Neuroscience and Decision Making
From PsychWiki - A Collaborative Psychology Wiki
Neuroscience and Decision Making
Introduction
“Human behavior, in general … is not under the constant and detailed guidance of careful and accurate hedonic calculations, but is the product of an unstable and unrational complex of reflex actions, impulses, instincts, habits, customs, fashion, and hysteria.” – Viner (1925)
For a long time, economists have argued that humans make decisions by obeying the laws of rationality. Expected utility theory has dominated our understanding of decision making by postulating that under the majority of circumstances, consumers make decisions and choices by maximizing the utility of any given outcome (Kahneman & Tversky, 1979). However, in observing our own behaviors and those of others we know that often this is not the case. We do not engage in a mental cost-benefit analysis to determine which choice to make. This holds true for both risky and non risky decisions. Many behavioral experiments have been conducted and different theories have been proposed to better explain our behavior under risky situations. The concept of risk has been associated with hazards that fill one with dread. Risk is different from uncertainty in that, risk has options that are well specified or transparent outcomes with certain probabilities of occurrence attached to them (Trepel et al., 2005); uncertainty incorporates no such thing. Over the last 30 years, behavioral decision making has accumulated a rich literature of decision making under risk; recently, however, neuroscience has started to play an important role in helping us understand the human brain and its role in explaining our behaviors. Even though neural mechanisms do not provide a complete explanation for rational choice and judgment, Camerer, Loewenstein and Prelec (2004) postulated that brain evidence has the potential to suggest a better theory (than the ones that currently exist) to explain our behaviors.
The purpose of this paper is to provide an integrative review of the area of Neuroscience and its relationship to Behavioral Decision Making. I will start by discussing Prospect theory and the role that neuroscience can play in understanding human behavior under risky situations. I will then discuss the Somatic Marker Hypothesis and its application in decision making. Further, I will highlight some techniques that are used to measure neural responses. Finally, I will end with future avenues of research where Neuroscience techniques can be applied in studying different Marketing phenomena.
Prospect Theory & Decision Making
From a purely economic perspective, expected utility theory has postulated that humans are rational beings and when making choices between options, our instinct is to maximize the benefit that we will derive from our choice while minimizing the cost that will be incurred from our decision. A seminal article by Kahenman and Tversky (1979) showed that often this was not the case. They gave examples where consumers made decisions which disobeyed the laws of ‘rational thinking’, specifically when these decisions were those that had to be made under risk and uncertainty. A classic problem that has been discussed to illustrate this point is the Asian Disease Problem which is stated as follows:
“Imagine that the United States is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed…If program A is adopted: 200 people will be saved. If program B is adopted: there is a 1/3 probability that 600 people will be saved and a 2/3 probability that no people will be saved.” (Tversky & Kahneman, 1981).
The authors ran this experiment with different sets of students and the results were astounding; they showed that the vast majority of people preferred program A – where 200 people would saved with certainty as opposed to program B where 600 people would saved with a probability of 1/3 (Note that both situations result in 200 people being saved). This was an interesting result and was described by the authors as a certainty effect where humans preferred a ‘sure thing’ or certain outcome as opposed to an outcome with a probability attached to it – regardless of whether the two outcomes yielded the same results. These authors and several others have documented this anomalous behavior under different experimental settings. Tverksy & Kahneman (1979) synergized these phenomena into Prospect Theory, which accounted for the risks or gambles that people take into account when making decisions. Under Prospect Theory, the value V of a simple prospect that pays $x with probability p is given by: V(x, p) = v(x)*w(p) (1) where v measures the subjective value of consequence x and w measures the impact of probability p on the attractiveness of the prospect.
The value function is shaped like an S-curve which represents a person’s diminishing sensitivity to the outcomes. As one goes away from the reference point (0 in this situation) the impact of a change in the outcome or probability of that outcome occurring goes down. We note that as we move away from the reference point, the value function is concave for gains and convex for losses. In fact, the value function is much steeper for losses than for gains. This means that people are more sensitive to losses than to equivalent amounts of gains, a phenomenon that has been characterized as loss aversion. Tversky and Kahenman (1981) summed this phenomenon by stating that ‘losses loom larger than gains’. Trepel et al. (2005) provided an example that most people would reject a gamble where they would gain $100 if a fair coin lands on heads and lose $100 when the coin lands on tails. They state that typically losses have atleast twice the impact of equivalent gains such that people would require a 50% chance of gaining atleast $200 to make up for a 50% chance of losing $100.
This example and others like these have lead to another avenue of research that has been studied in prospect theory which is the anticipation of rewards and the dread of losses. This notion of anticipation of rewards is tied to the decision utility of prospect theory. It has been shown that people actually respond differently – both neurally and physically (skin conductance response has shown this) when they are anticipating gains and losses. Neuroscience research has shown that different areas in our brain are activated when we anticipate rewards. It has been shown that the dopamine system, ventral striatum, prefrontal cortex and amygdala all play a role in the representation of decision utility (Trepel, Fox and Poldrack, 2005) – each will be discussed in turn.
Dopamine System
Dopamine is a modulatory neurotransmitter that is produced by regions in the midbrain. It is transmitted broadly to a set of cortical and subcortical regions (Cooper, Bloom and Roth, 2002). Trepel, Fox and Poldrack (2005) state that the dopaminergic system appears to be a primary substrate for the representation of decision utility. Specifically, increased firing of dopamine neurons has been documented when people are faced with unexpected rewards and in response to stimuli that predict future rewards.
Ventral Striatum
The VS is the center of integration of the ‘data’ between the prefrontal cortex, amygdala and hippocampus. Recent research has shown that it plays a critical role in the representation of the magnitude of anticipated reward (Knutson and Peterson, 2005). Further, it has been shown that since the VS is a primary target of the dopamine system, changes and activation in the VS system during stimuli processing (as seen under neuro imaging techniques) may allude to activation in the dopamine system.
Prefrontal Cortex
This is a large and heterogeneous brain region with many different functions in decision making. Since the dorsolateral prefrontal cortex (DLPFC) plays an important role in cognitive representations in working memory, it might also play a role in decision making by representing prospects and subsequent decision utility computations. The Iowa Gambling Task (Bechara & Damasio, 2005) showed that when patients had damage to the DLPFC – they tended to make non-optimal decisions which reflected their inability to use strategies and rules in decision making. Further, the ventromedial prefrontal cortex (henceforth, VMPFC) has also been used in understanding decision making, especially loss anticipation. The same gambling task showed that when people had lesions in the VMPFC, they had a hard time anticipating losses and how they would feel if they were to lose something.
Amygdala
The amygdala is heavily involved in emotion and learning. This is true especially for negative outcomes. The amygdala is responsible for producing fear responses in us and for the learning associated between particular stimuli and fear responses. It has been shown that the amygdala plays a key role in the representation of utility from a gain or dis-utility from losses. Once again, this was tested in the Iowa Gambling Task, where patients with damage to the amygdala did not learn that the choice of risky outcomes would result in losses, which would be disadvantageous to the individuals, and hence they chose those outcomes. On the contrary, those with no damage, learned to stay away from risky outcomes that resulted in losses (Bechara & Damasio, 2005). Understanding the function of these four systems in the brain, dopamine system, ventral striatum, prefrontal cortex and the amygdala, helps us gain more insight into the neural circuitry involved when explaining prospect theory and human decision making under risky conditions.
While prospect theory alludes to the fact that humans use emotions in decision making – it does not explicitly use the term ‘emotions’. It attributes our lapse in rational thinking when making decisions to some ‘other’ unexplained factor. It has been seen that emotions play a big role in driving many of our behaviors, including the decisions that we make. Accounting for emotions and affect in decision making has started to gain favor in the field of Marketing and it is an important topic to study as marketing managers often design ads to appeal to our senses and play off of our emotions. As this is an important area of research in marketing, I think that it is apt to explore the Somatic Marker Hypothesis proposed by Antonio Damasio (1994) as it delves into emotions and how they are characterized in our neural circuitry.
The Somatic Marker Hypothesis
The Somatic Marker Hypothesis (henceforth SMH) was proposed by Damasio (1994) after he observed that patients with lesions in their VMPFC, who had a defect in their emotional mechanism, were unable to make advantageous decisions when playing an investment gamble. The Damasio’s established emotions as being an integral part of decision making, which until then had been defined primarily by psychologists as an intense neural mental state, which evokes either a positive or negative physiological response, and arise purely subjectively rather than through conscious effort. The SMH characterizes an emotion as the relationship between both psychological and physiological states of the body. The object or event that predictably causes an emotion is designated as an emotionally-competent stimulus and this evocation of an emotion leads to physical responses in the body. Emotions are manifested both internally (unobservable to other people) and externally (other people can tell that you are experiencing an emotion). Internal physical changes in the body include an increase in the heart rate, endocrine release or smooth muscle contraction, while external, visible, changes include changes in facial expression, posture and particular behaviors.
To avoid the affect laden word of ‘emotion’, the SMH has characterized an emotion as a somatic state, which represents the relationship between the external and internal states of our bodies Bechara and Damasio (2005). There are 3 stages involved in the somatic response: 1. the central nervous system releases certain neurotransmitters (dopamine, seratonine etc.); 2. there is an active modification of the state of somatosensory maps such as those of the insular cortex (‘as-if-body-states) and finally 3. there is a modification of the transmission of signals from the body to the somatosensory regions. These somatic states can be induced from both primary and secondary inducers, where primary inducers are innate or learned stimuli that cause pleasurable and aversive states that automatically elicit a somatic response and secondary inducers are generated by recall of personal or hypothetical events. While the same external stimulus can be used to trigger either state, each has different neural processes that are responsible for triggering the states. The two primary neurological systems involved in the activation of these states are the amygdala (primary inducer) and the VMPFC (secondary state).
The amygdala joins the features of the primary inducers which can be processed subliminally via the thalamus or explicitly via early sensory and high-order association cortices, with the somatic state associated with the inducer. Once the somatic states are induced through the primary inducers, signals are relayed to the brain after which the presentation of a stimulus that evokes thoughts and memories about a specific primary inducer then operates as a secondary inducer. The amygdala triggers somatic states that are fast, short lived and have quick habituation (Buchel et al., 1998; Dolan et al., 1996). It is responsible for the ‘fight or flight’ response that is generated when we find ourselves in situations that require us to respond quickly. Typically, these responses are beneficial to us in daily life as they help us respond in a fast manner if necessary. However, Bechara et al. (2005) demonstrated that when the amygdala is damaged in brain lesioned patients, these necessary responses are no longer present and the lack of these triggers leads individuals to make decisions that are harmful to them. An example that is often cited is a person with damage to the amygdale; this person can no longer feel how painful it is to lose money and hence if put in situations where he is required to gamble money, he will freely make gambling decisions without thinking about the repercussions of his actions - the pain associated with losing money. The Ventro Medial cortex serves as a trigger structure for somatic states from secondary inducers. It serves as a ‘convergence-divergence’ zone where neurons can join some categories of events based on memory records, the effector structures that execute the somatic state and finally the neural patterns related to the non-conscious or conscious feeling of the somatic state (Bechara & Damasio, 2005). In contrast to the amygdala which is responsible for ‘lower-order’ response states, the VM cortex operates on a ‘higher level’, which means that both thought and reflection come into play before decisions are made. After the initial trigger by the amygdala, the VM cortex takes over and deliberation occurs for a long period of time. Unlike the amygdala, these responses last longer and are slow to habituate. Due to this slow deliberation process, it almost seems as though individuals are thinking about the future and the future consequences of their current actions. This is a very interesting finding and can be repeatedly seen in our daily interactions with each other and how we behave in different situations. I think this process and relationship between the primary and secondary inducers can be explained by an anti-smoking or anti-drug commercial that is shown on television. Picture an ex smoker who used to smoke when drinking a cup of coffee sees in the ad another person who is smoking – while drinking coffee. Instantaneously, his amygdala responds to the external visual stimulus of the person smoking a cigarette while drinking a cup of coffee and causes the ex-smoker to crave the cigarette. However, the craving can be quickly overcome if the smoker thinks about the long term impact of taking that cigarette, i.e., addiction, lung cancer, which causes them to resist the urge to smoke – which demonstrates the activation of the VMPFC.
I think that this concept can be extended to several different kinds of public policy commercials, especially when trying to combat different addictions that people go through – such as drugs, alcohol and smoking. The most interesting thing to note is that neuroscience brings tremendous insight into these situations and gives us a mechanism to understand how our brain works, and how emotions can be induced which finally can be used to prevent us from partaking in harmful behavior. There are two other important mechanisms that play a role in somatic markers, the body-loop and the as-if loop. When activated, the body loop, physically, engages the body and an appropriate somatic marker state is actually re-enacted in the body. The vagal route is especially critical (Bechara, 2002) in this process.
We notice that when a somatic state in the body has been activated, a trigger is sent up to the sensory and neurotransmitter nuclei, present in the brain stem, which then sends a trigger to the VM cortex. The activation of the VM cortex causes it to send a signal to the sensory and neurotransmitter nuclei in the insula and the effector structures, such as the hypothalamus, autonomic centers and PAG in the brain stem. The amygdala is also activated sending down triggers to the same effector structures. Finally these send responses back down to the body. Essentially, regions involved in body mapping (holding patterns of somatic states that help generate feelings), regions involved in triggering of somatic states (amygdala and VM cortex), regions involved in working memory (dorsolateral prefrontal cortex and other high order association cortices) are all triggered. And these states influence activity in regions concerned with motor responses and behavioral actions (Bechara & Damasio, 2005).
The as-if body loop is the other mechanism that plays a role in the activation of somatic markers. Under this scenario, the mental representation of a future state is enough to trigger a somatic state – regardless of its strength. This state is different from the body loop in the sense that somatic states are not physically reenacted in the body, however different neurotransmitter systems are activated. Refer to Figure 1 again; we notice that the sensory and neurotransmitter nuclei in the brainstem trigger responses in the Insula. Further, the VM responds by triggering things in the insula and the neurotransmitters in the brain stem, while the amygdala also triggers responses in the sensory neurotransmitters in the brain stem. Essentially, the two big differences between the as-if and body-loop are that, in the former, there is no involvement of the body and the effector structures are not triggered in this loop.
Applications of the Somatic Marker Hypothesis
There have been several examples of applying the SMH when studying people’s investment behavior. Studies have typically been done conducted with brain lesioned patients to document whether a lesion in their emotional mechanism impairs or enhances their ability to make decisions under risky conditions. Using emotions in decision making has been shown to have both positive and negative impacts on behavior. Using the Iowa Gambling Task, researchers have shown that individuals with damage to their emotional system perform poorly on this task than those who have an intact emotional system (Bechara et al., 1997; Damasio, 1994; Rogers et al., 1990). However, there are some cases where damage to the emotional mechanism can actually prove to be advantageous in some of the decisions that we make. Damasio (1994) gives the example of the patient with damage to the VMPFC driving down an icy road and his reaction to when the car skids. The patient with the VMPFC damage does not have his ‘fear system’ intact and this causes him to not worry about spinning out of control and he continues to drive safely over the icy patch. Damasio compared this to the case where a person with no damage to the VMPFC – thereby having an intact fear system – would have the tendency to panic and spin out of control as he crossed the icy patch.
In terms of investment behavior, it has been shown that individuals who are deprived of normal emotional reactions might, in some situations, make more advantageous decisions than those not deprived of such reactions (Shiv, Loewenstein, Bechara, Damasio & Damasio, 2005). These authors demonstrated this by having subjects play an investment game where they had to invest in every round because investment yielded higher expected value than not investing. They showed that normal participants behaved differently from participants who had lesions in the amygdale, orbitofrontal cortex or right insular or somatosensory cortex. They found that participants with lesions behaved in a more risk taking manner (invested money in every round) than normal subjects when faced with positive expected value gambles. These positive expected value gambles are those that humans often shy away from because we prefer certain outcomes to outcomes which have probabilities associated with it.
Neuroscience Techniques
In light of discussing some of these theories and applications of neuroscience in decision making, it is important to see what techniques are being used to study the brain. In the past few years, methods used in understanding brain patterns and neural activity have advanced tremendously. On a very primitive level, many of our physiological responses can be easily measured by just observing people. For example, pupil dilation has been correlated with mental effort (Kahneman & Peavler, 1969 – see Camerer, Lowenstein, and Prelec for cite). Further, anxiety, sexual arousal, mental concentration and other motivational states can be measured by blood pressure, skin conductance responses, and heart rate, (Camerer, Lowenstein & Prelec, 2004).
In recent years, more sophisticated techniques have emerged to understand neural activity and how our brains respond to stimuli presentation. They include Electroencephalograph (EEG), Magnetic Resonance Imaging (MRI) and Magnetoencephalography which all measure the changes in the electrical current in the brain – only using different techniques – MRI and MEG use magnets to measure brain waves while EEG’s use electrodes which are attached to the outside of your head. Next, Computerized Tomography (CT), which takes X-ray images of the brain; Positron Emission Tomography (PET) which measures emissions from radio-active particles in the blood. The last two are the most sophisticated methods – Functional Magnetic Resonance Imaging (fMRI) – which rely on magnetic properties to measure blood flow and the Single Neuron Measurement where tiny electrodes are inserted into the brain to measure the responses of single neurons. The Single Neuron Measurement is a very invasive procedure and is currently used only on animals. Looking at the above techniques, we can see that the techniques used to study the human brain have come a long way. Each of these techniques has different benefits and have some costs associated with them.
Neuroscience & Marketing – Avenues for further research
Neuroscience has recently seen a surge of applications in marketing research to understand consumer behavior. A very practical implication of using neuroscience techniques to understand consumer behavior is that if firms learn what activates different regions of a persons brain (for example the amygdale – to understand whether and how to tailor the emotional aspect of their message), they can design effective marketing campaigns to get people to purchase their products. I will discuss two unexplored avenues of research in the marketing discipline that have not yet seen the application of neuroscience techniques in better understanding consumer behavior.
The Elaboration Likelihood Model (henceforth ELM) was proposed by Petty & Cacioppo (1986), and states that people desire to attain correct attitudes but the extent and nature of their processing of persuasive argument depends on their motivation and ability for issue relevant thinking. Elaboration refers to the extent that people think about issue relevant arguments contained in persuasive messages. When there is high motivation to process the messages, the likelihood that people will elaborate on the message will be high.A common mode of thought is that when making decisions we elaborate on either the cognitive characteristics or on emotional or affect laden attributes of the message. Most research on ELM has focused on strict manipulations of the cues that can be used in studying the elaboration process in which people engage when making decisions. An important thing to note here is that elaboration or the process of deliberating (which is what we are interested in) is not really being measured. What is being measured is how people respond to different types of message cues. In marketing, there are two methods or proxies that are used to measure how much the extent of elaboration that we engage in when making decisions. First, in an experimental setting, individuals are asked to think and list reasons for why they purchase or choose the products that they do. Next, the amount that people elaborate is measured by counting the number of thoughts that people list which is then used as a proxy for the extent of elaboration. Another method that has been used is that of Response Time. This is used when experiments are conducted on the computer. As people are asked to make various decisions, we measure how much time it takes them to make that decision (the time till when they respond to the question). If this response time is long – we conclude that they have elaborated a lot and if the time is short, we conclude that they have hardly elaborated.
These two measures (number of responses and response time) are then used as proxies for understanding how much people elaborate on messages or ads. This amount is then used as an independent variable in a model to determine what products consumers choose. For example, do consumers elaborate more or less based on whether they are going to choose a sensory product (e.g., chocolate) or a functional product (e.g., glue stick). So if a person lists more thoughts or reasons as to why they choose a pack of chocolate as opposed to a glue stick, the conclusion has been that people think more about sensory products and do not want to give up these sensory products (loss averse) and hence they think more about it before giving it up. As one notices these are not the best measures of elaboration and one can go so far as to say that we are not really measuring the amount that people elaborate before making a decision to purchase a product.
It would be interesting to see whether we can use Neuroscience techniques here to measure the amount that people elaborate. Assuming that we do this using fMRI techniques, it will be interesting to show that when people are deliberating or thinking before they make a decision, certain parts of the brain activated and if we can measure how long these parts of the brain are activated, we should have a cleaner measure of the amount of elaboration that people do when making decisions. Further, we might also be able to gain insight into the neural process of activation as people deliberate and make decisions.
Another interesting avenue of research where neuroscience techniques can play a role is that of the relationship between goal progression and reward anticipation. The capacity to seek rewards as goals is essential for the survival and reproduction of all mobile creatures (Mcclure, York & Montague, 2004). Rewards are operationally defined as those stimuli that positively reinforce behavior. Research has been done on understanding our response to unlearned primary rewards such as food, water and sexual stimuli and also on conditionally learned rewards such as monetary gain. Other research has used the somatic marker hypothesis to understand human behavior in the anticipation of a reward or punishment (Behcara & Damasio, 2005). The authors showed that normal subjects generated SCRs (Skin Conductance Responses) in the anticipation of a reward or punishment, in other words immediately prior to the decision that they make or in the deliberation process of when they would make a decision.
It has been identified that reward anticipation activates different parts of the brain, and one can move on to see whether this can be applied to study goal progression in humans. Past research in this area has stated that the goal is to get a reward or to avoid a punishment. What I am proposing is different from this description of a goal. I suggest looking at the anticipated reward that one will get from successfully completing a goal. For example, one can look at the goal of successful completion of our Ph.D. programs and the anticipated reward associated with this completion (e.g., getting a good job). I think that it will be interesting to study whether different parts of our brain are activated as we progress through sub-goals and get closer to goal completion.
Limitations of Neuroscience
While there are several benefits of using neuroscience techniques in understanding human behavior and decision making, there are some questions that neuroscience cannot answer by itself and needs the help of experimental methodology and theories to understand why we behave in the manner that we do. The key limitation of neuroscience techniques, aside from being expensive, is that it is able to only identify that different regions of our brain are activated when we are in certain situations. These techniques are not able to provide an explanation or a reason (behavioral) as to why we respond in the manner that we do. An example of this is seen in a paper by McClure et al., (2004) which talks about how human brains respond towards Coke and Pepsi. The most interesting result from this paper was that in a blindfolded test consumers showed no significant difference in their preference for Coke or Pepsi. However, when primed with a picture of a Coke can or a Pepsi can, there was a significant difference in the preference for the two brands, where subjects demonstrated a much higher preference for Coke than for Pepsi. The authors concluded that brand knowledge biases preference decisions and choice for Coke products over Pepsi products. This is really interesting considering the fact that the products have very little difference in their chemical composition. It essentially states that Coke has done something different in their marketing strategy from Pepsi. While this paper was able to identify that humans respond differently when primed with Coke or Pepsi products, they were not able to explain (nor did they venture an explanation) why this might occur. I think this is where experimental methodology would help bolster the understanding as to why people make the choices that they do. Neuroscience techniques are becoming important in understanding human decision making. Initially, these methods were used to understand human behavior in risky situations. However, more recently, the techniques have been used to study things such as emotions in decision making and how people respond to different stimuli. There are some limitations to neuroscience techniques as it serves to tell us what happens in the brain or what is activated when we make decisions or are in the process of making decisions or responding to outcomes. It does not however give us any insight into why we make these decisions and why we respond in the manner that we do. At this juncture in research across various disciplines, especially Marketing, I think that a synergy between neuroscience techniques and behavioral experiments will provide tremendous insight into understanding human behavior and decision making.
References
Bechara, A., and Damasio, A. R., (2005). The Somatic Marker Hypothesis: A neural theory of economic decision. Games and Economic Behavior (52), 336-372
Bechara, A., Damasio, H., Tranel, D. and Damasio. A. (1997). Deciding Advantageously Before Knowing the Advantageous Strategy. Science 275, 1293–1294.
Buchel, C., Morris, J., Dolan, R.J., Friston, K.J. (1998). Brain systems mediating aversive conditioning: An eventrelated fmri study. Neuron 20 (5), 947–957.
Camerer, C., Loewenstein, G. and Prelec, D. (2004). Neuroeconomics: Why Economics Needs Brains. Scandanavian Journal of Economics, 106(3), 555-579
Cooper, J. R., Bloom, F., E., and Roth, R. H., (2002). The Biochemical Basis of Neuropharmacology, Oxford University Press, Cambridge, MA ]] Damasio, A. R. (1994). Descartes’ Error: Emotion, Reason, and the Human Brain, G. P. Putnam & Sons, New York.
Dolan, R.J., Fletcher, P., Morris, J., Kapur, N., Deakin, J.F.W., Frith, C.D. (1996). Neural activation during covert processing of positive emotional facial expressions. Neuroimage 4 (3), 194–200.
Kahneman, D. and Peavler, W. S. (1969). Incentive Effects and Pupillary Changes in Association Learning. Journal of Experimental Psychology, 79, 312–318
Kahneman, D. and Tversky, A. (1979). Prospect Theory: An Analysis of Decision under Risk. Econometrica, 47, 263–291
McClure, S. M., Li, J., Tomlin, D., Cypert, K. S., Montague, L. M. and Montague, P. R., (2004). Neural Correlates of Behavioral Preferences for Culturally Familiar Drinks. Neuron, 44, 379-387
Petty, R. E. and Cacioppo, J. T., (1986). Communication and Persuasion: Central and Peripheral Routes to Attitude Change; Springer Verlag
Pirouz, D. M. (2005). The Neuroscience of Behavioral Decision Making – Working Paper, University of California, Irvine
Shiv, B., Lowenstein, G., Bechara, A., Damasio, H. and Damasio, A. R. (2005). Investment Behavior and the Negative Side of Emotion. American Psychological Society, 16-6
Trepel, C., Fox, C. R., and Poldrack, R. A., (2005). Prospect theory on the brain? Toward a cognitive neuroscience of decision under risk. Cognitive Brain Research
Tversky, A. and Kahneman, D., (1981). The Framing of Decisions and the Psychology of Choice. Science, 453-458
Viner, J. (1925). The Utility Concept in Value Theory and Its Critics. Journal of Political Economy, 33, 369–387
