Specifying the affective influences on decision making...
Transcript of Specifying the affective influences on decision making...
Name: Xiaowen Li (4024167)
Module Title: Patterns of Action Dissertation (C83PAD)
Module Tutor: Professor David Clarke
Specifying the affective influences on decision making:
Rationale, efficacy and future directions
1
Abstract The study of human decision making (DM) behaviour has benefited from over 3
centuries of academic discourse, and the majority of research up to the 21st century
has almost solely been informed by economic theory. An alternative emotional
account has since been provided by advocates of the emotional perspective, and
objective evidence suggests that emotions have at least some influence on human DM
behaviour. The advancement of neuroimaging technology further allows the
investigation of neural processes involved in DM behaviour, and it has been shown
that certain neural loci involved in the mediation of emotions are similarly activated
in the performance of standard DM paradigms. While the neuroimaging research has
so far merely hinted at possible emotional influences in DM behaviour, the present
essay provides a framework to evaluate the specific operation of distinct emotions in
the regulation of human DM behaviour.
Introduction
Following from the largely economic nature of DM research to date, the present essay
shall undertake a brief review of the evolution of economic DM models. It is argued
in this review, that the systematic failure of Expected Utility (EU) theory (Von
Neumann & Morgenstern, 1944) to account for the empirical data stems from its
erroneous adoption of neoclassical assumptions. Contemporary models such as
Cumulative Prospect Theory (CPT – Tversky & Kahneman, 1992) on the other hand,
provide incrementally accurate descriptive accounts of human DM behaviour,
although an adequate explanation for the basis of such models remains at large.
Attempts at explaining CPT have taken on both cognitive (Stewart et al., 2006) and
affective (Hsee & Rottenstreich, 2001) perspectives, and the subsequent sections in
the essay shall attempt to forward an affective account of DM, on the basis of
theoretical and neuroanatomical research.
From an affective perspective, research has begun to suggest that the DM comprises
of more than mere cognitive computations of choice variables, and is to a large extent
influenced by emotional impulses as well. This alternative stance is championed by
the likes of regret theory (Loomes & Sugden, 1982) and the risk as feelings
2
hypothesis (Lowenstein et al., 2001), which recognise the importance of anticipated
and anticipatory emotions respectively. Proceeding from neuropsychological research,
the Somatic Marker Hypothesis (SMH, Damasio, 1994) implicates a number of neural
loci concurrently involved in DM and in emotional regulation, where specific neural
lesions which disrupt emotional control, are shown to result in deficient DM as well.
This neurological approach is similarly adopted by the present study, by reviewing the
functional imaging data implicating various neural loci concurrently involved in the
regulation of emotions and in the performance of DM tasks.
Technological advancement in neural imaging has enabled incrementally precise
functional imaging of neural activity during laboratory based DM paradigms – a.k.a.
neuroeconomics. Coupled with knowledge on the unique functions of specific neural
loci – from research in the psychological sciences, it becomes possible to infer the
underlying processes involved in human DM other than the known cognitive
computation of probabilities and outcomes. Indeed, preliminary studies have
demonstrated emotional centres in the brain (e.g. amygdale; orbitofrontal cortex –
Bechara & Damasio, 2005; insula – Sanfey et al., 2003) to be “active” during the
performance of DM tasks, thus emphasising the appeal of an emotional account of
human DM behaviour. The present essay addresses the mounting recognition for a
more specific emotional account of DM, by proposing a novel framework for
evaluating the specific influence of distinct emotions in regulating DM behaviour.
The present essay reviews the prior neuroimaging research in DM, in the hope of
facilitating an integration of economic (cognitive) and psychological (affective)
perspectives on human DM behaviour. The argument for an emotionally based model
of human DM behaviour is presented in 3 sections. The first of which provides a brief
description of the evolution of economic DM models, together with the argument that
these theories adopt an overly simplistic view of human nature. The second section
highlights the empirical and theoretical basis for incorporating emotions into models
of human DM behaviour, following largely from the SMH. The final section reviews
current neuroscientific trends in DM research, and proposes a novel approach to
evaluating the functional neuroanatomical data already available in the literature.
3
Evolution of Economic Decision Making Models
As part of the argument for living a Christian life, Blaise Pascal (1670, in
Cruickshank, 1983, sec. 233) famously articulated what was perhaps the first formal
statement of choice under uncertainty (Hacking, 1975). In the classic paradigm,
Pascal argued that one ought to consider living a Christian life, based on the
probability of God’s existence, together with the perceived utility of God’s existence
(see Table. 1).
State of the world
God exists God does not exist
Live a Christian life Salvation (∞ utility) Small inconvenience Choice
Live otherwise Damnation (-∞ utility) Normal Life
Table 1: Pascal’s Wager (adapted from Baron, 2000, p. 228)
A number of inadequacies relating to the Wager have since been identified, such as
the assumption of infinite utility from salvation (McClennen, 1994), the arbitrary
ascription of probability to God’s existence (Oppy, 1990) and moral objections to the
wager itself (Clifford, 1986). However, the critical contribution of the Wager to
decision theory was not so much in its philosophical or theistic value. Rather, the
computation of outcome probabilities and utilities in order to arrive at a rational
choice, as advocated by Pascal set the precedent for subsequent theories of decision
making (DM) under risk and uncertainty.
Such was the nature of the Expected Value approach, which defined rational choice
preference to be determined by comparing the EV of 2 (or more) alternatives – where
EV = probability x value of the potential outcome. Following from the largely
economic nature of subsequent DM research, choice alternatives were usually given
by probabilistically defined monetary gambles. Unlike the dichotomous utility of
God’s existence however, the utility of money is perhaps better understood as a
continuous function of its absolute value. Crucially, this necessitated a critical
assumption of a linear utility-wealth relationship in order to remain tenable as a
normative theory of DM.
4
This assumption was no more clearly violated than in the St Petersburg Paradox
(Bernoulli, 1738/1954), which defined a particular gamble where repeated instances
of a coin toss landing on tails is rewarded with the monetary outcome being doubled –
resulting in infinite EV. Yet it is perhaps obvious that a “reasonable man would sell
his chance, with great pleasure, for twenty ducats” (p.31). Bernoulli suggested that the
irregularity in the valuation of the monetary gamble may be due to the diminishing
marginal utility of absolute monetary gains, as defined by a logarithmic function
relative to total wealth (see Figure 1).
Figure 1: Utility of money as a function of total wealth (adapted from Bernoulli, 1738/1954)
While Bernoulli (1738/1954) distinguished between value maximising and utility
maximising strategies, the prescribed utility function he proposed (Figure 1) was
arbitrarily defined, and lacked an adequate empirical or theoretical basis to qualify as
a normative theory of DM. This left the door open for von Neumann & Morgenstern
(1944) to axiomatically define expected utility (EU) maximisation as a normative
model of human DM behaviour. By doing so, the Neumann-Morgenstern expected
utility (NM) theory made specific assumptions as to the nature of rational human DM
behaviour – an elaborate explication of the axioms is beyond the scope of this essay
and the interested reader is referred to the following authors (Baumol, 1972, p.545-
551; Schoemaker, 1982, p. 531-532) – allowing for empirical verification of the
theory. Suffice to say, the formally defined axioms created a test bed for subsequent
research to assess the descriptive validity of NM theory.
Since then, a large body of evidence has shown evidence of systematic and robust
violations to the independence (Allais, 1954; Ellsberg, 1961) and transitivity
(Lichtenstein & Slovic, 1971, 1973; Tversky, 1969) axioms as stated by NM theory.
Util
ity
Total Wealth
5
This led to the development of several modified theories of DM, which were still very
much grounded in the normative assumption of utility maximisation. Among these,
CPT (Tversky & Kahneman, 1992) is perhaps the most widely accepted account
(Plous, 1993), not least because of its ability to account for almost all the available
data (Baron, 2000). Fundamental to CPT, its inventors defined modified probability
weighting and ‘value’ (utility) functions, so as to accommodate for the previous
empirical violations to EU theory.
Figure 2: ‘value’ utility function as defined by CPT (from Fennema & Wakker, 1997)
An elaborate description of CPT is again beyond the scope of this essay, and the
reader is referred to the following sources (Chapter 2 – Kahneman & Tversky, 2000;
Fennema & Wakker, 1997) for more detailed accounts. Briefly however, the
concavity of the ‘value’ function for gains and its convexity for losses accounts for
the empirical evidence which suggests that people are risk averse in the domain of
gains, and risk seeking in the domain of losses (Kahneman & Tversky, 1979). Also,
the steeper ‘value’ function in the domain of losses suggests that “losses loom larger
than gains” (Kahneman & Tversky, 1979, p. 279). Finally, the position of the
reference point implied that utility is a function of changes to one’s current state of
wealth, rather than its prior conception as a function of total wealth (Bernoulli
(1738/1954). In short, the formal definition of the ‘value’ function was able to
account for the vast majority of the empirical evidence, and the conceptualisation of
the ‘value’ function predicted, for instance, the operation of the framing effect
(Tversky & Kahneman, 1981; Kuhberger, 1998) among others.
6
While CPT remains an excellent descriptive account of DM behaviour, it falls short in
so far as it is unable to provide an explanation as to why monetary DM should deviate
from NM theory – an explanation which has since been proposed by proponents of a
cognitive account. A study which modelled the debit (loss) and credit (gain)
frequency data in a vast sample of UK current bank accounts, derived a similar
function to that described in the ‘value’ function depicted in figure 2 (Stewart et al.,
2006). The ‘decision by sampling’ account of the research findings attributed the
human conformity to the ‘value’ function described by CPT, on the basis of its
cognitive familiarity with that in our environments. On the other hand, an affective
explanation has similarly been provided to account for the unique S-Shape of the
probability weighting function defined by CPT. Hsee & Rottenstreich (2001) propose
that in addition to the psychophysical methods employed by Tversky & Kahneman
(1992), the unique S-shape of the probability weighting function in CPT can be
attributed to affective influences as well. Specifically, they demonstrated that when
the prospect is affect-rich, then the S-shape of the associated weighting function is
considerably more pronounced than when the prospect is affect poor (see figure 3)
Figure 3: distinct probability weighting functions for affect-rich and affect-poor prospects – hypothetical (from Hsee & Rottenstreich, 2001)
Although there is ambiguity surrounding the precise cognitive and affective
explanations for CPT, it remains the most accurate account of human DM to date. The
same cannot be said for NM theory, where the empirical data seems to “suggests that
at the individual level EU maximization is more the exception than the rule”
(Schoemaker, 1982, p. 552). The failure of NM theory to account for the empirical
data can be understood by considering the neoclassical assumptions upon which it is
7
based. Firstly, according to the neoclassical view of the ‘economic man’(after Mill,
1848 – see Persky, 1995), the concept of rationality necessitates that the decision
maker have perfect knowledge about the relative probabilities and outcomes of choice
alternatives. This seems an unrealistic assumption, given that real world choices are
rarely made with the knowledge of all other possible alternatives. Perhaps more
plausibly, the notion of bounded rationality (Simon, 1957) suggests that “agents
experience limits in formulating and solving complex problems and in processing
information” (Williamson, 1981, p. 553).
Secondly, the neoclassical view assumes that the choice preference of rational
individuals is such as to maximise utility (Weintraub, 2007). ‘Rationality’ can broadly
be defined as “reasoning in a way which helps one to achieve one’s goals” (Evans et
al., 1993, p. 168), with utility being a measure of the extent to which these goals are
achieved. In the economic DM models however, utility has arbitrarily been defined as
some function of the monetary value. While it is acknowledged that money may be an
important facet of human desires, it is by no means the singular purpose of human
existence. It is therefore worth considering other aspects of human desire (e.g.
Maslow, 1943) in order to arrive at a more proximal measure of utility. From an
emotional perspective, regret theory (Loomes & Sugden, 1982) suggests that
decisions are made by computing the potential for anticipated regret/rejoice in making
a particular choice. According to regret theory therefore, to maximise utility in a
particular gamble, is to maximise the potential for rejoice and to minimise the
potential for regret.
Rationale for Incorporating Emotion into Decision Making Models
From regret theory, the role of emotions seemed to gain prominence as a means of
understanding human DM behaviour. This section of the essay shall review some of
the theoretical and empirical bases for suggesting that emotions play an integral part
in human DM. Beginning with regret theory (Loomes & Sugden, 1982), the radical
view taken by the SMH (Damasio, 1994) and more recent developments in the
literature have since suggested specific means of incorporating emotion in the
modelling of human DM behaviour (e.g. Lerner & Keltner, 2000). The evidence
supporting this emotional perspective will focus on the anomalies in investment (DM)
8
behaviour, where affective states have clearly been shown to mediate DM processes.
An additional rationale for reviewing investment behaviour is grounded in the
assumption that investors are thought to closely resemble the idealised neoclassical
standard of the ‘economic man’. Accordingly, clear violations of utility maximising
strategies would suggest that the neoclassical view of rationality cannot be upheld,
even where it would clearly be advantageous to do so.
In regret theory, the anticipated emotions of regret or rejoice are assumed to govern
DM, on the basis of various counterfactual comparisons. Other theories which have
adopted a similar view of the potency of anticipated emotions include decision affect
theory (Mellers et al., 1999), where choice preference is purportedly based on
maximising the potential for anticipated pleasure. Isen et al. (1988) similarly proposed
a mood maintenance hypothesis, which provides an interesting account of the concave
value function in CPT (see figure 2). According to Isen et al., people in good moods
are keen to preserve their current positive affective state, and are thus risk averse in
the domain of gains. Recently however, the validity of regret theory, and theories
incorporating anticipated emotion in general have been challenged on theoretical (e.g.
the assumption that regret and rejoice are equal and opposite – Humphrey, 2004) and
empirical (e.g. event splitting effects – Starmer & Sugden, 1993) grounds.
Furthermore, anticipated emotions are unable to provide a comprehensive explanation
beyond preference reversals (which regret theory was originally conceived to do), and
are thus insufficient as complete models of DM.
In a review of the literature concerning the affective influences on DM, Lowenstein &
Lerner (2003) distinguish between expected emotions and immediate emotions. They
further argue that theories incorporating merely expected emotions – like regret
theory – provide inadequate accounts of the affective influence on DM, likewise with
theories incorporating merely immediate emotions (e.g. Lowenstein et al 2001). In
their paper, Lowenstein & Lerner (2003) propose a hypothetical framework (see
figure 4) as to how expected and immediate emotions might interact to influence DM
behaviour, suggesting that a comprehensive model of DM need necessarily account
for both these sources of emotional influence.
9
Figure 4: Hypothetical framework describing the interaction between the determinants and consequences of immediate and expected emotions (from Lowenstein & Lerner, 2003)
The SMH (Damasio, 1994) proposed that human reasoning and DM are contingent on
multiple levels of neural operation, some of which are consciously overt, while others
– like emotions – are beneath our conscious awareness. Importantly, the SMH does
not distinguish between expected and immediate emotions, but is rather based on a
more holistic view of emotions. The SMH postulates that somatic markers – derived
from secondary emotions and regulated by the ventromedial prefrontal cortex
(vmPFC) – restrict the decision to a particular subset of choices, among all other
possible alternatives. The SMH was formulated on the basis of behavioural
observation of patients with specific lesions to the vmPFC, who demonstrated marked
reductions in visceral somatic responses. Subsequent empirical tests of the SMH took
the form of the Iowa Gambling Task (IGT, Bechara et al., 1994), which revealed a
relative impairment of vmPFC lesion patients in the task. It was deduced that an
impairment in the representation of emotional states (through somatic markers)
resulted in the failure to associate aversive (disadvantageous) card selections with
negative emotional states, thus resulting in a poorer IGT performance.
The SMH and the subsequent tests of its hypotheses provided a neuropsychological
basis for the role of emotions in DM. Other evidence alluding to the emotional
influence on DM arises from investment behaviour, where market research has
indicated a distinct effect of emotions on investment performance. For example,
Seasonal Affective Disorder (SAD) – brought about by prolonged hours of darkness
during winter – is a phenomenon known to affect a significant proportion of the
10
population, with effects similar to that of depression (Rosenthal et al., 1984). In a
study of stock markets in both the northern and southern hemispheres, Kamstra et al.
(2003, p. 324) demonstrate “a SAD effect in the seasonal cycle of stock returns…
even after controlling for well known market seasonals and other environmental
factors”. This effect is even more remarkable, given that the SAD effect is 6 months
out of phase in the northern and southern hemispheres, in line with the onset of
winter. More generally, the amount of morning sunshine has also shown to be
strongly correlated with the stock returns of that day (Hirshleifer & Shumway, 2003).
One shortcoming of the SMH is that it failed to distinguish between the expected and
immediate emotional influences on DM. Following from the distinction made by
Lowenstein & Lerner (2003), it is argued that a complete model of DM must
incorporate such a level of specificity. Furthermore, the SMH and the hypothetical
framework put forth by Lowenstein & Lerner, fail to provide an account for the
specific influence of distinct emotions on DM, such as anger and fear (Lerner &
Keltner, 2000). The final section of this essay addresses this inadequacy in the
literature reviewed thus far, by proposing a neuroscientific approach to investigating
the specific influence of distinct emotions in DM.
Neuroscience as a Means of Investigating the Role of Specific Emotional Influences on Decision Making
The advancement of neuroimaging techniques in recent years has enabled
psychologists to investigate with greater precision, the role of specific neural loci in
the experience of distinct emotions (see Adolphs, 2002). Similarly, decision scientists
have increasingly employed neuroimaging techniques to ascertain the neural loci
involved in performing DM tasks – a.k.a. neuroeconomics. It is thus somewhat
surprising that advocates of the emotional perspective have been slow to capitalise on
this established method, as a means of investigating the role of specific emotional
influences on DM. This section of the essay shall perform a cross comparison of
selected research in both the fields of affective neuroscience and neuroeconomics, in
order to highlight the potential for future advancement of the emotional perspective.
11
The beginnings of the neuroscientific approach can arguably be traced to the SMH,
where the vmPFC was first identified as being a critical neural structure for the
regulation of emotional states. Specifically, the vmPFC had been identified to help
“predict the emotion of the future, thereby forecasting the consequences of one’s own
actions” (Bechara & Damasio, 2005, p. 354). The amygdale was subsequently
recognised to play a crucial role in the emotional processing of choice alternatives, in
part due to its multitudinous efferent connections to the vmPFC (Bechara et al. 1999).
To put it simply, a lesion to the amygdale inhibits the subject from feeling the
unpleasant sensation of losing money, limiting the function of the vmPFC in so far as
it does not receive a signal from the amygdale. Conversely, a recent study by Hsu et
al., (2005) demonstrated that lesions to the orbitofrontal cortex (OFC) – an area
encompassing the vmPFC – led to a deficient incorporation of ambiguity (fear) into
DM. As is already well established in the literature, the amygdale is associated with
the experience of the fear emotion (Adolphs, 2002), therefore suggesting that fear
modulates DM behaviour.
Indeed, a study by De Martino et al. (2006) employed a reliable framing paradigm
originally devised by Tversky & Kahneman (1981), with subjects in a functional
magnetic resonance imaging (fMRI) scanner. Briefly, the framing effect accounts for
the concave (risk seeking) value function in the domain of losses, and the convex (risk
averse) value function in the domain of gains (see figure 2), by merely manipulating
the linguistic phrasing of 2 formally identical prospects. The fMRI imaging results
obtained revealed that the framing effect obtained was specifically associated with
activation of the amygdale. Furthermore, activity in the OFC reduced the
susceptibility to the framing effect, likely due to its modulating role between the raw
fear emotion – from the amygdale – and its influence on selecting choice preference.
In a previously mentioned study, Lerner & Keltner, (2000) argued that the fear
emotion would result in an increased perception of risk. With respect to the fMRI
study described earlier (De Martino et al., 2006), an interpretation of the framing
effect – risk aversion in the domain of gains – can thus be explained in terms of the
subject’s “fear” of losing their current endowment. More recently however, Tom et al.
(2007) demonstrate evidence which suggests that the amygdale activation is absent in
loss aversion. Regardless of the eventual outcome, the current discourse in the subject
12
of emotional involvement in DM is the first step to an integration of both the
economic and affective sciences.
Having identified the amygdale as being the neural locus for the fear emotion,
subsequent neuroimaging research in DM (De Martino et al., 2006), targeted
functional analysis to this region during the performance of DM tasks. In a similar
way, the neural loci for other emotions should be targeted as the focus of future
neuroeconomic research, so as to ascertain their specific influence (and the influence
of their associated emotions) on standard DM paradigms. Proceeding with this
guideline for future research, the already established influence of emotions on DM
can be defined with greater specificity.
One such neural locus which has since been targeted is the insula, established as being
the neural locus for the emotion of disgust (Calder et al., 2000; Adolphs et al., 2003).
In an fMRI study of the ultimatum game, rejected (perceived unfair) offers were
found to be associated with significantly heightened activity in the insula (Sanfey et
al. 2003). In the ultimatum game, 2 players are assigned to different roles – the
proposer and the responder. The proposer is credited with a particular sum of money,
and makes an offer as to how much of that sum he/she decides to share with the
responder. The responder can then either choose to accept or reject the offer. If the
offer is accepted, then the money is split as defined by the offer. If however, the
responder rejects the offer, then neither player receives any money.
Contrary to the neo-classical expectations of rationality, responders typically reject
around 50% of the offers below 20% of the original sum (Camerer, 2003). The neo-
classical assumption of rational action predicts an unconditional acceptance of the
proposition by the responder, because whatever the sum of money offered, it is
infinitely better than receiving none. In an attempt to account for this apparent
deviation in rational behaviour, economic theories have modeled fairness as being a
key motivator of behaviour (Fehr & Schmidt, 1999). Alternatively, the affective
consequence of inequity (fairness) has been proposed to account for the apparently
non-rational rejection of propositions (van Winden, 2007). Taking the imaging data
from before (Sanfey et al., 2003), heightened insula activation serves to suggest that
the reason for the responder’s rejection is due to his/her disgust at the proposition.
13
Once again, robust deviations from apparent rational behaviour can be explained in
terms of the affective influence on DM behaviour.
By selectively focusing future research on known emotional neural loci, the full
gamut of deviations from the NM model, in addition to other theories of rational DM
may gradually be accounted for by affective explanations. Indeed, limited research
has already implicated a number of emotional centres in the brain to be involved in
DM tasks – e.g. anterior cingulate cortex (Kennerley et al., 2006); ventral striatum
(Bhatt & Camerer, 2005). However, the emerging research is slow to draw the link
between the activation of these neural regions in DM tasks, and their known function
in the regulation of emotions.
Conclusion
The essay thus far has provided a concise rationale for incorporating emotions in the
investigation of DM behaviour. Stemming from the inadequacies of prior economic
theory (e.g. NM theory), which assume a specific form of rationality,
neuropsychological evidence (SMH) has since alluded to the importance of
incorporating emotions into the analysis of DM. Early theories attempting to
incorporate emotions in the modelling of DM behaviour have enjoyed limited
success, with the large majority either failing to incorporate certain aspects of emotion
(e.g. regret theory) or failing to provide a specific enough account of the various
emotions (e.g. SMH). The proliferation of neuroimaging research gives fresh impetus
to furthering an affective account of DM, where an integration of research in both
neuroeconomics and the psychological sciences would serve as its ammunition. As it
is, there is substantial evidence to suggest that the emotions of fear (amygdale) and
disgust (insula) are responsible for DM deviations from traditional economic theory.
Other neural loci with established emotional function should therefore be
incorporated in future functional analyses of DM, so as to elucidate our understanding
of the emotional contribution to DM.
4199 words
14
References Adolphs, R. (2002). Neural Systems for Recognising Emotion. Current Opinion in
Neurobiology, 12(2), 169-177. Adolphs, R., Tranel, D. & Damasio, A.R. (2003). Dissociable Neural Systems for
Recognizing Emotions. Brain Cognition, 52(1), 61-69. Allais, M. (1953). Le Comportement de l'Homme Rationnel Devant le Risque:
Critique des Postulats et Axiomes de l'ecole Americane. Econometrica, 21(4), 503-546.
Baron, J. (2000). Thinking and Deciding (Third Edition). Cambridge: Cambridge
University Press. Baumol, W.J. (1972). Economic Theory and Operations Analysis. Englewood Cliffs,
N.J.: Prentice-Hall. Bechara, A., Damasio, A.R., Damasio, H. & Anderson, S.W. (1994). Insensitivity to
Future Consequences Following Damage to Human Prefrontal Cortex. Cognition, 50, 7-15.
Bechara, A., Damasio, H., Damasio, A.R. & Lee, G.P. (1999). Different
Contributions of the Amygdala and Ventromedial Prefrontal Cortex to Decision Making. Journal of Neuroscience, 19, 5473-5481.
Bechara, A. & Damasio, A.R. (2005). The Somatic Marker Hypothesis: A Neural
Theory of Economic Decision. Games and Economic Behavior, 52, 336-372. Bernoulli, D. (1738). Specimen Theoriae Novae de Mensura Sortis. Commentarri
Academiae Sdentiarum Imperialis Petropolitanae, V, 175-92. Translated by Sommer, L. (1954) as "Expositions of a New Theory on The Measurement of Risk". Econometrica, 22, 23-26.
Bhatt, M. & Camerer, C. (2005). Self-Referential Thinking and Equilibrium as States
of Mind in Games: fMRI Evidence. Games and Economic Behavior, 52(2), 424-459.
Calder, A.J., Keane, J., Manes, F., Antoun, N. & Young, A.W. (2000). Impaired
Recognition and Experience of Disgust Following Brain Injury. Nature Neuroscience, 3, 1077-1078.
Camerer, C. (2003). Behavioral Game Theory, Princeton: Princeton University Press. Clifford, W.K. (1986). The Ethics of Belief. In G.D. McCarthy (ed.). The Ethics of
Belief Debate, Atlanta, Georgia: Scholars Press. Cruickshank, J. (1983). Pascal, Pensées / John Cruickshank. London: Grant & Cutler.
15
Damasio, A.R. (1994). Descartes’ Error. New York: Putnam. De Martino, B., Kumaran, D., Seymour, B. & Dolan, R.J. (2006). Frames, Biases, and
Rational Decision-Making in the Human Brain. Science, 313, 684-687. Ellsberg, D. (1961). Risk, Ambiguity and the Savage Axioms. Quarterly Journal of
Economics, 75, 643-669. Evans, J.S., Over, D.E. & Manktelow, K.I. (1993). Reasoning, Decision Making and
Rationality. Cognition, 49(1-2), 165-187. Fehr, E., & Schmidt, K. (1999). A Theory of Fairness, Competition and Cooperation.
Quarterly Journal of Economics, 114, 817–868. Fennema, H. & Wakker, P. (1997). Original and Cumulative Prospect Theory: A
Discussion of Empirical Differences. Journal of Behavioral Decision Making, 10, 53-64.
Hacking, I. (1975). The Emergence of Probability. Cambridge: Cambridge University
Press. Hirshleifer, D. & Shumway, T. (2003). Good Day Sunshine: Stock Returns and the
Weather. The Journal of Finance, 58(3), 1009-1032. Hsee, Y. & Rottenstreich, C.K. (2001). Money, Kisses and Electric Shocks: On the
Affective Psychology of Risk. Psychological Science, 12(3), 185-190.
Hsu, M., Bhatt, M., Adolphs, R., Tranel, D. & Camerer, C.F. (2005). Neural Systems Responding to Degrees of Uncertainty in Human Decision-Making. Science, 310, 1680-1683.
Humphrey, S.J. (2004) Feedback-Conditional Regret Theory and Testing Regret-Aversion in Risky Choice. Journal of Economic Psychology. 25(6), 839-857.
Isen, A.M., Nygren, T.E., & Ashby, F.G. (1988). Influence of Positive Affect on the
Subjective Utility of Gains and Losses: It is Just Not Worth the Risk. Journal of Personality and Social Psychology, 55(5), 710-717.
Kahneman, D. & Tversky, A. (1979). Prospect theory: An Analysis of Decision
Under Risk. Econometrica, 47(2), 263-291. Kahneman, D. & Tversky, A. (2000). Choices, Values, and Frames. Cambridge:
Cambridge University Press. Kamstra, M.J. Kramer L.A. & Levi M.D. (2003). Winter Blues: A SAD Stock Market
Cycle. The American Economic Review, 93(1), 324-343.
16
Kennerley, S.W., Walton, M.E., Behrens, T.E.J., Buckley, M.J. & Rushworth, M.F.S. (2006). Optimal Decision Making and the Anterior Cingulate Cortex. Nature Neuroscience, 9, 940-947.
Kühberger, A. (1998). The Influence of Framing on Risky Decisions: A Meta-
Analysis. Organisational Behaviour and Human Decision Processes, 75(1), 23-55.
Lerner, J.S. & Keltner, D. (2000). Beyond Valence: Toward a Model of Emotion-
Specific Influences on Judgment and Choice. Cognition and Emotion, 14(4), 473-494.
Lichtenstein, S. & Slovic, P. (1971). Reversals of Preference between Bids and
Choices in Gambling Decisions. Journal of Experimental Psychology, 89(1), 46-55.
Lichtenstein, S. & Slovic, P. (1973). Response-Induced Reversals of Preference in
Gambling: An Extended Replication in Las Vegas. Journal of Experimental Psychology. 101(1), 16-20.
Lowenstein, G.F. & Lerner, J.S. (2003). The Role of Affect in Decision Making. In
R.J. Davidson, K.R. Scherer & H.H. Goldsmith (eds.). Handbook of Affective Sciences. Oxford: Oxford University Press.
Loewenstein, G.F., Weber, E.U., Hsee, C.K. & Welch, N. (2001). Risk as Feelings.
Psychological Bulletin, 127(2), 267-286. Loomes, G. & Sugden, R. (1982). Regret Theory: An Alternative Theory of Rational
Choice Under Uncertainty. Economic Journal, 92, 805-824. Maslow, A.H. (1943). A Theory of Human Motivation, Psychological Review, 50,
370-396 McClennen, E. (1994). Finite Decision Theory. In J. Jordan (ed.). Gambling on God:
Essays on Pascal's Wager, London: Rowman & Littlefield Mellers, B., Schwartz, A. & Ritov, I. (1999). Emotion-Based Choice. Journal of
Experimental Psychology: General, 128(3), 332-345. Mill, J.S. (1948). Principles of Political Economy: With Some of Their Applications to
Social Philosophy. London: Parker. Oppy, G. (1990). On Rescher on Pascal s Wager, International Journal for
Philosophy of Religion, 30, 159-168. Persky, J. (1995). Retrospectives: The Ethology of Homo Economicus. The Journal of
Economic Perspectives, 9(2), 221-231. Plous, S. (1993). The Psychology of Judgment and Decision Making. USA: McGraw-
Hill.
17
Rosenthal, N.E., Sack, D.A., Gillin, J.C., Lewy, A.J., Goodwin, F.K., Davenport, Y., Mueller, P.S., Newsome, D.A. & Wehr, T.A. (1984). Seasonal Affective Disorder: a Description of the Syndrome and Preliminary Findings with Light Therapy. Archives of General Psychiatry, 41(1), 72–80.
Sanfey, A.G., Rilling, J.K., Aronson, J.A., Nystrom, L.E. & Cohen, J.D. (2003). The
Neural Basis of Economic Decision-Making in the Ultimatum Game. Science, 300, 1755-1758.
Schoemaker, P.J.H. (1982). The Expected Utility Model: Its Variants, Purposes,
Evidence and Limitations. Journal of Economic Literature, 20, 529-563. Simon, H.A. (1957). Models of Man: Social and Rational. New York: John Wiley and
Sons, Inc. Starmer, C. & Sugden, R. (1993). Testing for Juxtaposition and Event-Splitting
Effects. Journal of Risk and Uncertainty. 6(3), 235-254. Stewart, N., Chater, N. & Brown, G.D.A. (2006). Decision by Sampling. Cognitive
Psychology, 53, 1-26. Tversky, A. (1969). Intransivity of Preferences. Psychological Review, 76, 31-48. Tversky, A. & Kahneman, D. (1981). The Framing of Decisions and the Psychology
of Choice. Science, 211, 453-458. Tversky, A. & Kahneman, D. (1992). Advances in Prospect Theory: Cumulative
Representation of Uncertainty. Journal of Risk and Uncertainty, 5, 297-323. Tom, S.M., Fox, C.R., Trepel, C. & Poldrack, R.A. (2007). The Neural Basis of Loss
Aversion in Decision-Making Under Risk. Science, 315, 515-518. Von Neumann, J. & Morgenstern, O. (1944). Theory of Games and Economic
Behavior. Princeton, NJ: Princeton Univ. Press. Van Winden, F. (2007). Affect and Fairness in Economics. Social Justice Research,
20(1), 35-52. Weintraub, E.R. (2007). Neoclassical Economics. In D.R. Henderson (ed.). The
Concise Encyclopedia of Economics. USA: Library of Economics and Liberty. Williamson, O. (1981). The Economies of Organization: The Transaction Cost
Approach. American Journal of Sociology, 87(2), 548-577.