background information and a state of the art overview.
In particular, we will discuss the principles and applications of
expected utility theory, (cumulative) prospect theory and regret
theory as these models have dominated the scarce literature in travel behavior analysis on decision-making under uncertainty.1 In
each section, we will rst discuss the basic assumptions and specications of the model and some variations, and then discuss selected
results of empirical studies. In addition to providing a state-ofthe-art overview, we will also reect on the appropriateness and
limitations of these models to applications in transportation planning
and management, and suggest some avenues of future research.

Notation
Most models in travel behavior research on choice and decisionmaking assume that the probability of choosing a particular choice
alternative is some function of the attributes of the choice alternatives and a set of socio-demographic variables. The position of the
choice alternatives on these variables is represented by a single value. This represents choice and decision making under conditions
of certainty. In case of choice and decision making under uncertainty, the characterization of the choice alternatives is captured
in terms of probability distributions. It implies that individuals
are or cannot be sure about the exact state of the choice alternative
along these uncertain dimensions or about the outcome of his decisions. This is the realm of choice and decision-making under conditions of uncertainty.
To create an overall framework for the various models in an attempt to allow the researcher to compare the various approaches
with classic discrete choice model, in this section we will rst
introduce some notation. Expanding the notation suggested by
Liu and Polak (2007), assume that each decision maker i is faced
with a set C = {sn; C fSn  n  Ng of N risky choice alternatives
1
These three theories are just a small subset of theories and models of choice
behavior under uncertainty. For example, Hey and Chris (1997) list the following
additional theories that were motivated by the inability of expected utility theory to
explain observed behavior. Allais (1952) theory, Anticipated Utility theory, Cumulative Prospect theory, Disappointment theory, Disappointment Aversion theory,
Implicit Expected (or linear) Utility theory, Implicit Rank Linear Utility theory,
Implicit Weighed Utility theory, Lottery Dependent Expected Utility theory, Machinas Generalised Expected Utility theory, Perspective theory, Prospective Reference
theory, Quadratic Utility theory, SSB theory, and Yaaris Dual theory. Some of these
theories are special cases or generalisations of those discussed in this paper; others
are based on different concepts.
2
Although there are differences between risk and uncertainty, we will use the
terms interchangeably in the present paper.

(prospects, lotteries).2 Each choice alternative sn in C consists of a

set of J possible outcomes or states sn fSnj ; 1  j  Jg. Each outcome
j of the nth risky choice alternative is dened by a vector of observable attributes X nj fxnjk ; 1  k  Kg. Associated with each risky alternative is a set of given probabilities pn fpnj ; 1  j  Jg, such that
PJ n
n
n
j1 pj 1, where pj is the probability that outcome Sj is realized
n
in s .
Choice under risk implies that a decision maker has to integrate
(i) information about the attributes characterizing the risky outcomes, and (ii) information about the probability of each outcome.
We assume that each attribute k inuencing outcome j of prospect
n is valued according to mapping function h, which translates the
values of the observable variables xnjk ; 8j; k into valuation scores
mnijk ; 8j; k. In turn, the valued variables are integrated according to
function g to derive to arrive at an valuation mnij ; 8j of the jth outcome of the nth prospect. Thus,

mnijk hxnjk ; 8 i; j; k

mnij gmnijk ; 8 i; j

or,

mnij ghxnjk ; 8 i; j; k

3
uni

Finally, we assume that the overall valuation (or utility)

of
prospect sn by decision maker i is a function f of the valuations
of the possible outcomes j of the nth prospect, the given probabilities of these outcomes pn and a set of model parameters
ui fuir ; 1  r  Rg that characterize the decision making process
in risky situations. Hence,

uni f mnij ; ui pnj ;

j 1; :::; J

Expected utility theory

Principles
Expected utility theory can be traced back to the work of Bernoulli in 1738 to solve the famous St. Petersburg paradox. This paradox concerns the problem how much a single gambler should pay
a casino to enter a game in which he would toss a fair coin, doubles
a start gain of 1 unit every time a head appears, and the game ends
the rst time a tail appears. Thus, the gambler would win 2k1
units if the coin is tossed k times before the game ends. The exP 1
pected payoff of this gamble is equal to E 1
k1 2 1 Consequently, the expected payoff for the gambler would be an innite
amount of money, implying that the gambler should play the game
at any price. Yet, few people would consider paying any price, giving rise to the paradox: the discrepancy between what people
seem willing to pay to play the game and the innite expected value. The problem led to a considerable amount of work on decision
making under uncertainty. Neumann and Morgenstern (1947)
should be viewed as the basis of modern expected utility theory
in that they provided a set of axioms (completeness, transitivity
and continuity) to lay the foundation of modern expected utility
theory.
The most basic version of expected utility theory (EUT), especially used in a normative context, is the expected value model,
which states that the overall evaluation or utility uni of prospect
sn by decision maker i, given pn can be derived by taking the expectation of the outcome evaluations xnij 8j over the probability distribution pn. That is,

uni

J
X
j1

pnij xnij

81

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

By assuming a decision rule, expected utility can be linked to

choice. The expected utility model assumes that an individual i will
choose risky alternative sn from choice set C iff

uni > uni

8sn sn 2 C

When applied in a non-normative context, several lines of research

have been developed. Note that the prediction in the standard case
depends on (a) given probabilities, (b) outcome values rather than
the valuation of these values, and (c) deterministic decision rules.
Each of these components has been relaxed/replaced.
Subjective probabilities
In normative applications of expected utility theory, the probability of occurrence of a decision outcome is given. In real-life,
however, probabilities of outcomes will not be given, but have to
be construed by individuals. It may be assumed that to account
for this difference, given objective probabilities can be replaced
by subjective probabilities or beliefs pni , pni fpnij ppnj ;
1  j  Jg, where p is some probability weighting function,
so that

uni

J
X
pnij xnij

j1

For example, Chew and MacCrimmon (1979) suggested a weighting

function of the following form that gave rise to weighted utility
theory:

wsnj =

J
X
wsnj0 pnj0

j0

Sugden (2004) suggested using

wv nj v nj

where a is a risk attitude parameter. Hu and Mehrotra (2012) argued that the BoxCox transformation is promising. It can be expressed as:

wv nj

if a0

10

lnv nj if a 0

wkj ppnj pnj1 . . . pnJ  ppnj1 pnj2 . . . pnJ

pnij v nij

12

j1

The prediction of the expected utility model then depends on

the form of h(). Several functional forms have been used in the literature, the popularity of the function often dependent on the domain of application (cf. Stott, 2006). Table 1 gives an overview.
HARA is in fact a family of models; the CARA, CRRA and DCRA
discussed below can be shown to be special cases.
Fig. 1 displays the nature of these value functions.
Risk attitudes
A major disadvantage of the most basic expected utility model,
the expected value model concerns its insensitivity to the dispersion of the utilities of the possible outcomes of the choice alternatives: decision-makers are assumed to be risk-neutral. To account
for the effect of dispersion and different risk attitudes, several nonlinear transformations of gmnij ; ui have been considered, where u
is a vector of parameters that affect the shape of g and capture
the decision-makers risk attitudes. If gmnij ; ui is concave, the expected utility of a risky prospect will be lower than the expected
value, implying risk averse behavior. In contrast, if gmnij ; ui is convex, the expected utility of a risky prospect will be higher than the
expected value of the prospect, implying that risk-seeking behavior
is captured.
Arrow (1965) pointed out the compensation required for a risk
averse individual to accept a gamble. He calculated the probability
as

1 1 u00 x
 d
2 4 u0 x

13

where d is payoff for the gamble. By dening absolute risk aversion

as

a

u00 x
u0 x

14

which is a positive constant and integrating Eq. (14), we obtain

The value p

...
is the subjective weight attached to the probability of receiving a payoff at least as good as
the payoff of j, while ppnj pnj1 . . . pnJ is the subjective weight
attached to the probability of receiving a payoff strictly better then
the payoff of j. Obviously the choice of p(.) is critical in specifying a
rank dependent utility function. If p(.) is concave then p(p) 6 p for
all p, which would represent a pessimistic individual, who
over-weights lower ranked outcomes (bad states of the world)
and under-weights higher ranked outcomes. Vice versa, a convex
p(.) represents an optimistic individual, who over-weights the
higher ranked (good states of the world) outcomes.
pnJ

15

Eq. (15) is known as Constant Absolute Risk Aversion (CARA).

By dening relative risk aversion which is dependent on expected value as

b

xu00 x
u0 x

16

integrating, and separating results by b = 1 and b 1, the CRRA

(Constant Relative Risk Aversion) utility can be derived:

11

where p(.) is a nonlinear weighting function with p(0) = 0 and

p(1) = 1.
pnj1

J
X

ux eax

This transformation was also used in de Lapparent (2010).

Quiggin (1982) developed rank-dependent expected utility theory. While the decision weights are independent of the choice set
in the previous models, according to this theory, the value of an
outcome depends on the probability of realizing that outcome
and the ranking of the outcome relative to the other possible outcomes. More specically, given an increasing order of outcomes,
the decision weight is equal to:

pnj

uni

( v n 1

Valuation of outcomes
Secondly, rather than using the objective outcome values, the
valuation of these values has been used. It follows that

Table 1
Functional forms value function.
Name

Equation

Linear
Logarithmic
Power
Quadratic
Exponential
Bell
HARA
Log quadratic
Sumex
Linear  expo

m(x) = x
m(x) = ln(a + x)
m(x) = xa
m(x) = ax  x2
m(x) = 1  exp(ax)
m(x) = bx  exp(ax)
m(x) = (b + x)a
m(x) = ln(x+1)+a(ln(x+1)2)
m(x) = a exp(bx) + cexp(dx)
m(x) = (ax + b)exp(cx)

82

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

Fig. 1. Value function for different parameter values by different methods.

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

ux lnx if b 1
ux

x1b
1b

17

if b1

18

The HARA (Hyperbolic Absolute Risk Aversion) specication is

most exible in that it encompassed both constant and relative risk
aversion. By dening risk tolerance as the inverse of risk aversion,
we obtain

u0 x
u00 x

19

Assuming linear function ax + b for this tolerance, HARA

emerges as a simple hyperbolic function of x:

u00 x
1

u0 x ax b

20

If a = 0, the CARA models is obtained; if b = 0 HARA turns into CRRA.

For the general case a 0; b 0 and assuming a > 0 and x >  ab
and integrating



b
if a 1
ux log x

21

1a1

ux

x ab
1  1a

if a1

22

Probabilistic choice
Classic expected utility theory assumes deterministic choice
behavior. Individual i will choose risky alternative sn from choice
0
0
set C iff uni > uni 8 sn sn 2 C. This assumption of deterministic
choice is however unrealistic as individuals may demonstrate different choices when faced with the same choice problem. In travel
behavior research, Polak et al. (2008) therefore suggested to combine expected utility theory and random utility theory, which however can be interpreted as replacing the assumption of
deterministic choice with that of probabilistic choice.3 This is
accomplished by adding an error term to the utility of the prospect,
as commonly assumed in random utility theory.
Consequently, assuming an additive utility function, Eq. (4)
then becomes

uni

J
X
f v nij ; ui pnj eni

83

assumptions with respect to the nature of the preference function,

choice rules and/or distributions of error terms, constant error
models, probit models, Luce choice models, generalized extreme
value models, etc. can be considered.
Examples
Polak et al. (2008) investigated the inuence of travel time variability on mode and departure time choice, using a sample of 215
respondents drawn from travelers the LondonBirmingham and
LondonReading corridors. Respondents were presented with eight
choice scenarios regarding their trip, which were varied in terms of
uncertainty in their arrival time at the destination due to variability in travel time. Respondents, who indicated to have a choice of
mode, were requested to make within and between mode comparisons. Respondents, who did not have any other transport mode
available, were only asked to make within mode comparisons. Each
alternative was described in terms of a known departure time, travel cost and a distribution of possible travel times. This resulted in
a distribution of possible arrival times, relative to a preferred arrival time. Respondents were shown ten equiprobable positive or
negative delays, relative to the preferred arrival time, xing the
range of variation in travel time with reference to the duration of
the reported trip and the exibility in arrival time at the destination reported by respondents. They compared the performance of
the expected value model and a constant absolute risk aversion
model, also allowing for taste variation. Findings indicated that
the latter type of model outperformed models based on expected
value, suggested that risk attitude plays a role in choice of departure time. Particularly, in this study, travelers were found, on average, to be mildly risk averse. However, results evidenced a
substantial degree of heterogeneity in risk aversion.
De Palma et al. (2007, 2012) analyzed the decisions of drivers
whether to acquire information and which route to take on a simple congested road network. They considered four information regimes: No information, free information, information available for
a fee, and private information, which is available free to a single
individual. Based on their assumption, they showed that private
information is individually most valuable, while the benets from
free and costly information cannot be ranked in general. They also
derive that free or costly information can decrease the expected
utility of sufciently risk-averse drivers.

23

j1

Prospect theory
n
i

where e is an unobservable component of utility associated with

risky choice alternative n for individual i and mnij Z nij gnij , where
Z nij and gnij are respectively the observable and unobservable components of the value function associated with outcome j of risky alternative n for individual i. The probability than individual i will then
choose risky alternative sn is

"
#
J
J
X
X
0
n
n
n
n0
n0
n0
pi s jC Pr
f v ij ; ui pj ei >
f v ij ; ui pj ei
8 sn sn
n

j1

j1

2C
24
Assumptions about the error terms commonly used in discrete
choice modeling can be used to calculate choice probabilities. Most
applications in travel behavior analysis have used simple logit or
mixed logit formulations. However, dependent on theoretical
3
As in random utility theory, alternative interpretations can be given to the error
terms, ranging from an analysts imperfect knowledge and/or measurement of the
relevant attributes, individual taste variation, stochastic preferences and deterministic choice, deterministic preference and probabilistic choice, and even uncertainty.

Principles
Kahneman and Tversky (1979) questioned the validity of expected utility theory as a descriptive theory of human choice
behavior. To explain violations of normative expected utility theory as reected peoples actual choice behavior, such as Allais paradox and the empirical nding that people tend to be risk-averse
over high probability gains but risk seeking over high probability
losses, with magnitude of losses higher than of gains,4 they
formulated an alternative theory of decision making under risk,
called prospect theory. According to their original formulation,
decision-making processes consist of two stages. In the rst stage
the editing stage various decision rules are used to frame
possible outcomes in terms of gains and losses, relative to some
neutral reference point. Gains refer to outcomes that exceed this reference point, while losses refer to outcomes that fall short. In the
second stage the evaluation phase the decision maker evaluates
the outcomes of each alternative according to some value function
4

Other examples are discussed in Avineri and Prashker (2004).

84

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

and transforms objective probabilities into subjective probabilities

according to a non-linear probability weighting function. To account
for violations of classic expected utility theory, the utility function
should be concave over gains and convex over losses. Empirical ndings of risk-seeking versus risk-avoiding behavior require the disutility of a loss to be valued higher than the utility of an equivalent gain.
Let s dene a reference point in the outcome domain. Prospect
theory then states that the utility of prospect n is dened as:

uni

J
X

ppnj v nij xnj  s

25

j1
0

Risky choice alternative sn is preferred to sn iff

uni > uni

8sn sn 2 C

26

Tversky and Kahneman (1992) suggested the following functional

form for the value function:

n n
ij xj

 s

xnj  sa

if xnj  s  0

kjxnj  sjb

if xnj  s < 0

27

Parameter k > 1 captures the degree of loss aversion, while

parameters a, b < 1 measure the degree of diminishing sensitivity
to change in both directions from the reference point. The curve
at zero, being steeper for small losses than for small gains implies
risk averse when outcomes are considered a loss and risk seeking
when outcomes are considered a gain, implying that people tend
to be more sensitive to decreases of their wealth than to increases.
The probability weighting function p is a monotonically
increasing function, with discontinuities at 0 and 1, such that it
systematically overweighs small probabilities and underweights
large probabilities. Several functional forms have been suggested
for the decision weight function, as shown in Table 2.
Fig. 1 graphs these weighting functions for different values of k
and u.
It shows that some of these probability weighting functions
have a reference point, while others do not, and that some functions are symmetric, while others are not. In many cases, it will
be impossible to reasons theoretically why one function would
be better than any other. Hence, researchers should compare different specications. In travel behavior research, applications have
typically applied the functions suggested by Tversky and Kahneman, Wu and Gonzalez and Prelec. Fig. 2 graphs these weighting
functions for different values of k and u
Later, Tversky and Kahneman (1992) extended prospect theory
by including rank-dependent probabilities, allowing for different
probability weighting for gains and losses (Palma et al., 2008). In
particular, decision weights in this so-called cumulative prospect
theory are equal to:

ppnj wpnj pnj1 . . . pnJ  wpnj1 pnj2 . . . pnJ

j 1; . . . ; J  1 and ppnj wpnj

28

Table 2
Functional forms for decision weight function.
Goldstein and Einhorn (1987)

kpn

j
ppnj kpnu 1p
n u
j

Tversky and Kahneman (1992)

Prelec (1998)
Prelec II
Wu and Gonzalez
Luce

ppnj

pnj
u
pnj

1=u
1pnj u 

ppnj explnpnj u
ppnj expklnpnj u
u

ppnj

pnj
u

u k

pnj 1pnj 

Eq. (39) shows that cumulative prospect theory belongs to the

broader class of rank dependent utility models.
Examples
Relatively late, in the early 2000s, transportation researchers
started to explore the adequacy of (cumulative) prospect theory
to predict traveler behavior under uncertainty. Most applications
of prospect theory have been concerned with departure times
and route choice behavior under travel time uncertainty.5
In examining the applicability of prospect theory, Jou and
Kitamura (2002) assumed two reference points: earliest acceptable
arrival time and ofcial work start time. Gains and losses were dened as non-linear functions around these two reference points,
resulting in four-segmented value functions on the earliest arrival
time, and the work starting time for a given commuter. No empirical results were reported. Later, Senbil and Kitamura (2004) added
preferred arrival time. Senbil and Kitamura (2006) incorporated
delay/early arrival travel time variability directly in the utility
function
Schwanen and Ettema (2007) (see also Schwanen, 2007) also
examined the applicability of cumulative prospect theory. Their
problem concerned picking up children at the day care. In particular, they examined the role of reference points in a cumulative
prospect model. Three reference points were used: (i) The time
that most other parents pick up their child; (ii) the time imposed
by day care management that the children should be picked up,
and (iii) the time that the day care ofcially closes. In the choice
task, only two possible arrival times, rounded to 5 min as much
as possible, were shown. Probabilities were used as commonly
encountered in everyday life and/or rounded to 5% or 10%, and presented in terms of in numbers (1 out of x cases) and percentages.
Some parameters of the cumulative prospect model were arbitrarily set. A logit specication and a genetic algorithm were used
to calibrate the model. Estimated parameters indicated the value
function is slightly convex for early or on time arrivals and concave
for late arrivals, supporting the commonly assumed inversely
S-shaped form. The authors also found evidence of loss aversion,
albeit of small magnitude. Finally, they found evidence of overweighting of small probabilities and underweighting of larger
probabilities, but again the effect was relatively small. The model
based on cumulative prospect theory tted the data better than a
model, based on expected utility theory.
Xu et al. (2011a,b) contend that the budgeted time to ensure
that a traveler arrives at a certain destination with some desired
probability serves as the reference point. This time will depend
on trip purpose and risk attitude. A low desired probability implies
that either the trip is not important or the individual is risk seeking. The budgeted time reects their subjective beliefs on travel
times based on their preferences, and can thus potentially serve
as reference points (e.g., Lo et al., 2006). Through past experiences,
travelers are assumed to develop a reference time for each O-D pair
which is a function of the budgeted time for each path. In particular, it is assumed that the reference time (point) is the minimum of
the budgeted times of all paths. Heterogeneity is accounted for by
grouping travelers of a OD pair into different classes with respect
to their desired on-time arrival probabilities. A prospect-based
user equilibrium model where the resulting reference points are
consistent with the equilibrium ow pattern was derived.
Earlier, Connors and Sumalee (2009) also investigated network
equilibrium for different prospect theoretic values of risky links
in a hypothetical network. Based on assumed values of key

ppnj exp uk 1  pnj k 

k measures the elevation of the weighting function and u > 0 represents its degree
of curvature.

5
Van de Kaa (2010) offers a considerably more detail overview of relevant studies,
not only in the context of route and departure time choice, but also in other domains.
Another recent review of the literature is offered by Li and Hensher (2011).

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

85

Fig. 2. Weighted probability function for different parameter values.

parameters of prospect theory, they found that a change in reference point values has a signicant inuence on the equilibrium
achieved in the network.
Other studies have looked at route choice behavior.
Katsikopoulos et al. (2000, 2002) found support for the applicability
of prospect theory in route choice behavior. Participants in a driving
simulator received descriptive information about travel time ranges
of two routes and were asked to choose a route under different
trafc scenarios. They found that if travel times on the alternative
route are on average shorter than those on the reference route,
participants demonstrate risk aversion in their route choice
behavior. In contrast, in case of losses, route choice risk-seeking
behavior is observed when the alternative route is riskier relative
to the reference.
Avineri and Prashker (2004, 2005, 2006) also applied prospect
theory in a route choice context. In their rst study, respondents

were invited to choose between different routes, characterized by

different probabilities of travel times. They found evidence of
non-linear decision weights and loss aversion. Masiero and
Hensher (2011) investigated effects of a negative shift of the reference point. A pivoted stated choice experiment framework was
used to estimate the indirect freight transport costs associated
with a temporary closure of a main road in Switzerland. Logistics
managers completed two experiments: the rst designed around
the initial reference alternative; the second after road closure
around the second best road alternative. Three pairs of models
were estimated. The rst pair of models is based on the assumption
of symmetry. The second pair of models is based on reference
dependence, estimating different parameters for gains and losses,
while the third pair of models further allows for asymmetric nonlinearity in gains and losses. Results provided support to the prospect theoretic assumption of a change in preference structure, due

86

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

to a shift in reference point. On average, respondents showed

evidence of increased loss aversion for cost and time attributes,
and a decrease in loss aversion for punctuality.
While these studies focus on stated choice data, Hu et al. (2012)
extracted travel time distribution from historical, revealed preference data. They compared the predictive performance of several
models based on expected utility theory and non-expected utility
theory and concluded that model ts are generally similar with
only prospect theory performing marginally better over expected
utility theory.
Han et al. (2005) applied prospect theory in a typical framing
context. They assumed that travelers compliance rates with travel
information does not only depend on the (in)congruence between
the information provided and subjective beliefs, but also on the
whether the information is provided in a positive or in a negative
sense. Another interesting study on framing was conducted by Fujii
and Kitamura (2004), who postulated that drivers perceive uncertain travel time as an interval, and that they frame travel such continuous to compose 5 alternatives of departure time choice in
terms of the dichotomy of being on time and being late.
The domain of application of original prospect theory was concerned with gambling. Consequently, the denition of gains and
losses is clear-cut. The denition of reference points and the corresponding gains and losses are however not naturally dened in
applications to route, departure and other choices in travel
behavior research. Moreover, reference points likely differ between individuals. Avineri and Bovy (2008) discussed three possible ways to set the value of the reference point in studies on
route choice behavior. A rst option is to use the mean or others
parameters of the travel time distribution, such as the median value or the mode, as a reference point. More specically, Avineri
and Prashker (2005) suggested to set the reference point to the
actual travel time experienced on average in the population of
the target traveler group, such as an average of about 30 min
one-way commuting time in many countries. In principle, the approach can be further detailed in terms of travel purpose, mode,
user group and/or spatial context. Avineri and Prashker (2004) argued that data may be derived from national (or regional) travel
surveys. However, such an approach would deny any regional
and local variability, and will thus likely introduce substantial
bias.
A second option is to ask travelers directly about their desired
or ideal travel times. The question here is whether the notion of
gains and losses is conceptually equivalent to the concept of desired or ideal travel times. Moreover, respondents may experience
difculty in expressing what they consider as desired and ideal travel times. Thus, the validity of measurements of desired travel
times and the value of the reference points remains ambiguous.
More methodological research into the measurement of reference
points and the factors affecting perceived reference points is
needed.
A third option is deriving reference point from stated/revealed
preference data. This approach would involve comparing model
predictions for different values of the reference point. Further research here is required, however, as the parameter space will show
many discontinuities.
A more general problem is that reference points are likely heavily context-dependent, and may not only depend on past experiences, but also on aspirations and social comparison. Moreover,
taste variation may be substantial.
Avineri and Bovy (2008) argued that reference-dependent preferences should be based on the perception of reference values.
Consequently, they advocated using fuzzy rather than crisp representation and extended the concept of reference point to a fuzzy
set of reference values, in order to address the vague perception
of reference travel times for travelers (see also Avineri, 2009).

The empirical studies summarized in this section, indicates that

the travel behavior literature shows some examples of applications
of cumulative prospect theoretic models. However, as argued by Li
and Hensher (2011), most of these studies have been limited in
scope and approach either because (i) they only included particular
aspects of prospect theory, (ii) they did not estimate all parameters
in a consistent manner, and/or may not have treated the issue of
(multiple) reference point in the best possible way. We would
add that compared to random utility models, most studies have
also been relatively weak in accounting for stochastic valuation,
probabilistic choice and taste variation.
Regret theory
Principles
Seminal regret theory as formulated by David (1982), Fishburn
(1982) and Loomes and Sugden (1982, 1987) was originally as a
model of pairwise choice between lotteries (Machina, 1987). It is
concerned with the problem of which of two risky choice alternatives, characterized by a set of states that may occur with some
probability and that have a (monetary) outcome, will be chosen.
In contrast to expected utility theory and prospect theory, regret
theory is based on the notion that individuals utility or satisfaction
of choosing an alternative is not only based on the anticipated payoff of each individual choice alternative across different states of
the world, but also on anticipated payoff of the other alternative.
More specically, individuals are assumed to consider the possibility that the non-chosen alternative turns out to have a higher payoff than the chosen one after all. Thus, regret theory incorporates
the idea that the utility individuals derive from the outcomes of
their decisions is inuenced by their perception of what would
have occurred if they had made different choices. Individuals feel
regret when they experience that they would have been better
off if they had chosen another alternative, and rejoice when they
experience that they are better off. In some choice situations, these
emotions are responses to ex post evaluations of choices made, relative to foregone alternatives. In other choice contexts, regret is
based on subjective anticipations. Individuals are assumed to
trade-off the utility purely derived from the attributes of the choice
alternatives against their desire to avoid and minimize future regret (and maximize future rejoice). More specically, individuals
are assumed to anticipate for each possible state of the choice
alternatives the associated regret, dened as the extent to which
the chosen alternative performs worse than the non-chosen one),
and then aggregate these regrets across all possible states of the
world. Thus, regret theory assumes that alternative sn will be cho0
0
sen to alternative sn jsn ; sn 2 C iff:
J
X
pnj Rsn sn0 snj  > 0

29

j1

with
0

Rsn sn0 snj Rsn0 sn snj

30

and R is regret.
Note that the reference point in regret theory depends on the
composition of the choice set and the distribution of attribute values across choice alternatives, not on some (context-dependent)
arbitrary reference point. It rules out any choice mechanism that
is based on aspiration levels that are not currently attained.
Chorus et al. (2008b) explored the applicability of regret theory
to model travel choices, both under conditions of certainty and
uncertainty. Such application needs generalization of the basic
model from pairwise choice to choice among multiple (risky)

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

choice alternatives, and from a single attribute to multiple attributes. Adopting Quiggins (1994) principle of Irrelevance of Statewise Dominated Alternatives, which states that a choice from
any given choice set is not be affected by adding or removing an
alternative that is inferior for every state of the world, the regret
associated with any choice, assuming a state of the world to occur,
depends only on the actual outcome and the best possible outcome
that could have been attained for that state of the world. It implies
that for any state of the world, regret associated with a choice
alternative only depends on the best available choice alternative.
To generalize to the multi-attribute case, Chorus et al. (2008a)
postulated that choice alternatives are valued in terms of the associated regret on an attribute-by-attribute basis. More specically,
regret associated with choice alternative sn, based on a comparison
0
of a particular attribute with alternative sn is equal to zero if sn is
n0
valued equal or better than s on that particular attribute, and a
non-decreasing function of the difference in attribute-values
otherwise. Formally, the linear regret function for attribute k can
be expressed as:
0

wk max j0; maxfxnjk  xnjk gj; sn 2 C

31

and

Rsn

X
wk

32

Evidently, the linear regret function is a straightforward formulation, but may not sufciently capture the magnitude of regret and
its relationship with attribute differences. Chorus et al. added a
parameter to capture non-linearities in regret. They added this
parameter to the overall regret function. Nonlinear regret, dened
at the level of each attribute, would be equal to:
n0

n0

33

Compared to linear regret, a concave regret function emphasizes

the importance of small preference differences, whereas a convex
regret-function discounts differences. If # =1, across all attributes,
regret minimization reduces to utility-maximization.
Later, Chorus (2011) (see also Chorus and Bierlaire, 2013) suggested a different non-linear function while in addition it was assumed the comparison is not only based on the best choice
alternative, but on all foregone alternatives. Thus,

N X
K
X

ln j1

expbk fxnk

off again with high differences in attribute values. Such preference

would be captured by a non-symmetric S-shaped curve, as suggested in Timmermans et al. (2000). However, this function is more
difcult to estimate.
The comparison against all alternatives rather than the best
foregone alternative is also not necessarily an improvement. It is
reasonable to assume that if an individual could have chosen multiple choice alternatives that perform better that the chosen one,
regret is higher than if only a single non-chosen alternatives performs better, but cognitively this may be quite demanding, particularly in the context of a potentially very large number of
overlapping routes.
Finally, the assumption of assessment of regret on an attributeby-attribute basis can be problematic in some contexts. Consider,
for example, the problem of destination choice. Imagine, a foregone
alternative would have been slightly better in terms of a specic
attribute, but it is located substantially farther away than the chosen alternative. Would an individual regret his choice; most likely
not, but the assumption of comparison on an attribute-by-attribute
basis would imply that the distance attribute would not matter in
this example.
The discussion of regret theory thus far has been based on decision-making under conditions of certainty. For decision-making
under conditions of uncertainty, additional assumptions need to
be made with respect to the beliefs of the decision-maker about
the occurrence of a particular value (or discrete classes) of the
attributes for each of the choice alternatives. The most general
assumption would allow for covariances within and between
choice alternatives. Let these beliefs be represented by a multidimensional probability density function:

Hj Hxnjk ; 8n; k

Hxnjk

35

n;k

wk max0; maxfxnjk  xjk g; s 2 C

Rns

87

0
xnk gj

34

Then, expected regret is equal to:

ERsn

36

Individuals are then assumed to choose the risk alternative with the
minimum regret. Developments in discrete choice models can be
used to derive a probabilistic choice model. For example, realizing
that regret minimization is equivalent to maximizing minus regret,
a mixed logit model would imply
n

Prs 2 C

Z
g1 ;d1

n0 n k1

This improved model was motivated by some practical problems

in the estimation of the model. The max operators imply a nonsmooth likelihood function, creating difculties in the derivation
of marginal effects and elasticities. Moreover, it requires dedicated
software to estimate the model. However, this specication creates
other problems. First, the specication is theoretically no longer
consistent with the concept of regret as the value of the regret function is positive also if the chosen alternative turns out to be better
than all foregone choice alternatives. It can also be easily shown
that Eq. (51) does not approximate zero if the attribute differences
are small and/or the number of competing options is large. Second,
if regret would be based on better performance of foregone alternatives, the new specication is the same as the old one, except that a
constant is added. The potential problem here maybe that the addition of an arbitrary constant has more effect if attribute differences
are small, and negligible effects if these differences are high. Empirically, this function was found to perform better than the linear
function. One might argue, however, that regret is small for small
differences, then rapidly increases with increasing attribute differences between the chosen and foregone choice alternative, to tailor

Rsn Hjdj

!
expRi;sn gi
P
f gi ; di dgi ; di
0
0
sn 2C expRi;sn gi

37

where di is an individual-specic error term, and gi captures taste

variation.
Very recently, Chorus et al. (2013) suggested a hybrid model
specication in which some attributes are assumed processed in
terms of regret, while others in the classic random utility fashion.
Although the model was applied to decision making under certainty, it can be extended in a straightforward manner to decision
making under uncertainty. Hess et al. (2012) investigated a mixture model, allowing some individuals choice behavior to reect
regret minimization, and others utility maximization, while also
allowing for taste heterogeneity.
Examples
Chorus et al. (2008a) compared the performance of regrettheoretical models against utility-maximizing models in the
context of non-risky choice of transport modes and acquisition of
travel information. Using fractional factorial experimental a stated
choice experiments, 252 respondents expressed their choice
between 31 pairs of car and train, varying trip purposes (business,

88

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

commute, social, leisure), travel times, costs, waiting times and

seat availability. Estimation results indicated that regret-based
model outperformed the random utility model. Respondents
weighted differences in performance between the chosen and
non-chosen alternative more heavily and tended to discount larger
differences.
Ramos et al. (2011) used a travel simulator to investigate route
preferences. They asked respondents to make 40 consecutive
choices between three routes in the context of a meeting with
co-workers or job interview during the morning commute (departure at 8:00 a.m.; arrival within 1 h). The routes were described as
follows: (i) route 1 consisted mainly of highways, and is the fastest
route, (ii) route consisted mainly of rural roads, and is most reliable, while route 3 consisted of a mix of highway and urban roads,
going through the city centre, and has intermediate performance.
Moreover, information provision was varied according to three
conditions: (i) no information provision, (ii) provision of travel
time in minutes and (iii) provision of queue length in kilometers.
At the end of each simulated trip, true travel times were provided.
Using this data, the authors compared the performance of expected
utility theory, prospect theory and regret theory. They concluded
that the prediction ability of these competing theories varied by
type of information provided. Expected utility theory and regret
theory perform slightly better than cumulative prospect theory if
no information is provided. In contrast, if information is provided,
cumulative prospect theory clearly outperformed the other theories. Overall, there was not much difference.
Beck et al. (2013) compared utility maximizing and regret minimizing models for vehicle choice and found regret minimization
to be the preferred behavioral mechanism for groups and individuals within groups who show a high degree of responsibility for
the choice of the group. In contrast, there is no difference between
regret minimization and utility maximization specications for
individuals and group respondents with a lesser role in the ultimate group decision.

Reection
This review paper has discussed a selection of theories, models
and results of choice and decision-making under conditions of
uncertainty. In particular, expected utility theory, (cumulative)
prospect theory and regret theory and examples of their application to mainly route and departure choice decisions were discussed. While some elaborations of expected utility theory,
prospect theory and regret theory have in common the goal of
offering a model that allows the explanation of empirical anomalies of choices under conditions of uncertainty implied by the classic normative expected utility theory, they emphasize different
mechanisms and consequently also differ in terms of their mathematical specication. For example, while both prospect theory and
regret theory use a reference point, in prospect theory this point is
exogenous and captures the distinction between gains and losses,
whereas the reference point in regret theory is exogenously dened in terms of the best foregone choice alternative. In addition,
the models differ in terms of the valuing functions and decisions
weight being used.
Each of these theories was originally formulated in disciplines
other than transportation research, and applied to domains other
than travel behavior analysis. Original theories were formulated
with gambling in mind. This raises the question about the validity
and usefulness of these theories and model in travel behavior analysis. Can aspects of travel behavior be viewed as being congruent
to gambling?
To provide input to a discussion of this question, it should be
realized that all three discussed theories, as they have been applied

in travel behavior research, relate to one-shot decisions and clearly

formulated payoffs. In contrast, route and departure time choices
are never made in isolation. Rather, these decisions are part of daily activitytravel schedules, which often are based on contextdependent, routine-like, habitual behavior. It is not necessarily
clear that the problem is viewed in terms of gains and losses, as
prospect theory assumes. Uncertainty is viewed as just a single aspect of a more complex scheduling problem. Moreover, to the extent uncertainty is important, decision-makers will face multiple
sources of uncertainty. A path to a destination will normally consist of multiple links and several of those may show high variability
in travel times, while travelers have the exibility of choosing a
path by choosing alternative links between some key nodes in
the network. Uncertainty is not only related to travel times on
routes, but for many activities also on social norms, congestion at
the destination, uncertainty in service levels at the destination,
etc. Furthermore, many sources of uncertainty may be interdependent. Decisions about departure under uncertainty will inuence
the uncertainty of travel times, and subsequent uncertainty of arrival times and activity duration may impact uncertainty in start
times and duration of other activities later during the day.
Another major issue that we wish to identify is that decisionmaking in one shot situations may be inherently different from repeated choice. In one-shot decision situations, decision-makers do
not have any built-up experience about outcome probabilities. At
best their choices are made on general contentions. Moreover, they
cannot alleviate any negative impact of their decisions. In contrast,
repeated choices, such as activitytravel decisions, can in principle
be based on accumulated experience. Travelers learn about the
outcome of their decisions and, based on these experienced outcomes, adapt their behavior in reinforcing (the subjective probability of) positive outcomes and avoiding (the subjective possibility
of) negative or less positive outcomes.
Research agenda
Thus, we argue that most applications of theories and models of
decision-making under uncertainty in travel behavior research
have dealt with toy-problems: very simple problems, represented
in often too limited hypothetical choice problems, and modeled
in ways that do not much justice to context-dependent heterogeneity and taste differences among travelers. The value of such studies is that they show light and increase our understanding of choice
behavior in the controlled settings. However, if the ultimate aim is
to develop comprehensive models of activitytravel behavior of
individual and household decision-making under uncertainty, that
can even start to compete with current dynamic models of travel
demand, a tremendous research effort is needed. The good thing
about it is that a research agenda is unfolding easily. Recent attempts of improving econometrics underlying models of choice
and decision-making under conditions uncertainty (e.g., Polak
et al., 2008; Hu et al., 2012) are an important step forward as it allows incorporating probabilistic choices, taste variation, and
incomplete knowledge on the part of the researcher. In this context, Zhang et al. (2013) proposal to use relative utility theory
which inherently deals with multiple reference points and can
incorporate non-linear functions deserves further attention. The
next step should be examining more complex decision problems,
such as transport mode chains, path choice, combinations of different choice facets and ultimately complete daily and weekly activitytravel schedules, multiple sources of uncertainty, sequential
decisions under uncertainty, uncertainty in household decisionmaking and in social networks. Work along these lines in travel
behavior research is still very scarce. Gao et al. (2008, 2010)
suggested a routing policy choice model, based on a cumulative
prospect theoretic model, to predict travelers route adaptation to

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

real-time information and risk attitudes. The approach explicitly

captures travelers strategic route choice adjustments according
to information on realized network conditions in stochastic networks. Sun et al. (2012) considered multiple sources of error and
estimated a decision tree model, allowing for latent classes of risk
attitudes.
We argue this research agenda is important because basic
assumptions underlying the discussed choice theories do not necessarily hold under these more complex, dynamic conditions.
Barron and Erev (2003) found evidence of risk attitude reversals
when feedback is introduced in repeated choice experiments.
Subjects showed a tendency to avoid risks when faced with losses
and accept more risks when faced with gains. In route-choice,
Ben-Elia et al., 2008) obtained similar ndings. They concluded
that travelers mainly exhibit risk-seeking behavior in the short
run. In the longer run, under reinforcement, they show a tendency to avoid risk. They also did not nd any evidence of loss
aversion and the prospect theoretic model specication was not
better in terms of model t compared to an expected utility specication. Before, Erev et al. (2008) also questioned the validity of
prospect theory in explaining decision-making in repetitive situations, and argued that it is better to be described in terms of
diminishing sensitivity to absolute payoffs. Avineri and Prashker
(2005) found that the higher the variance in travel times is, the
lower travelers sensitivity to travel time differences. This behavior leads to patterns, which differ signicantly from those predictions of cumulative prospect theory.
Learning also plays a case role in regret theory (Ben-Elia et al.,
2012). In order to judge the difference in utility between a chosen
alternative and foregone alternatives, decision-makers need to
experience and learn the probability of outcomes of alternative
choices. It creates however an interesting problem. Traveler will
experience and learn the state of the world for the chosen alternatives, but not for the foregone alternatives, unless they actively collect information and that information is available.
All theories considered in this review paper are based on the
notion that individuals will maximize some individual parameterized (subjective) (reference-dependent)-utility or value function.
However, experienced travel times are the aggregate result of
accumulated individual decisions. It means that the expected travel time at any moment of choice does not only depend on the
decision of an individual, but also on the aggregated results of
the decisions of all other individuals. Thus, in a fundamental sense,
expectations should take the expected decision of all others into
account. Rational behavior of any individual may include strategic
behavior in the sense it is based on the anticipated decision of
other individuals. The situation is however more complex in the
sense that the other individuals may also act strategically, leading
to an iterative loop. Under normal circumstances, the described issue may just be random noise. However, when travel recommendation is provided, its importance may be quite different. Han
et al. (2008) is the only work along these lines in route choice
behavior. They found that individuals were able to anticipate the
strategic behavior of other travelers under uncertainty and adjust
their own choice accordingly. The topic of strategic behavior under
conditions of uncertainty can be more systematically included in
the various approaches discussed in this paper.
In summary, the need to focus research in travel behavior on
choice behavior under uncertainty cannot only be argued from a
theoretical point of view, but is also supported by results of a limited number of empirical studies, which lead to the unavoidable
conclusion that observed dynamic patterns are not congruent with
predictions of commonly used theories. Further development and
the formulation and exploration of alternative, more encompassing
theories and models are required. We hope that this invited review

89

article may be a trigger and inspiration for scholars to address the

identied issues and avenues of future research.
References
Allais, M., 1953. Le Comportement de lHomme Rationnel devant le Risque: Critique
des Postulats et Axiomes de lEcole Americaine. Econometrica, 503546.
Arentze, T.A., Hofman, F., van Mourik, H., Timmermans, H.J.P., Wets, G.P.M., 2000.
Using decision tree induction systems for modeling space-time behavior. Geogr.
Anal. 32, 5272.
Arrow, K.J., 1965. The theory of risk aversion. In: Aspects of the Theory of Risk
Bearing. Yrjo Jahnssonin Saatio, Helsinki. Reprinted in: Essays in the Theory of
Risk Bearing. Markham, Chicago 90109 (1971). ISBN: 0841020019.
Avineri E., 2009. The fuzzy meaning of reference-based perceptions in travel choice
modeling. In: Proceedings 88th Annual Meeting of the Transportation Research
Board, Washington, DC.
Avineri, E., Bovy. P.H.L., 2008. Parameter identication of prospect theory model for
travel choice analysis. In: Proceedings 87th Annual Meeting of the
Transportation Research Board, Washington, DC.
Avineri, E., Prashker, J.N., 2004. violations of expected utility theory in route-choice
stated-preferences: the certainty effect and inating of small probabilities. In:
Proceedings 83rd Annual Meeting of the Transportation Research Board,
Washington, DC.
Avineri, E., Prashker, J.N., 2005. Sensitivity to travel time variability: travellers
learning perspective. Transp. Res. C 13, 157183.
Avineri, E., Prashker, J.N., 2006. The impact of travel time information on travellers
learning under uncertainty. Transportation 33, 393408.
Barron, G., Erev, I., 2003. Small feedback-based decisions and their limited
correspondence to description-based decisions. J. Behav. Decis. Making 16,
215233.
Batty, M., 1976. Urban Modelling. Cambridge University Press, Cambridge.
Beck, M., Chorus, C., Rose, J.M., Hensher, D.A., 2013. Random regret and random
utility in the household purchase of a motor vehicle. In: Proceedings Annual
TRB Meeting, Washington, DC, USA.
Ben-Akiva, M., Lerman, S., 1985. Discrete Choice Analysis. MIT Press, Cambridge,
MA, USA.
Ben-Elia, E., Erev, I., Shiftan, Y., 2008. The combined effect of information and
experience on drivers route-choice behavior. Transportation 35, 165177.
Ben-Elia, E., Ishaq, R., Shiftan, Y., 2012. If only I had taken the other road...: Regret,
risk and reinforced learning in informed route-choice. Transportation, online,
25 pp.
Bhat, C.R., 2005. A multiple discrete-continuous extreme value model: formulation
and application to discretionary time-use decisions. Transp. Res. B 39 (8), 679
707.
Chew, S., MacCrimmon, K., 1979. Alphanu choice theory: A generalization of
expected utility theory. University of British Columbia, Faculty of Commerce
and Business Administration, Working Paper No. 669, Vancouver.
Chorus, C.G., 2011. A new model of random regret minimization. Eur. J. Transp.
Infrastruct. Res. 10, 181196.
Chorus, C.G., Arentze, T.A., Timmermans, H.J.P., 2008a. A random regretminimization model of travel choice. Transp. Res. B 42, 118.
Chorus, C.G., Arentze, T.A., Timmermans, H.J.P., 2008b. A comparison of regretminimization and utility-maximization in the context of travel mode-choices.
In: Proceedings 87th Annual Meeting of the Transportation Research Board,
Washington, DC.
Chorus, C.G., Bierlaire, M., 2013. Travel choice models that generate preferences for
compromise alternatives: an empirical comparison. In: Proceedings 92nd
Annual Meeting of the Transportation Research Board, Washington, DC.
Chorus, C.G., Rose, J., Hensher, D.A., 2013. Hybrid models of random utility
maximization and random regret minimization: results from two empirical
studies. In: Proceedings Annual Meeting TRB, Washington, DC.
Connors, R.D., Sumalee, A., 2009. A network equilibrium model with travellers
perception of stochastic travel times. Transp. Res. B 43 (6), 614624.
David, E., 1982. Regret in decision making under uncertainty. Oper. Res. 30, 961
981.
De Lapparent, M., 2010. Attitude toward risk of time loss in travel activity and air
route choices. J. Intell. Transp. Syst. Technol. Planning Operations 14, 166178.
De Palma, A., Lindsey, R., Picard, N., 2007. Congestion, Risk Aversion and the Value of
Information. Working Paper 2, University of Alberta, Edmonton.
De Palma, A., Lindsey, R., Picard, N., 2012. Risk aversion, the value of information,
and trafc equilibrium. Transp. Sci. 46, 126.
Erev, I., Ert, E., Yechiam, E., 2008. Loss aversion, diminishing sensitivity, and the
effect of experience on repeated decisions. J. Behav. Decis. Making 21 (5), 575
597.
Fishburn, P.C., 1982. Non-transitive measurable utility. J. Math. Psychol. 26 (1), 31
67.
Fujii, S., Kitamura, R., 2004. Drivers mental representation of travel time and
departure time choice in uncertain trafc network conditions. Networks Spatial
Econom. 4, 243256.
Gao, S., Frejinger, E., Ben-Akiva, M., 2008. Adaptive route choice models in
stochastic time-dependent networks. Transp. Res. Rec. 2085, 136143.
Gao, S., Frejinger, E., Ben-Akiva, M., 2010. Adaptive route choices in risky trafc
networks: a prospect theory approach. Transportat. Res. C 18, 727740.
Goldstein, W.M., Einhorn, H.J., 1987. Expression theory and the preference reversal
phenomenon. Psychol. Rev. 94, 236254.

90

S. Rasouli, H. Timmermans / Travel Behaviour and Society 1 (2014) 7990

Han, Q., Dellaert, B., van Raaij, F., Timmermans, H.J.P., 2005. Integrating prospect
theory and stackelberg games to model strategic dyad behavior of information
providers and travelers: theory and numerical simulations. In: Proceedings
83rd Annual Meeting of the Transportation Research Board, Washington, DC.
Han, Q., Timmermans, H.J.P., Dellaert, B., van Raaij, F., 2008. Route choice under
uncertainty: effects of recommendation. In: Proceedings 87th Annual Meeting
of the Transportation Research Board, Washington, DC.
Hensher, D.A., 1981. Applied Discrete Choice Modelling. Wiley, New York.
Hey, J.D., Chris, O., 1994. Investigating generalizations of expected utility theory
using experimental data. Econometrica 62, 12911326.
Hess, S., Stathopoulos, A., Daly, A., 2012. Allowing for heterogeneous decision rules
in discrete choice models: an approach and four case studies. Transportation 39,
565591.
Hu, J., Mehrotra, S., 2012. Robust decision making using a risk-averse utility set.
working paper, Department of Industrial Engineering and Management
Sciences, Northwestern University.
Jou, R.C., Kitamura, R., 2002. Commuter Departure Time Choice: A Reference-Point
Approach. Proceedings EWGT, Bari, Italy.
Kahneman, D., Tversky, A., 1979. Prospect theory: an analysis of decisions under
risk. Econometrica 47, 263291.
Katsikopoulos, K.V., Duse-Anthony, Y., Fisher, D.L., Duffy, S.A., 2000. The framing of
drivers route choices when travel time information is provided under varying
degrees of cognitive load. Hum. Factors 42, 470481.
Katsikopoulos, K.V., Duse-Anthony, Y., Fisher, D.L., Duffy, S.A., 2002. Risk attitude
reversals in drivers route choice when range of travel time information is
provided. Hum. Factors 44, 466473.
Li, Z., Hensher, D.A., 2011. Prospect theoretic contributions in understanding
traveller behaviour: a review and some comments. Transp. Policy 31, 97115.
Liu, X., Polak, J.W., 2007. Nonlinearity and the specication of attitudes towards risk
in discrete choice models. In: Proceedings 86th Annual Meeting of the
Transportation Research Board, Washington, DC.
Lo, H.K., Luo, X.W., Siu, B.W.Y., 2006. Degradable transport network: travel time
budget of travelers with heterogeneous risk aversion. Transp. Res. B 40 (9), 792
806.
Loomes, G., Sugden, R., 1982. Regret theory: an alternative theory of rational choice
under uncertainty. The Econ. J. 92, 805824.
Loomes, G., Sugden, R., 1987. Some implications of a more general form of regret
theory. J. Econ. Theor. 41 (2), 270287.
Luce, R.D., 1959. Individual choice behavior: a theoretical analysis. Wiley, New York.
Machina, M.J., 1987. Choice under uncertainty: problems solved and unsolved. J.
Econ. Perspect. 1, 121154.
Masiero, L., Hensher, D.A., 2011. Shift of reference point and implications on
behavioral reaction to gains and losses. Transportation 38, 249271.
McFadden, D., 1974. Conditional logit analysis of qualitative choice behaviour. In:
Zarembka, P. (Ed.), Frontiers in Econometrics. Academic Press, New York, pp.
105142.
Neumann, V., Morgenstern, O., 1947. Theory of Games and Economic Behavior,
second ed. Princeton University Press, Princeton, NJ.
Palma, A. de., Lindsey, R., Picard, N. 2008. Congestion, risk aversion and the value of
information. In: Proceedings 87th Transportation Research Board Annual
Meeting, Washington, DC.
Polak, J.W., Hess, S., Liu, X., 2008. Characterizing heterogeneity in attitudes to risk in
expected utility models of mode and departure time choice. In: Proceedings
87th Annual Meeting of the Transportation Research Board, Washington, DC.

Prelec, D., 1998. The probability weighting function. Econometrica 66, 497527.
Quiggin, J., 1982. A theory of anticipated utility. J. Econ. Behav. Organ. 3, 323
343.
Ramos, G.M., Daamen, W., Hoogendoorn, S.P., 2011. Expected utility theory,
prospect theory, and regret theory compared for prediction of route choice
behavior. Transport. Res. Rec. 2011 (2230), 1928.
Rasouli, S., Timmermans, H.J.P., 2012a. Uncertainty in travel demand forecasting
models: literature review and research agenda. Transp. Lett. 4, 5573.
Rasouli, S., Timmermans, H.J.P., 2012b. Uncertainty, uncertainty, uncertainty:
revisiting the study of dynamic complex spatial systems. Environ. Planning A
44, 17811784.
Schwanen, T., 2007. The When dimension of coupling constraints. In: Proceedings
86th Annual Meeting of the Transportation Research Board, Washington, DC
Schwanen, T., Ettema, D.F., 2007. Coping with unreliable transportation when
collecting children: examining parents behavior with cumulative prospect
theory. In: Proceedings 86th Annual Meeting of the Transportation Research
Board, Washington, DC.
Senbil, M., Kitamura, R., 2004. Reference points in commuter departure time choice:
a prospect theoretic test of alternative decision frames. Intell. Transp. Syst. 8,
1931.
Senbil, M., Kitamura, R., 2006. Valuing expressways under time pressures. In:
Proceedings 85th Annual Meeting of the Transportation Research Board,
Washington, DC
Stott, H.P., 2006. Cumulative prospect theorys functional menagerie. J Risk
Uncertainty 32, 101130.
Sugden, R., 2004. Alternatives to expected utility. In: Hammond, P., Christian, S.
(Eds.), Handbook of Utility Theory, vol. 2. Kluwer, Dordrecht, pp. 685755.
Sun, Z., Arentze, T.A., Timmermans, H.J.P., 2012. A heterogeneous latent class model
of activity rescheduling, route choice and information acquisition decisions
under multiple uncertain events. Transp. Res. C 25, 4560.
Timmermans, H.J.P., Arentze, T.A., Joh, C.H., 2000. Modeling learning and
evolutionary adaptation processes in activity settings: theory and numerical
simulations. Transp. Res. Rec. 1718, 2733.
Train, K., 2003. Discrete Choice Models with Simulation. Cambridge University
Press, New York.
Tversky, A., Kahneman, D., 1992. Advances in prospect theory: cumulative
representation of uncertainty. J. Risk Uncertain. 5, 297323.
Van de Kaa, E.J., 2010. Applicability of an extended prospect theory to travel
behaviour research: a meta-analysis. Transp. Rev. 30, 771804.
Walker, J., Ben-Akiva, M.E., 2002. Generalized random utility model. Math. Soc. Sci.
43, 303343.
Wilson, A.G., 1974. Urban and Regional Models in Geography and Planning. Wiley,
New York.
Xu, H., Lou, Y., Yin, Y., Zhou, J., 2011a. A prospect-based user equilibrium model with
endogenous reference points and its application in congestion pricing. Transp.
Res. Part B 45 (2), 311328.
Xu, H., Zhou, J., Xu, W., 2011b. A decision-making rule for modeling travelers route
choice behavior based on cumulative prospect theory. Transp. Res. Part C 19 (2),
218228.
Zhang, J., Yu, B., Timmermans, H.J.P., 2013. Extending relative utility model with
multiple reference points to incorporate asymmetric, non-linear response
curvature. In: Proceedings 92nd Annual Meeting of the Transportation Research
Board, Washington, DC.