Comparing Apple and Religion: Perspectives Related to Brainwaves

Center for Neural Decision Making

Fox School of Business

Temple University

Co-hosted by MIT and the University of Michigan

3rd

ANNUAL INTERDISCIPLINARY SYMPOSIUM

ON DECISION NEUROSCIENCE

PROGRAM and ABSTRACTS

May 3 – 5, 2013

MBA Commons, 7th

Floor

Alter Hall

Organizing Committee:

Angelika Dimoka Temple University Vinod Venkatraman Temple University Drazen Prelec MIT Carolyn Yoon University of Michigan

Center for Neural Decision Making

Fox School of Business, Temple University

The Center for Neural Decision Making (CNDM) is directed by Professor Angelika Dimoka, and is comprised of academic researchers and industry partners interested in the potential of brain imaging to drive both theory and practice. The Center is founded through the generous support of the Fox School of Business and Management and the Institute for Business and Information Technology (IBIT).

The CNDM offers training and research opportunities for undergraduate and graduate students, along with scholarships and research assistantships for doctoral students and post-docs.

3

Friday, May 3

Neuroscience in Practice: Where Minds Intersect

12:00 PM Lunch 1:00 Session 1: Consumer Neuroscience: Bridging the Gap between

Research and Practice

1:00 Consumer Neuroscience: Progress and Promises Vinod Venkatraman, Temple University 1:10 Brain activity correlates of television advertising effectiveness Richard Silberstein, Neuro-Insight, Swinburne University of Technology 1:30 Examples of Validation in Consumer Neuroscience from Innerscope Research

Carl David Marci, Innerscope Research, Massachusetts General Hospital, Harvard Medical School

1:50 Neuroscience Applications for Advertising and Branding Insights Karl Rosenberg, MSW Neuroscience 2:10 Neuroscience and Behavior Change Emily Falk, University of Michigan 2:25 “Neuro 2.0”: An Academia - Practice Research Study Paul A Pavlou, Temple University

2:35 Coffee Break

3:00 Session 2: Constructs in Consumer Neuroscience: Definitions, Methods, and Best Practices Leaders from academia and practice join in discussions on how to define key constructs commonly used in practice, and identify the best neuroimaging and physiological tools to measure them.

7:00 PM Opening Reception

4

Saturday, May 4

8:00 AM Continental Breakfast 8:50 Welcome Angelika Dimoka Director, Center for Neural Decision Making, Temple University

9:00 Opening Remarks M. Moshe Porat Dean and Laura H. Carnell Professor, Fox School of Business, Temple University Michele M. Masucci Vice Provost for Research and Graduate Education, Temple University

9:15 Session 1: Decision Neuroscience – Yesterday, Today and Tomorrow Chaired by: Drazen Prelec, MIT

9:15 Decision Neuroscience: Where Have We Been, Where Are We Going? Paul Glimcher, New York University

9:40 Identifying Emotions on the Basis of Neural Activation George Loewenstein, Carnegie Mellon University

10:05 Predicting the choice of a subject before it was made, using direct recording from the human brain

Moran Cerf, New York University

10:30 Coffee Break

11:00 Session 2: Influencing Consumer Choice Chaired by: Carolyn Yoon, University of Michigan

11:00 Neuro-cultural Mechanisms of Choice Justification

Shinobu Kitayama, University of Michigan

5

11:25 Paying it forward: Generalized reciprocity and the limits of generosity

Michael Norton, Harvard Business School 11:50 Reward substitution: How consumers can be incentivized to choose smaller

food portions Martin Reimann, University of Southern California Antoine Bechara, University of Southern California

12:15 PM Lunch Poster Session

2:00 Session 3: Risky Choice Chaired by: Vinod Venkatraman, Temple University

2:00 Individual differences in gist representation versus sensation-seeking in neuroimaging of framing effects

Valerie Reyna, Cornell University

2:25 Asymmetric belief formation in human brain Tali Sharot, University College of London

2:50 Personality influences intertemporal choice: Behavioral and brain evidence Joshua Manning, MIT 3:15 Rational preference reversals in real time Joseph McGuire, University of Pennsylvania

3:40 Coffee Break

4:15 Data Blitz 7:30 PM Social Event – Water Works Restaurant and Lounge

6

Sunday, May 5

8:30 AM Continental Breakfast 9:30 Session 4: Neuromarketing and Consumer Neuroscience Chaired by: Antoine Bechara, University of Southern California

9:30 Building Multi-dimensional models of human choice Read Montague, Virginia Tech & University College of London 9:55 Predicting consumer attention and choice: Insights from Computational

Neuroscience Milica Mormann, University of Miami

10:20 Sleep on It? The effects of sleep on subjective preference-based choice Uma Karmarkar, Harvard Business School

10:45 Coffee Break 11:15 Session 5: Tools and Advances in Decision Neuroscience Chaired by: Angelika Dimoka, Temple University

11:15 Do EEG Oscillations Predict Consumer Choice? Ale Smidts, Erasmus University

11:40 Correlation and Causation: Understanding and Altering Decision-Making in the human brain

Sameer Sheth, Columbia University Medical Center 12:05 Using fNIR to study the effects of music on behavior

David Gefen, Drexel University

12:30 PM Lunch and Closing Remarks William E. Aaronson Associate Dean for Graduate Programs and Research, Fox School of Business,

Temple University

7

3rd annual Interdisciplinary Symposium on Decision Neuroscience

Abstracts

Neuroscience in Practice: Where Minds Intersect

Brain activity correlates of television advertising effectiveness. Richard Silberstein, Neuro-Insight, Swinburne University of Technology Examples of Validation in Consumer Neuroscience from Innerscope Research Carl D. Marci, M.D. Innerscope Research, Inc., Massachusetts General Hospital, Harvard Medical School Neuroscience Applications for Advertising and Branding Insights Karl Rosenberg, MSW Neuroscience Neuroscience and Behavior Change Emily Falk, University of Michigan

8

Author

1st Author: Richard Silberstein (Presenter)

2sd

Co-Author: Heather Andrew

3rd

Co-Author: Geoffrey Nield

4th

Co-Author: Pranav Yadav

Affiliations

1st Author: Neuro-Insight Pty Ltd, Melbourne, Australia / Swinburne University of Technology,

Melbourne, Australia

2sd

Co-Author: Neuro-Insight-UK, Newcastle, United Kingdom.

3rd

Co-Author: Neuro-Insight Pty Ltd, Melbourne, Australia

4th

Co-Author: Neuro-Insight-USA, New York, United States of America.

TITLE: Brain activity correlates of television advertising effectiveness.

ABSTRACT: In this presentation, we report on a study utilizing Steady State Topography (SST), an evoked

potential methodology to examine the relationship between brain activity while viewing television

advertising and a surrogate indicator of advertising effectiveness, specifically, in-market status or whether

the participant is in the target group of the advertisement or not. We have previously reported that SST

measures of lateral prefrontal brain activity are robustly correlated with second-by second changes in long-

term memory encoding (Silberstein, et al 2000 International Journal of Psychophysiology. 2000; 39:79-85).

In a subsequent study (Silberstein & Nield, International Journal of Advertising. 2008; 27: 359 – 380) we

demonstrated that the levels of long-term memory encoding occurring at the time of branding is correlated

with an increase consumer brand preference for the advertised brand. As such, we suggest that this measure

may constitute an indicator of advertising effectiveness. In this presentation we report on an SST study

where brain activity measures were collected by Neuro-Insight in a series of 6 syndicated studies conducted

in the USA and UK between 2010 and 2012 involving a total 687 participants and 40 advertisements. Each

study involved between 110 and 119 participants with closely matching males and female numbers. Age

ranged from 18 – 65 and matched the general population. Participants viewed television advertisements

organized in advertising breaks within a prime time television program. All participants completed a series of

questionnaire to determine their membership of the target groups for each of the client advertisements. For

each of the advertisements, we determined the level of long-term memory encoding coinciding with the last

occurrence of explicit branding. We hypothesize that the SST measure of long-term memory encoding at

branding will be higher for the target group than the non-target group. Our hypothesis was robustly

confirmed with a paired t-test (t=5.4 , df=39 , p=2.8X10-6 ). To see whether the effect was restricted to the

point of branding or was a characteristic of the entire advertisement, we repeated the analysis applying it to

the SST measure of long-term memory encoding averaged over the entire duration of the advertisement

(typically 30 sec). In this case, there was no significant difference between the target and non-target groups

(t=1.9 , df=39 , p=0.07). The implications of these findings will be discussed in terms of the importance of

temporal resolution in the study of time varying cognitive stimuli such as television advertisements.

9

Author

1st Author: Carl D. Marci, M.D.

Affiliations

1st Author: Innerscope Research, Inc., Massachusetts General Hospital, Harvard Medical School

TITLE: Examples of Validation in Consumer Neuroscience from Innerscope Research

ABSTRACT: Validation is an often used but little defined construct in market research in general but in

consumer neuroscience in particular. In fact, consumer neuroscience has been criticized for not

publishing or sharing examples of validation from their own research, often citing competitive or

proprietary rationale. Dr. Marci, Chief Science Officer and co-founder of Innerscope Research, will

present case studies, peer review studies and proprietary examples of validation using biometrics

measures of emotional engagement for discussion and questioning. The presentation will focus on four

types of validation in particular: 1) industry review, 2) scientific/peer review, 3) client review, and 4)

partner review.

10

Author

1st Author: Karl Rosenberg

Affiliations

1st Author: MSW●ARS Research, Inc.

TITLE: Neuroscience Applications for Advertising and Branding Insights

ABSTRACT: Modern Psychology periodically reminds the marketing research practitioner that most

decisions and choices are made intuitively and automatically, without conscious control or effort and

that the practitioner’s nearly exclusive reliance on the survey questionnaire likely limits what can be

learned about these processes. Daniel Kahneman’s “Thinking, Fast and Slow” has served as one such

eloquent and compelling reminder that the marketing research industry must look beneath the usual

“System 2” survey responses if we are to more fully understand how consumers interact with and are

influenced by our brands.

MSW●ARS presents an overview of its commercial applications using EEG, GSR, and Eye Tracking to

help advertisers produce more sales effective advertising and branding elements. Techniques are

described and cases presented to illustrate how these tools are being applied in the marketplace.

11

Author

1st Author: Emily Falk,

Affiliations

1st Author: Assistant Professor of Communication Studies, Department of Psychology, University of

Michigan

TITLE: Neuroscience and predicting behavior change

ABSTRACT: Can neural responses predict behavior change outside of the lab? In this talk, I will

provide a brief overview of recent research in our laboratory in which neural activity collected using

fMRI has been used to predict individual behavior change and population level outcomes.

I will discuss ways in which consumer research can move beyond the scanner to linking neuroscience

data with health, policy and consumer choice relevant outcomes.

12

3rd annual Interdisciplinary Symposium on Decision Neuroscience

Abstracts

Session 1: Decision Neuroscience Yesterday, Today and Tomorrow

Decision Neuroscience: Where Have We Been, Where Are We Going? Paul Glimcher, New York University Identifying Emotions on the Basis of Neural Activation George Loewenstein, Carnegie Mellon University Predicting the choice of a subject before it was made, using direct recording from the human brain Moran Cerf, New York University

13

Author

1st Author: Paul Glimcher

Affiliations

1st Author: Center for Neural Science – New York University

TITLE: Decision Neuroscience: Where Have We Been, Where Are We Going?

ABSTRACT: Although barely 15 years old, decision neuroscience has racked up an astonishing set of

accomplishments. Where two decades ago we knew little about where in the brain value of any kind was

represented, today complex models which capture both neurobiological constraints and economic

insights are becoming commonplace. Monographs, textbooks, articles, international scientific societies

and even an academic journal define the landscape in which decision-neuroscientists operate today.

Perhaps what is most striking about this is how rapidly this discipline has coalesced. In 2001 what can

plausibly be called the first academic conference on neuroscientific and economic approaches to

decision-making was held at Princeton University for a handful of social and natural scientists who had

little knowledge of each other’s fields. Today a growing group of truly interdisciplinary scholars, well-

represented by the attendees of this meeting, have been able to accomplish far more than those early

pioneers had imagined was possible. In this talk Glimcher will provide an overview of the last 15 years

of progress in decision-neuroscience highlighting both what has been accomplished and what important

challenges lie ahead.

14

Author

1st Author: Karim S. Kassam

2nd Co-Author: Amanda R. Markey

3rd Co-Author: Vladimir L. Cherkassy

4th Co-Author: George Loewenstein (Presenter)

5th Co-Author: Marcel Adam Just

Affiliations

1st Author: Carnegie Mellon. Dept Social and Decision sciences

2nd Co-Author: Carnegie Mellon. Dept Social and Decision sciences

3rd Co-Author: Carnegie Mellon. Dept Social and Decision sciences

4th Co-Author: Carnegie Mellon. Dept Social and Decision sciences

5th Co-Author: Carnegie Mellon. Dept of Psychology

TITLE: Identifying Emotions on the Basis of Neural Activation

ABSTRACT: We present tests of a novel methodology for identifying specific emotion that applies

machine learning algorithms to data derived from neuroimaging. We demonstrate the ability to identify

specific emotions experienced by an individual at well over chance accuracy in three contexts: 1)

identification on the basis of neural activation of the same individual in other trials, 2) identification on

the basis of neural activation of other individuals who experienced similar trials, and 3) identification on

the basis of neural activation of the same individual to a qualitatively different type of emotion

induction. The reliability of these identifications suggests the possibility of inferring a person’s

emotional reaction to an arbitrary stimulus – a flag, a brand name, or a political candidate, for example –

on the basis of neural activation. Finally, we employ a factor analysis to uncover groups of voxels that

respond similarly to these emotional cues and which enable successful identification.

15

Author

1st Author: Moran Cerf, PhD

Affiliations

1st Author: NYU Stern School of Business, UCLA department of Neurosurgery

TITLE: Predicting the choice of a subject before it was made, using direct recording from the human

brain

ABSTRACT: Using invasive recording of activity directly from the brain of patients undergoing brain

surgery we are able to trace individual neurons whose activity predicts volition and choice. We are able

to probabilistically decode the choice of a two-alternative-forced-choice and make predictions on

outcomes prior to volition and prior to conscious urge to make the choice, as reported by the subject.

Furthermore, in a number of trials, we are able to use the predictions to affect the coming choice and

alter the outcome such that the decision is manipulated by events in accessible to the patient. This taps

into the underlying mechanisms of choice and allows us to estimate the brain mechanisms involved in

decision-making.

16

3rd annual Interdisciplinary Symposium on Decision Neuroscience

Abstracts

Session 2: Influencing Consumer Choice

Neuro-cultural Mechanisms of Choice Justification Shinobu Kitayama, University of Michigan Paying it forward: Generalized reciprocity and the limits of generosity Michael Norton, Harvard Business School Reward substitution: How consumers can be incentivized to choose smaller food portions Martin Reimann, University of Southern California Antoine Bechara, University of Southern California

17

Author

1st Author: Shinobu Kitayama

Affiliations

1st Author: University of Michigan

TITLE: Neuro-Cultural Mechanisms of Choice Justification

ABSTRACT: How will choice produce choice-justifying preference changes? Drawing on both

previous behavioral work and more recent neuroimaging work in the area, Kitayama, Chua, Tompson,

and Han (2013, NeuroImage) have proposed that when choosing between equally attractive options, the

chooser experiences negatively arousing cognitive conflict (i.e., dissonance), which in turn motivates her

to look for positive distinctive features in one of the options in an effort to reduce the dissonance. If the

positive distinctive features are found and, moreover, they are judged to be relevant to the self,

preferences will be updated. The chooser will make a choice on the basis of the updated preferences.

This model shares with traditional social psychological theories the premise that dissonance plays a

crucial role in choice justification and, yet, differs from the latter in assuming that choice justification is

accomplished during the choice, not afterward. More broadly, the model implies that the chooser will

have good reasons to believe that their choice is evidence-based and, thus, rational and, yet, the reasons

that make her believe her own rationality are likely identified in self-gratifying, and arguably irrational,

fashion. The model has been supported by recent neuroimaging evidence and, moreover, it sheds new

light on known cross-cultural variations in the conditions in which choice justification is observed.

18

Author

1st Author: Kurt Gray

2nd

Co-Author: Adrian F. Ward

3rd

Co-Author: Michael I. Norton (Presenter)

Affiliations

1st Author: Department of Psychology, Rutgers University

2nd

Co-Author: Department of Psychology, Rutgers University

3rd Co-Author: Harvard Business School

TITLE: Paying It Forward: Generalized Reciprocity and the Limits of Generosity

ABSTRACT:

When people are the victims of greed or recipients of generosity, their first impulse is often to pay back

that behavior in kind. What happens when people cannot reciprocate, but instead have the chance to be

cruel or kind to someone entirely different—to pay it forward? In 5 experiments, participants received

greedy, equal, or generous divisions of money or labor from an anonymous person and then divided

additional resources with a new anonymous person. While equal treatment was paid forward in kind,

greed was paid forward more than generosity. This asymmetry was driven by negative affect, such that a

positive affect intervention disrupted the tendency to pay greed forward. Implications for models of

generalized reciprocity are discussed.

19

Author

1st Author: Martin Reimann (Presenter)

2nd

Co-Author: Deborah MacInnis

3rd

Co-Author: Antoine Bechara (Presenter)

Affiliations

1st Author: University of Arizona, Eller College of Management

2nd

Co-Author: University of Southern California, Marshall School of Business

3rd

Co-Author: University of Southern California, Department of Psychology / Brain & Creativity

Institute

TITLE: Reward Substitution: How Consumers Can Be Incentivized to Choose Smaller Food Portions

ABSTRACT: This research establishes that offering non-food rewards (i.e., lottery tickets) bundled

with smaller food portions as an alternative to full-portion meals can substantially decrease chosen

portion sizes. Study 1, a behavioral experiment and field survey, establishes this effect, which I call

“reward substitution,” among adult consumers from the general population. The impact of possible

boundary conditions and moderating variables on this effect are investigated and reported. Study 2, a

functional magnetic resonance imaging (fMRI) experiment, finds that the reward substitution effect is

based on a common neurophysiological process that becomes evident in structures of the mesolimbic

dopamine system. The current research introduces reward substitution as another determinant of smaller

portion choice and has implications for portion size research, the psychology of self-control, and

classical theories of motivation and economic utility. Implications for future research are provided, and

suggestions for practical implementation and public policy making are discussed.

20

3rd annual Interdisciplinary Symposium on Decision Neuroscience

Abstracts

Session 3: Risky Choice

Individual differences in gist representation versus sensation-seeking in neuroimaging of framing effects Valerie Reyna, Cornell University Asymmetric belief formation in human brain Tali Sharot, University College of London

Personality influences intertemporal choice: Behavioral and brain evidence Joshua Manning, MIT Rational preference reversals in real time Joseph McGuire, University of Pennsylvania.

21

Author

1st Author: Valerie Reyna (Presenter)

2nd

Co-Author: Christina Chick

3rd

Co-Author: Rebecca Weldon

4th

Co-Author: Jonathan Corbin

5th

Co-Author: Evan Wilhelms

Affiliations - Cornell University

TITLE: Individual differences in gist representation versus sensation-seeking in neuroimaging of

framing effects

ABSTRACT: Although dual-process theories attribute framing effects to an emotional bias (Roiser et

al., 2009), behavioral evidence suggests that people who take more risks in real life actually show

diminished framing effects (and, often, reverse framing; Reyna, Estrada, DeMarinis et al., 2011). We

aimed to identify neural correlates associated with individual differences in gist representations vs.

sensation-seeking in framing behavior.

Thirty-two young adults completed a risky choice framing task while being scanned. Subjects made

60 decisions, involving either lives or money. Subjects also completed personality questionnaires

including a 19-item sensation seeking scale (SSS; Zuckerman, 2008), the Adolescent Risk Questionnaire

(ARQ; Gullone, 2000), and other measures.

As standard framing increased, activation during “framing” decisions increased in the bilateral

superior and inferior parietal lobes, and in the middle and superior frontal lobes (e.g., BA 10). As

sensation seeking increased, activity during decisions contrary to the typical framing pattern increased in

the superior and medial temporal lobes. In particular, higher sensation seekers had more activation in

the hippocampus and amygdala when choosing risky options (compared to safe options) in the gain

frame in the verbatim condition. Sensation seeking predicted real-life risk taking as measured by the

ARQ.

These results suggest that although emotional and motivational factors may influence framing

behavior, this influence runs in the opposite direction predicted by most dual-process theories. In

particular, higher sensation seeking was associated with less standard framing and with higher activation

in emotional processing areas, such as the amygdala, during “no framing” decisions. In contrast,

standard framing effects were most reliably associated with activation in the superior and inferior

parietal lobes, as well as some regions (such as BA 10) that most dual-process theories predict would

decrease framing effects—not underlie them. This suggests that framing effects are driven by cognitive

(in addition to emotional) circuitry (Dennis, Bowman & Vandekar, 2012), and that reverse framing

effects are driven by emotional (in addition to cognitive) processing.

22

Author

1st Author: Tali Sharot

Affiliations

1st Author: University of College London

TITLE: Asymmetric belief formation in the human brain

ABSTRACT: Humans display an aversion to learning from bad news, discounting the strength of the

evidence. In contrast, when the information is favorable people effectively incorporate new information

into existing beliefs in a Bayesian manner. This phenomenon has important societal implications,

potentially contributing to the generation of financial market bubbles and ill preparedness in the face of

natural disasters. How can good and bad news have such different impacts on belief formation? Our

neuroimaging results reveal that the tendency to discount bad news, while adopting good news, is

related to a failure of distinct frontal regions (including the IFG and MPFC) to code for negative errors.

This failure is absent in depressed individuals. Furthermore, we show that the ability to learn from

negative information can be selectively improved (using TMS and/or psychological stress) or disrupted

(using L-DOPA) and has a different developmental trajectory to that of learning from positive

information. Together, the findings suggest that the biological mechanisms for processing favorable and

unfavorable information can be differentiated. The volatility of the latter may lead to the creation of

positively biased beliefs.

23

Author

1st Author: Joshua Manning (Presenter)

2nd

Co-Author: Trey Hedden

3rd

Co-Author: Drazen Prelec

4th

Co-Author: John D. E. Gabrieli

Affiliations

1st Author: Dep. of Brain and Cognitive Sciences, McGovern Institute for Brain Research, MIT

2nd

Co-Author: Massachusetts General Hospital, Harvard Medical School

3rd

Co-Author: Sloan School of Management, Department of Brain and Cognitive Sciences, MIT

4th

Co-Author: Dep. of Brain and Cognitive Sciences, McGovern Institute for Brain Research, MIT

TITLE: Personality Influences Intertemporal Choice: Behavioral and Brain Evidence

ABSTRACT: We investigated how personality is associated with individual differences in temporal

discounting and neural activity. Participants completed a Big-Five personality inventory. In a functional

magnetic resonance imaging scanner (fMRI), participants then completed a discounting task where they

chose between a series of two options of monetary rewards, a smaller monetary reward with a shorter

delay and a larger monetary reward with a longer delay up to one year. Discount functions were

estimated for each individual using the two-parameter diminishing sensitivity discount function: ( ) ( ( ) ). represents impatience and represents diminishing temporal sensitivity,

becoming more hyperbolic as approaches 0. We correlated each participant’s personality scores with

each parameter. Conscientiousness was significantly correlated with lower and neuroticism marginally

correlated with higher . In addition, conscientiousness was significantly correlated with higher and

neuroticism was significantly correlated with lower . High levels of conscientiousness were associated

with more patience and relatively more consistent time preferences, while high levels of neuroticism

were associated with more patience and less consistent time preferences. Next we correlated each

individual’s subjective value (SV) of the reward and personality to neural activity. When choosing the

shorter delay, SV and conscientiousness were correlated with greater activations in the ventral striatum,

bilateral dorsolateral prefrontal cortex (DLPFC), orbitofrontal cortex (OFC), and bilateral parietal

cortex. There were no correlations with neuroticism and SV when choosing the shorter delay. When

choosing the longer delay, SV and neuroticism were also correlated with greater activations in the

ventral striatum, bilateral DLPFC, OFC, and bilateral parietal cortex. There were no correlations with

conscientiousness and SV when choosing the longer delay. These two contrasts reveal a network of

common regions that are associated with reward, cognition and executive control. These regions are

engaged when choosing the option that is relatively less preferred for each personality type (i.e. high

conscientiousness with short delays/inconsistent time preferences and high neuroticism with long

delays/consistent time preferences). It is likely that choosing the relatively less preferred delay for each

personality type requires the integration of the reward system with regions involved in cognition and

planning.

24

Author

1st Author: Joseph T. McGuire (Presenter)

2nd

Co-Author: Joseph W. Kable

Affiliations - University of Pennsylvania

TITLE: Rational preference reversals in real time

ABSTRACT: Obtaining a desirable long-run outcome, such as physical fitness, financial stability, or

skill development, requires sustaining a preference for that outcome over time. Otherwise reversals can

occur: a delayed reward might be favored at first but abandoned before it is attained. Reversals of this

sort are often referred to as delay-of-gratification failures, and invite explanation in terms of ego

depletion, temptation, or dual-system competition. However, such a reversal could be rational if new

information signaled that the reward was less valuable than one first believed. When temporal

uncertainty exists, the mere passage of time can provide such a signal. A long delay so far can imply a

longer-than-expected delay yet to come. We hypothesize that decision makers reappraise future rewards

as time passes, and that this dynamic reappraisal manifests both in overt decisions and in the activity of

value-responsive brain regions.

Methods: Experimental participants had repeated opportunities to wait for monetary rewards at

uncertain delays, striving to maximize their overall reward in a fixed total time. By manipulating the

distribution of possible delays, we created a "low-persistence" environment in which the best strategy

was to quit each trial after 20 seconds if the reward had not yet arrived, and a "high-persistence"

environment in which the best strategy was to wait as long as necessary on each trial (up to 40 seconds).

Each participant learned about the two environments through direct experience. After training,

participants performed the task while undergoing fMRI.

Results: Participants were responsive to our manipulation of delay-length statistics, showing greater

willingness to wait in the high-persistence environment. Neuroimaging analyses focused on brain

responses that differed across physically identical delay intervals in the two environments. Ventromedial

prefrontal cortex showed a trial-onset-locked response that increased more steeply in the high-

persistence environment than the low-persistence environment, matching theoretical subjective value

estimates derived from a foraging model.

Conclusions: Reversing patterns of choice over time can stem from a rational, dynamic reappraisal

of awaited outcomes. This idea stands in contrast to accounts that attribute reversals to self-control

failure or resource depletion. Our results imply that an effective way to alter persistence behavior is to

modify decision makers' temporal expectations. Neural findings are consistent with the possibility that

both successes and failures of persistence reflect the operation of a unitary system for subjective

evaluation.

25

3rd annual Interdisciplinary Symposium on Decision Neuroscience

Abstracts

Session 4: Neuromarketing and Consumer Neuroscience Building Multi-dimensional models of human choice Read Montague, Virginia Tech & University College of London Predicting consumer attention and choice: Insights from Computational Neuroscience Milica Mormann, University of Miami Sleep on It? The effects of sleep on subjective preference-based choice Uma Karmarkar, Harvard Business School

26

Author

1st Author: P. Read Montague (Presenter)

2nd Co-Author: Ann Harvey,

3rd Co-Author: Terry Lohrenz,

4th Co-Author: Kevin Hill

Affiliations

1st Co-Author: Dept of Physics, Virginia Tech - The Wellcome Trust Centre for Neuroimaging,

University College of London

2nd Co-Author: Virginia Tech

3rd Co-Author: Virginia Tech

4th Co-Author: Virginia Tech

TITLE: Building multi-dimensional models of human choice

ABSTRACT: One goal of modern neural underpinnings of decision-making is to build a coherent

picture of how a human agent chooses so that predictions can be made about choices in novel

settings. These are still early days for such expectations. Hoever, using batteries of simple decision-

making tasks in large cohorts of subjects, we have begun to estimate multiple decision-making

parameters from single subjects with the idea of amalgamating these into a working model of simple

choice in humans. These multiple probes of single subjects, when combined with functional

neuroimaging measures has provided new possibilities for decision-making biomarkers potentially

useful in practical settings and perhaps useful in the characterization of psychopathologies. This talk

will outline our early work in this area.

27

Author

1st Author: Milica Mormann (Presenter)

2nd Co-Author: Blythe R. Towal

3rd Co-Author: Christof Koch

Affiliations

1st Author: School of Law, University of Miami, Coral Gables, FL

2nd Co-Author: Division of Biology, California Institute of Technology, Pasadena, CA

3rd Co-Author: Computation and Neural Systems, California Institute of Technology, Pasadena, CA

TITLE: Predicting Consumer Attention and Choice: Insights from Computational Neuroscience

ABSTRACT: Every year companies across the globe spend massive amounts of time and money on

eye-tracking studies and A/B testing of marketing materials. All the while, no systematic understanding

exists of how visual features, such as color or brightness, affect consumer attention, and no

recommendations are available for marketing practitioners on how to design attention-grabbing

materials. Further, there is no tool as yet to measure, much less predict, visual attention and its effects on

downstream marketing factors, such as consideration and choice.

We here introduce a class of neuro-computational models of visual attention to predict what attracts

consumers’ attention and use eye-tracking to explain how the deployment of said attention affects

consumer choices.

We present hungry subjects with high-resolution photos of store shelves filled with 28 typical snack

food items, such as chips and cookies. We investigate the extent to which subjects’ preferences, on one

hand, and visual features of the food items, on the other hand, drive subjects’ eye-movements as they

search for and choose a food item that they want to eat. The results show that consumers indeed choose

their most preferred items most of the time (liking-rating of 4 or 5 out of 5). However, among similarly

rated food items, subjects are more likely to choose those that are visually salient, i.e., whose packaging

stands out compared to that of the alternatives. Using eye-tracking with high temporal resolution, we

demonstrate three distinct mechanisms of visual attention that give rise to this effect of attention on

choice: subjects look earlier, more often, and longer at visually salient items than at visually less salient

alternatives even if the latter are similarly or, sometimes, even more liked.

Our results increase our understanding of how attention operates and how it affects decision making.

We also offer concrete benefits to the industry in its quest for guidance on attention management,

especially at the design and testing stages for marketing materials, including websites, product

packaging, magazine ads, video-game ads, and many more.

28

Author

1st Author: Uma R. Karmarkar (Presenter)

2nd

Co-Author: Baba Shiv

3rd

Co-Author: Rebecca M.C. Spencer

Affiliations

1st Author: Harvard Business School

2nd

Co-Author: Stanford Graduate School of Business

3rd

Co-Author: Department of Psychology, University of Massachusetts, Amherst

TITLE: “Sleep on it?” The effects of sleep on subjective preference-based choice

ABSTRACT: Conventional wisdom and studies of unconscious processing (1) suggest “sleeping on it”

improves decisions. Supporting this, recent work found that after sleep, performance improved on the

Iowa Gambling Task, a procedure thought to involve affect-guided, unconscious choice (2).

Additionally, sleep benefits processes intimately related to choice, such as memory and insight in

problem solving. Yet whether sleep has a meaningful impact on decision-related perceptions remains

unknown.

We examined the effects of sleep in a subjective decision context where the “optimal” outcome was

dependent on individual preferences. In session 1, participants were given positive and negative

attributes of four laptop satchels. In session 2, following a 12-hr interval containing overnight sleep or

daytime wake, participants made an incentive compatible choice between these satchels. Perceptions of

the satchels and memory for their attributes were measured in both sessions.

Sleep had a main effect on recall such that the Sleep group remembered more in session 2 (compared to

session 1) than the Wake group (F(1,53)=4.73, p<.05). We examined the amount of positive and

negative information recalled, defined by participants’ own valence ratings. Compared to Wake, the

Sleep group appeared to remember proportionally more positive than negative information

(F(1,53)=3.51, p=.067). In addition, liking for the preferred (eventually chosen) satchel showed a

marginal increase for the Sleep but not Wake group (F(1,53)=3.84, p=.055).

Despite these positive trends, sleep decreased the reported desire to purchase the preferred bag

(F(1,53)=9.383, p<.005). Furthermore, after choosing, individuals in the Sleep group reported

significantly lower perceived decision quality compared to the Wake group (F(1,53)=10.064; p<.004).

Specifically, they rated the choice as more difficult and were less confident and less satisfied with their

decision.

Sleep effects were found in comparison to a group who spent a waking interval separated from the

decision, suggesting an influence beyond unconscious deliberation (1). In addition, circadian controls

showed that time-of-day effects could not account for the changes observed.

Our results demonstrate that sleep has distinct and complex effects on the decision process. However,

despite the benefits of a more positive view of the chosen option, if the goal is to feel good about a

decision, we find some of the first evidence that “sleeping on it” can hurt more than it helps.

29

3rd annual Interdisciplinary Symposium on Decision Neuroscience

Abstracts

Session 5: Tools and Advances in Decision Neuroscience Do EEG Oscillations Predict Consumer Choice? Ale Smidts, Erasmus University Correlation and Causation: Understanding and Altering Decision-Making in the human brain Sameer Sheth, Columbia University Medical Center Using fNIR to study the effects of music on behavior David Gefen, Drexel University

30

Author

1st Author: Ale Smidts (Presenter)

2nd Co-Author: Maarten Boksem

Affiliations - Rotterdam School of Management, Erasmus University

Erasmus Center for Neuroeconomics

TITLE: Do EEG oscillations predict consumer choice?

ABSTRACT: The goal of this study is to assess whether EEG oscillations predict consumer preference

and choice, above and beyond self-report. We focus on frontal theta wave activity related to cognitive

processing (attention, learning and memory), parietal alpha activity related to arousal, and asymmetry in

frontal alpha activity related to approach and avoidance motivation.

In the experiment, participants watched 18 movietrailers of 2 minutes each while their brain activity was

recorded with EEG. Movies were selected from four genres (thriller, adventure, action and drama) and

differed in US box office success. Participants watched trailers of movies they had not seen before from

their three most favorite genres. Each trailer was subsequently rated on overall liking, PANAS, and

willingness to pay. After having watched all 18 trailers, participants were asked to categorize these

movies into three categories of liking (most liked, neutral, least liked). Finally, they were asked to pick

three DVDs to take home (buying behavior).

Preliminary results indicate that frontal theta wave activity and alpha activity predict both buying and

US box office. Moreover, this brain activity explains additional variance beyond self-reported liking,

suggesting that the brain signal contains unique information. Importantly, the brain response in the

neural focus group seems predictive of movie success in the general population. Further detailed results

will be presented at the symposium.

31

Author

1st Author: Shaun R. Patel

2nd

Co-Author: Sameer A. Sheth (Presenter)

3rd

Co-Author: Matt K. Mian

4th

Co-Author: Jimmy C. Yang

5th

Co-Author: Emad N. Eskandar

Affiliations 1

st Author: Department of Neurosurgery, Massachusetts General Hospital, Boston, MA

2nd

Co-Author: Columbia University Medical Center

3rd

Co-Author: Department of Neurosurgery, Massachusetts General Hospital, Boston, MA

4th

Co-Author: Department of Neurosurgery, Massachusetts General Hospital, Boston, MA

5th

Co-Author: Department of Neurosurgery, Massachusetts General Hospital, Boston, MA

TITLE: Correlation and Causation: Understanding and Altering Decision-Making in the Human Brain

ABSTRACT: Objective: Single neuronal computations underlying financial decision-making remain unclear.

We explored single neuronal activity from the nucleus accumbens (NAc) and subthalamic nucleus (STN) in

patients undergoing deep brain stimulation while actively engaged in a financial decision-making task. In a

subsequent experiment, we explored the effects of intermittent electrical stimulation of the STN on financial

decision-making.

Methods: Subjects are presented with a computerized gambling task modeled as a simplified version of the

card game “war”. The subject is dealt a card and asked to make a high or low wager ($5 or $20). Immediately

following their choice they are shown their opponent’s card–the player with the highest card wins. To simplify the

task, cards were limited to even cards of one suit from 2 through 10. We recorded 25 NAc and 20 STN neurons

from 13 patients undergoing deep brain stimulation for neuropsychiatric or movement disorders. In a subsequent

experiment, we applied intermittent stimulation to 15 subjects implanted with a DBS electrode (STN) while

engaged in the same financial decision-making task.

Results: We found single neuronal evidence of a behavioral and prediction error signal in both the NAc and

STN. The behavioral signal predicted–-during a discrete 500 ms interval–-their upcoming financial decision well

before they physically manifested their choice. Interestingly, we found that this activity continued to predict

upcoming decisions even when the stimulus was not a predictor of outcome; in other words, when there was a

50/50 chance of winning and losing. The prediction error signal-–the difference between expectation and

outcome–-again occurred during a discrete 500 ms interval, immediately after receiving feedback on the outcome

of the current trial. Activity in both structures encoded both positive and negative prediction errors. Furthermore,

in the second experiment, we were able to modulate their decisions (i.e. make them bet low) by applying 1 s of

electrical stimulation in the STN at the same period where we observed the behavioral signal; we found no

changes in behavior when stimulation was delivered at any other task epoch.

Conclusions: In conclusion, these findings demonstrate the ubiquitous nature of two reinforcement-learning

signals during financial-decision making: (i) binding predicted stimulus value to action, and (ii) evaluating

discrepancies between expectation and outcome. In addition, we demonstrate a causal relationship by applying

intermittent electrical stimulation to bias the decision signal and ultimately alter financial behavior.

32

Author

1st Author: David Gefen (Presenter)

2nd

Co-Author: Hasan Ayaz

3rd

Co-Author: Kurtulus Izzetoglu

4th

Co-Author: Banu Onaral

Affiliations

1st Author: LeBow College of Business - Drexel University

2nd

Co-Author: School of Biomedical Engineering, Science & Health Systems, Drexel University

3rd

Co-Author: School of Biomedical Engineering, Science & Health Systems, Drexel University

4th

Co-Author: School of Biomedical Engineering, Science & Health Systems, Drexel University

TITLE: Music, Neural Correlates, and Behavior in a Trust Game

ABSTRACT: Romeo apparently manipulated Juliet’s trust and judgment by playing music to her.

What music he played and what neural correlates this music may have had, and therefore why it may

have affected her trust is unknown. In fact, very little is known why music affects behavior and

assessments at all, except that certain types of music in certain situations do so. In this presentation we

report an experiment we performed to assess how two types of music affect behavior in a trust game and

its neural correlates in the prefrontal cortex as recorded with functional near infrared (fNIR)

spectroscopy. The results show that investment amount in the trust game decreased significantly when

subjects heard a classical allegretto moderato piano piece in F minor compared to there being no music

at all, and a parallel decrease in activity in the right dorsal region of the medial prefrontal cortex. This

effect was not evident when the subjects heard an allegro vivace piano piece in D major. What this may

indicate regarding trust theory will be discussed as well as some ideas about the kinds of music aspiring

Romeos should prefer.

33

3rd annual Interdisciplinary Symposium on Decision Neuroscience

Abstracts

Session: POSTERS

Presenter in the conference:

Jones, William J.

Lin, Meng-Hsien (Jenny)

Massoni, Sébastien

Minas, Randall K.

Shenhav, Amitai

Yoon, Haewon

Walden, Eric A.

Webb, Ryan

Wu, Yi-Chia

34

Author

1st Author: William J. Jones (Presenter)

2nd

Co-Author: Terry L. Childers

3rd

Co-Author: Yang Jiang

Affiliations

1st Author: University of South Dakota

2nd

Co-Author: Iowa State University

3rd

Co-Author: University of Kentucky

TITLE: Buying-related negativity in consumer choice

ABSTRACT: In this paper, we explore the role of response-locking of ERP during consumer choice.

Specifically, we focus on error-related negativity (ERN), a negative deflection within the ERP after a

subject indicates a response and usually within a feedback-related experimental paradigm. Although past

work has generally described ERN as a neural response to the commission of errors (Gehring et al.,

1990), other work suggests that ERN may be an indicator of a broader action monitoring system

(Gehring & Knight, 2000). We recorded data from 32 EEG channels while 39 consumers participated in

a modified version of a shopping task previously used by Jones, Childers, and Jiang (2012). Consumer

were shown a series of products with percentage promotions and asked to indicate whether they would

buy the offering or not. Prior to participating in the lab session, consumers also completed scales to

assess their math anxiety level and overall numeracy. In order to elicit an ERN-like response, accurate

feedback was provided to consumers after making a decision to buy or not. ERP was then epoched based

on consumers’ responses both as a function of their decisions to buy or not as well as the accuracy of

their decisions. ERPs were also created based on stimulus locking of correct versus incorrect feedback.

For comparison purposes to Jones, Childers, and Jiang (2012), ERPs were constructed for price

evaluation processes and during the prompt to make a buying decision. ERN like deflections maximal at

FCz were observed 0-200 ms after responses for buys relative to non-buys and for incorrect versus

correct decisions. Comparisons of effect size measures revealed more pronounced differences for buying

versus accuracy comparisons. Assessing ERN within separate 100 ms intervals revealed that effects of

math anxiety occur within the later interval and numeracy the earlier. The direction of these effects is

such that increased anxiety and numeracy were associated with increased ERN. Interpreting ERN, we

draw on Vermetten and Bremner (2002) who proposed that interactions between the amygdala and the

extrapyramidal motor system might precipitate responses to threat.

35

Author

1st Author: Meng-Hsien (Jenny) Lin (Presenter)

2nd

Co-Author: Samantha N. N. Cross

3rd

Co-Author: Terry L. Childers

Affiliations - Iowa State University

TITLE: Olfactory Imagery and Emotions: Neuroscientific Evidence

ABSTRACT: Introduction: Olfactory imagery is defined as “being able to experience the sensation of

smell when an appropriate stimulus is absent” (Stevenson and Case (2005). At times, odors associated

with a product are not accessible in the marketplace. For example, packaged products are shown visually

through pictures or ads or are sold online. Our study investigates the impact of visual information in

stimulating other sensations such as olfaction. The literature on mental imagery shows that odor imagery

can be processed in the brain similar to the sensory processing of actual stimuli (Djordjevic et al 2005).

There also is a close relationship between the sense of smell and emotions (Aggleton & Mishkin 1986).

However, what has not been explicitly examined is whether odor imagery can affect the emotions. We

examine this issue in our study.

Methods: An Event-Related-Potential (ERP) experiment was designed to detect emotions using

encephalography (EEG) methods. The design is a 3 (odor-associated valence: neutral vs. pleasant vs.

unpleasant) × 2 (tasks: passive view vs. olfactory mental imagery) within subject design. A pool of 90

pictures was constructed and pretested; including non-odor associated pictures as well as pictures of

pleasant or unpleasant odor associated items. Pictures are presented for 1500ms with data collected from

20 participants. Late Positive Potential (LPP) at Pz in the window of 500 to 700ms was analyzed.

Results and Conclusions: Unpleasant odor-associated pictures generated a larger LPP compared to

pleasant and neutral in both view and imagery tasks. An interaction between visual and olfactory

information on emotion processing moderated the valence of the stimuli. In the pleasant odor pictures,

imagery generates a larger LPP than in the view task. For unpleasant odor pictures, the view condition

generated a stronger LPP compared to the imagery condition. This suggests olfactory imagery can

enhance emotion processing of positive odor-associated pictures. In comparison, visual processing may

dominate the emotional processing of unpleasant odor-associated pictures. In a follow-up study in

process, we are investigating the impact of individual differences in sense of smell on olfactory imagery.

We expect differences in ability to smell will influence the magnitude of emotions elicited during

olfactory imagery.

36

Author

1st Author: Sébastien Massoni (Presenter)

2nd

Co-Author: Nicolas Roux

Affiliations - CES – University of Paris 1 & Paris School of Economics

TITLE: Optimal Group Decision: A Matter of Confidence Calibration

ABSTRACT: The failure of groups to make optimal decisions is an important topic in human sciences.

Recently this issue has been studied in perceptual settings where the problem could be reduced to the

question of an optimal integration of multiple signals. The main result of these studies asserts that

inefficiencies in group decisions increase with the heterogeneity of its members in terms of

performances. By using a suboptimal decision rule groups overweight the recommendations of the least

able member. On the contrary, we propose to relate those results to biases in subjects' confidence

calibrations. We assume that subjects' beliefs about their perceptive abilities are initially not related to

their actual perceptive abilities. If everyone holds similar beliefs about his performances, the most able

subjects tend to be relatively underconfident as compared to the least able subjects. Consequently, a

group will put too much weight on the least able member, so that heterogeneity induces greater

collective inefficiencies. Therefore our explanation of collective inefficiencies does not rely on the

incapacity of humans to aggregate heterogeneous information. We rather see them as an inevitable

consequence of the lack of information subjects have access to. We perform a perceptual task in which

we elicit subjects' confidence at each trial. Each subject of a dyad observes the stimuli in isolation and

answer individually (choice and confidence). Then the two group members must reach an agreement on

choice and confidence. After the group decision is made, subjects answer anew to check if they agree

with group decision. Finally they observe whether they were right or wrong. Our results give evidence

that predicting the performance of a group is improved by taking into account its members' confidence

in their own reliability. First, the heterogeneity of performances has no more impact on the failure of

group decisions when we take into account the heterogeneity of confidence calibration. Second, a

models comparison gives support to our model based on beliefs rather than the model based on

performances heterogeneity. This experiment allows us to emphasize the importance of the

metacognitive ability in optimal group decisions. Our work is focused on the calibration question but it

could be extend to the discrimination ability and then offers some new perspective of works in terms of

group decisions, social interactions and metacognition.

37

Author

1st Author: Randall K. Minas (Presenter)

2nd Co-Author: Robert F. Potter

3rd Co-Author: Alan R. Dennis

4th Co-Author: Valerie Bartelt

5th Co-Author: Soyoung Bae

Affiliations

1st Author: Operations and Decision Technologies Department, Indiana University

2nd Co-Author: Department of Telecommunications, Indiana University

3rd Co-Author: Operations and Decision Technologies Department, Indiana University

4th Co-Author: Department of Information Systems, Texas A&M University-Kingsville

5th Co-Author: Department of Telecommunications, Indiana University

TITLE: Overloaded or biased? Using EEG and NeuroIS to elucidate information processing in virtual teams

ABSTRACT: Virtual teams are common in today’s organizations, yet they often make poor decisions. Virtual

teams use a variety of collaboration tools to interact (e.g., instant messaging, video conferencing). Teams that use

text-based tools, such as chat rooms, usually share more information than their face-to-face counterparts, but

typically reach decisions that are no better because team members do not seem to act upon new information

provided by others. There are at least two theoretical explanations for why virtual team members ignore the

information from others. One is that team members are so overloaded with information that they reach a point at

which some information is missed, which results in members ignoring new information they received from others.

A second reason why decision making is no better in the information-rich virtual environment is that team

members engage in biased information processing that selectively focuses on information confirming their

personal preferences rather than seeking unbiased information.

This research aims to understand which of these theoretical mechanisms occurs in small virtual team

interactions. Conventional research methods have been unable to disentangle whether the lack of processing of

new information is due to information overload or confirmation bias. The use of neurophysiological and

psychophysiological methods provides an unique opportunity to better understand individual information

processing in team settings. This experiment uses a 14-channel electroencephalography (EEG) system to examine

alpha activation near the surface of cortex, as well as electrodermal activity (EDA), and facial electromyography

(EMG). A typical hidden-profile task was used, wherein the participant is asked to make an individual decision

based on incomplete information before interacting with the team. During the team interaction (which was done

using a simulator), the participant was exposed to new information and the opinions of other team members that

either challenged or supported the participant’s pre-decision.

Our findings indicate differential alpha activity in processing of information that confirms an individual’s pre-

decision preference, while information challenging an individual’s pre-decision preference is processed similarly

to irrelevant information. We observed differential alpha activation between information that confirmed and

challenged pre-discussion preferences in frontal and temporal cortices, which are implicated in both working

memory and assimilation of new information. These findings suggest the presence of confirmation bias in

information processing during virtual discussions.

38

Author

1st Author: Amitai Shenhav (Presenter)

2nd

Co-Author: Randy L. Buckner

Affiliations

1st Author: Dept of Psychology, Harvard University; Princeton Neuroscience Institute, Princeton

University

2nd

Co-Author: Dept of Psychology, Harvard University; Dept of Radiology, Dept of Psychiatry, MGH,

Harvard Medical School; Howard Hughes Medical Institute

TITLE: Neural Correlates of Dueling Affective Reactions to “Win-Win” Choices

ABSTRACT: Background: When offered a choice between multiple highly desired options, we can

simultaneously feel pleasure and anxiety. The sources of these dueling subjective experiences are still

poorly understood, as are the correlates of decision and reevaluation processes for the kinds of choices

that give rise to such feelings – namely, choices that typically require trade-offs between multi-attribute

options of quantitatively similar but qualitatively different value (e.g., between favorite desserts). We

therefore explored how activity in different neural circuits during such choices related to both subjective

and objective aspects of the decision and reevaluation process.

Methods: Participants in two experiments (N= 84) were scanned while making incentivized time-

pressured choices between products. Participants chose between products they had valued similarly low

(Low-Low), similarly high (High-High), or dissimilarly (Low-High). They retrospectively rated positive

and anxious experiences for each choice pair, and were offered an (unanticipated) opportunity to

reevaluate earlier choices.

Results: Participants reported feeling both most positive and most anxious during High-High

choices. Crucially, we found that dissociable networks tracked these two competing experiences.

Ventral versus dorsal regions of ACC and striatum differentially tracked positive and anxious states,

consistent with their proposed roles in evaluating potential rewards versus evaluating the response costs

required to obtain them (e.g., choice conflict). Activity in dorsal ACC also predicted post-scan choice

reversals (i.e., changes of mind), as did bilateral rostrolateral PFC. Surprisingly, an additional

dissociation was found within ventral mPFC. Whereas rostral ACC was most sensitive to the positive

experience associated with choice (and therefore the value of the best option), medial OFC was most

sensitive to whether participants were making the most difficult kind of choice (High-High) but did not

otherwise differentiate between chosen, relative, or total values.

Conclusions: Our results suggest that separate neural systems that assess the value of options versus

the cost of choosing amongst them may be to blame for an experience of win-win choices that is

simultaneously positive and stressful. Moreover, within vmPFC we find evidence for dissociable regions

that are differently sensitive to the value and positive experience associated with choosing (rACC)

versus a potential interaction between the value and integrative demands of one’s choice (mOFC).

39

Author 1

st Author: Haewon Yoon (Presenter)

2nd

Co-Author: Gretchen B. Chapman

Affiliations - Department of Psychology, Rutgers University

TITLE: Exploring qualitative differences in intertemporal choice models

ABSTRACT: In intertemporal choice research, Mazur (1987) proposed the hyperbolic discounting

function, which can account for dynamic inconsistency (Thaler, 1981) -- a preference reversal whereby

the agent initially prefers the larger later reward but later changes to prefer the smaller sooner reward.

Since then, researchers have proposed several different alternative discounting functions based on new

empirical findings. However, the hyperbolic discounting function is still the most widely accepted

model by many researchers. Why is that? We suspect that model comparisons have not convinced other

researchers enough because they are simply showing the quantitative differences between alternative

models, using curve fitting to see which model best fits the data. Unfortunately, a curve fit does not

provide a decisive conclusion about whether the model is qualitatively correct or not. It only tells which

function fits the data better than the other.

To investigate the qualitative predictions of different discounting models of intertemporal choice, we

have developed a new modeling framework (iPRP: intertemporal preference reversal prediction) that can

analyze the preference reversal patterns across different delays, reward magnitudes, and individual

discount rates based on specific discounting models. From this modeling framework, we found that for a

theoretical agent with a high discount rate and a hyperbolic discounting function, the proportion of

consistently choosing a smaller sooner option is capped at 50% of the objectively defined parameter

space. This prediction is counterintuitive because one would expect increasing preference for the smaller

sooner option as a function of a discount rate. We evaluated an alternative discounting model that does

not make this problematic prediction and is still capable of demonstrating dynamic inconsistency, which

is the essence of the hyperbolic discounting function. This alternative model uses exponential

discounting with subjective time perception.

We conducted an experiment that focused on the qualitative differences between these two models.

The experiment supported the alternative model. Roughly 1/4 of participants consistently chose the

smaller sooner option in more than 50% on the parameter space. In addition, participants showed the

dynamic inconsistency phenomenon. These results contradicted to the exponential discount function

(due to the presence of the dynamic inconsistency) and the hyperbolic discounting function (because

participants with high discount rates selected the smaller sooner option in more than 50% of the

parameter space).

40

Author

1st Author: Glenn J. Browne

2nd

Co-Author: Eric A. Walden (Presenter)

Affiliations - Rawls College of Business Texas Tech Univeristy

TITLE: Information Search and Stopping: A NeuroIS Investigation

ABSTRACT: Information search is now a ubiquitous aspect of users’ interactions with information

systems and a fundamental first step in their decision-making processes. Prior research has investigated

cognitive stopping heuristics, or rules, in various types of search (e.g., Browne and Pitts 2004; Browne,

Pitts, and Wetherbe 2007; Ho, Bodoff, and Tam 2011). The purpose of the present research is to

investigate information search and stopping using functional Magnetic Resonance Imaging (fMRI) data.

Subjects were 21 students from a business school subject pool at a large university in the

southwestern U.S. The experimental task required a subject to search for information about three

products individually within a product class, such as televisions, and this was repeated for seven product

classes. When subjects indicated they had gathered sufficient information (that is, they indicated they

wanted to stop), they were given a question about why they chose to stop. This type of design is

different from typical inhibition studies in neuroscience, such as stop-signal and go/no-go tasks, because

the subjects rather than the researcher controlled the decision to stop.

Results revealed that a number of brain networks showed significantly higher blood-oxygen-level-

dependent (BOLD) activation during stopping than during searching. The first was an executive control

network comprised of the dorsolateral pre-frontal cortex and the anterior cingulate cortex. Together the

greater activation of these two areas suggests that subjects were actively employing top down executive

control to attend to the stop decision and evaluate the value of a stop response.

Second, the insula and the caudate/putamen were more active on both sides of the brain during

stopping than during searching. This is a typical network of activation in stop-signal tasks, and is

indicative of inhibition, which represents the behavior of stopping. These areas inhibit the urge for a

person to continue with what he is doing, which suggests that he will stop.

We also found activation in the supramarginal gyrus near Broca’s area. In part this is because our

stimuli are verbal, but we believe it is higher during stopping that searching because subjects are

mentally repeating the verbal stimuli to feed forward the total verbal information into the executive

control networks.

Together, these areas of activation provide strong evidence that different areas of the brain are

involved in searching and stopping. Implications for decision making more generally will be discussed.

41

Author

1st Author: Ryan Webb

Affiliations

1st Author: Center for Neuroeconomics, New York University

TITLE: Neural stochasticity begets drift diffusion begets random utility: a foundation for the

distribution of stochastic choice

ABSTRACT: The random utility model has become the standard empirical framework for modelling

stochastic behaviour in applied settings, and the neural random utility model extends this framework to

the neural decision process. In this framework, the distribution of random utility has important

implications for both testing behavioural theories and predicting behaviour. Moreover, the random

utility framework has been agnostic about the dynamics of the decision process that are of considerable

interest in psychology and neuroscience, in particular bounded accumulation models which relate

decision response times to stochastic behaviour. We demonstrate that the random utility model can be

derived from a bounded accumulation model, of which the classic drift diffusion model is a special case,

with the features of this dynamic process impacting the distributional structure of random utility. This

clarifies that the random utility model is essentially a reduced form of a dynamic neural model, and

proposes how the large literature on the neural implementation of bounded accumulation frameworks

can offer advances in modelling choice at the level of random utility.

42

Author

1st Author: Yi-Chia Wu (Presenter)

2nd

Co-Author: Michael S. Minor, Ph.D

3rd

Co-Author: Monica Hernandez, Ph.D

Affiliations

1st Author: University of Texas-Pan American

2nd

Co-Author: University of Texas-Pan American

3rd

Co-Author: Texas A&M University-Corpus Christi

TITLE: Comparing Apple and Religion: Perspectives Related to Brainwaves

ABSTRACT: This research examines how the Apple brand has transformed its relationship with users,

possibly equivalent to a religion for some, becoming part of an individual’s life. Religion enables human

beings to create meaning for their existence (Muniz and Schau 2005). In a fashion similar to religion,

material objects can be embedded with meaning and interpreted by consumers as validating their

existence. Material consumption can also formulate and create meaning through objects that come to

symbolize one’s perspective. It is therefore reasonable to attempt to understand material consumption

through the lens of religion (Muniz and Schau 2005).

This paper is in the data collection process. Utilizing a 2×2 factorial design, we identify two

classification factors: Apple and religion. Participants will be recruited based on their experience: Apple

enthusiasts (high vs. low) and religious enthusiasts (high vs. low), based on self-reports. Subjects will be

recruited from a southwestern university. All the participants will see randomly assigned picture stimuli

associated with Apple and religion within subjects. Each of the 22-picture stimuli appear for six seconds

and are followed by a six-second break. Electrode recording will be collected by ProComp 2 to enable

measuring and analyzing EEG with BioGraph Infinite software from Thought Technology. After the

experiment, a short survey will be conducted to triangulate the participant’s response in the experiment.

We will use factorial ANOVA to compare cell means. Ten PhD and MBA student volunteers have

participated in a pretest, the purpose of which was to explore whether the picture stimuli were

meaningful to the subjects and to test the short survey. The next step is the main test.

Neurologists have explained their process of dealing with religion by explaining how human neurons

react in order to process religious sentiments (Aamodt and Wang 2008). This study intends to find

whether a brand can develop a similar meaning to an individual as religion denotes in one’s life and

whether brand enthusiasm rises to the level of a spiritual state. The managerial implications lie on the

possibility that the Apple brand has come to represent a symbol of comfort similar to that of religious

belief. This study would shed light on the opportunities of developing a brand to that stage, beyond the

mere relationship of product and user.

3rd

ANNUAL INTERDISCIPLINARY SYMPOSIUM

ON DECISION NEUROSCIENCE

We would like to thank all sponsors!

Gold Sponsors:

Silver Sponsors

Bronze Sponsors

Supporters

47