University of Southern California Dissertations and Theses

Visual and audio priming of emotional stimuli and their relationship to intertemporal preference shifts

(USC Thesis Other)

Visual and audio priming of emotional stimuli and their relationship to intertemporal preference shifts

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS

Visual And Audio Priming Of Emotional Stimuli And
Their Relationship To Intertemporal Preference Shifts
By
Andrew James Melrose

Thesis Submitted in Partial Fulfillment
of the Requirements for the Degree of
Masters of Arts
In the Department of Psychology

University of Southern California
August 2015

PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 2
Abstract
There is considerable evidence that when affect or emotions are experimentally
manipulated it can lead to systematic preference shifts in intertemporal choice. Through
two experiments, the current work sought to replicate and extend past works by
developing a new paradigm, and investigating potential mediating variables. Experiment
1 sought to replicate and simplify a prior study using happy, fearful and neutral face
primes prior to intertemporal choice. Results indicated that with our procedure, both
happy and fearful face primes increased impulsive choice. Experiment 2 extended prior
research by developing a novel paradigm utilizing positively and negatively valenced
auditory affective primes prior to intertemporal choice. Results indicate that similar to
Experiment 1, both pleasant and unpleasant auditory primes increased impulsive choice,
though only unpleasant auditory primes met significance. Results further implicate that
intertemporal preference shifts do not appear to be mediated by either trial-by-trial
variation in sympathetic nervous system activity or fixation-based visual attention.
Together, these experiments add to the growing literature on the role of emotion and
affect during delay discounting, and also contributes to the current understanding of
underlying mechanisms by which affective primes may be manipulating preferences.
Keywords: Delay Discounting, Emotion, Affect, Eye-Tracking

PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS
Table of Contents
Abstract…………………………………………………………………………..…..2
Introduction…………………………………………………………………..………4
Experiment 1…………………………………………………………………………9
Methods……………………………………………………………………..10
Analysis……………………………………………………………………..13
Results………………………………………………………………………13
Discussion…………………………………………………………………..15
Experiment 2………………………………………………………………………..15
Methods……………………………………………………………………..18
Analysis……………………………………………………………………..23
Results………………………………………………………………………25
Discussion…………………………………………………………………..31
General Discussion………………………………………………………………....31
References…….…………………………………………………………………....35

PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 4
Introduction
Relative to other species, humans are exceptional in their ability to forego smaller
immediate rewards in exchange for larger delayed rewards. Where such behavior appears
to exist on similar scales in other species, such as in squirrels’ burying nuts that they later
dig up when food is scarce, the behavior has been shown to be the product of specialized
instincts that make the behavior immediately gratifying. Humans can, without domain-
specific instincts, wait weeks, months or even years for a large enough delayed reward.
Nevertheless, people are not perfect in their capacity to delay gratification. Even humans
have a nearly universal tendency to “discount” options that are delayed (referred to as
“delay discounting”), often opting for the smaller of two rewards depending on the size
and delay of the larger reward. The degree of this tendency varies greatly across
individuals, but tends to be stable within an individual, as evidenced by high test-retest
reliability with an intervening interval of one year (Kirby, 2009). However, despite a
relative stability of discounting, it has been noted that these preferences are susceptible to
visceral factors (G. Loewenstein, 1996), which is the topic of the current work.
Monetary intertemporal choice is undoubtedly the most frequently used
experimental assessment of delay discounting. In an intertemporal choice trial, the
individual must decide between a smaller but more immediate monetary reward (the
smaller sooner reward, or SS) or a larger but more temporally distant monetary reward
(the larger later reward, or LL). For example, the participant may be asked whether they
prefer $25 today (SS), or $30 in a week (LL). Should the individual decide they would
prefer to receive the $25 today, they are implicitly saying the cost of waiting 7 days is not
worth an increase of $5 in net pay. This behavior has a long and extensive history of
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 5
investigation spanning a wide range of academic disciplines (e.g., Ainslie, 1975;
Samuelson, 1937) and as such, many models of discounting behavior have been proposed
(for overview see Doyle, 2013). Currently, the most commonly utilized model within
experimental psychology and behavioral economics is the hyperbolic model of delay
discounting (see equation 1 below). The prominence of this model is likely due to its
ability to partially account for preference reversals with the passage of time, and its
ability to be reflected by a single free-parameter (Kirby, Petry, & Bickel, 1999).
In contrast to axiomatic rationality and utility maximization of neoclassical
economics, behavioral economic models of choice incorporate effects of emotions on
decision-making both generally, and specifically within the intertemporal choice
literature (e.g., Loewenstein & Lerner, 2003). Recognition of the importance of emotion
and affect in judgment and decision-making has led to the development of several models
attempting to describe both why and how emotions are used to inform decision-making
(e.g. the Somatic Marker Hypothesis (Bechara & Damasio, 2005), and the Affect-As-
Information hypothesis (Schwarz & Clore, 1983)). With some models forming
predictions based on a general maladaptive influence of incidental emotions (e.g. Bechara
& Damasio, 2005), and others making emotion-specific predictions (e.g., Clore &
Huntsinger, 2007). Though across these models there seems to be a consensus that
emotions and affect can and do influence preferences during decision-making, the
manner in which different emotions and affective states exert their force on choice
remains very much an open question.
Several studies to date have utilized either generally appetitive or aversive primes
to manipulate intertemporal choice preferences (though no single study to the author’s
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 6
knowledge has attempted to manipulate both). Among studies utilizing appetitive priming
stimuli, pleasant food-cues (Li, 2008), attractive faces (Wilson & Daly, 2004), sexually
appealing stimuli (Van den Bergh, Dewitte, & Warlop, 2008), and appealing brand logos
(Murawski, Harris, Bode, Dominguez, & Egan, 2012) have all been shown to increase
impulsive intertemporal choice preference. In contrast, Tuk, Trampe, & Warlop, 2011
used increased bladder pressure (i.e. need to pee) as an aversive cue and results indicated
a decreased preference for impulsive intertemporal choice. These studies all evince that
the appetitive and aversive states being experimentally manipulated is consistently being
used to inform preferences. These works seem to support the notion that valence of affect
or emotion may be predictive of preference shifts, with pleasant primes increasing
impulsive choice, and aversive primes decreasing impulsive choice, though a mechanistic
model of how affective information influences intertemporal preferences runs into
problems if predictions are to be made based on the valence of emotions alone. Issues
arise specifically due to the heterogeneity of antecedent appraisals, facial expressions,
autonomic physiology and neural response across specific emotions within valence which
have all been implicated as being relevant to decision-making (Lerner & Keltner, 2000),
making a valence-based mechanism for preference shifts conceptually unlikely.
The influences of emotion-specific cue primes on intertemporal choice have an
even less straightforward interpretation, and further argue against a valence-based model
for affect’s role in delay discounting. Results are mixed following experimental
manipulations of happiness and positive affect, with results showing both increased
(Ifcher & Zarghamee, 2011), as well as decreased impulsive choice (Liu, Feng, Chen, &
Li, 2013; Pyone & Isen, 2011). In Lerner, Li, & Weber, 2012 two separate negative
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 7
emotions were manipulated (sadness and disgust) prior to intertemporal choice with
results indicating sadness related increases in impulsive choice whereas there was no
preference shift following disgust. Similar to positively valenced emotion cue primes,
these findings again highlight divergent results within negatively valenced primes. These
studies hint that the valence of the emotional manipulation will likely not lead to a
predictable systematic shift in delay discounting preference, implying a different
theoretical model may be needed.
Most relevant to the current work, two studies to date have attempted to
manipulate the affective state of the individual in a within subject design on a trial-by-
trial basis. In Raeva, Mittone, & Scharzbach, 2010, regret and rejoice were
experimentally manipulated via the results of a gamble made by the participant prior to
intertemporal choice. Results indicated that rejoice was associated with decreased
impulsive choice, whereas regret was associated with increased impulsive choice. It is
important to note, however, that these results may be adulterated by gamble outcomes
manipulating risk preferences, which may be associated with the attractiveness of the
delayed option (Patak & Reynolds, 2007). Using a much cleaner experimental emotion-
cue manipulation, Luo, Ainslie, & Monterosso, 2014 sought to invoke incidental affect
by having participants remember happy, fearful or neutral computer-generated faces
during the course of intertemporal choice trials. Results indicated that fear was associated
with decreased impulsive choice when compared to happy face primes, though neither
affective condition was shown to statistically differ from neutral primes.
Past works combining affective-cues and intertemporal choice have focused
almost exclusively on characterizing behavioral effects (Ifcher & Zarghamee, 2011;
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 8
Lerner et al., 2012; Li, 2008; Liu et al., 2013; Murawski et al., 2012; Pyone & Isen, 2011;
Raeva et al., 2010; Tuk et al., 2011; Van den Bergh et al., 2008; Wilson & Daly, 2004),
neural effects (Shan Luo et al., 2014; Murawski et al., 2012), and interactions between
states and traits of the individual (Augustine & Larsen, 2011; Hirsch, Guindon, Marisano,
& Peterson, 2010) leaving many unanswered questions. This literature paints a picture of
divergent results if a theoretical framework is to be drawn either in an emotion-specific,
or in a valence based way. Utilizing prior work (Gray, 1999) and preliminary data, the
Somatic Marker Hypothesis posits that the presence of incidental affect or emotion
during decision-making “shifts decisions in the direction of short-term goals” regardless
of affective valence or emotion via the interactions between the brain (specifically the
ventral medial prefrontal cortex (vmPFC), somatosensory regions, and the amygdala) and
the body through the autonomic nervous system (Bechara & Damasio, 2005). This
conceptual framework builds off focal brain lesion studies indicating that time is
represented in the vmPFC moving from the near future being represented in the posterior
regions towards the more distant future being represented in the anterior portions, which
may explain why immediacy may have a stronger relationship to somatic responses. This
theoretical model, which is partially adopted for the current work, would predict that
incidental emotions will disrupt intertemporal choice preferences, and will push the
individual towards the immediate option in a valence independent manner. Furthermore,
this model predicts that this shift towards short-term goals will likely be associated with
the strength of emotion-related markers of autonomic arousal.

PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 9
Experiment 1
In what is primarily intended as a replication attempt of the behavioral results
seen in Luo et al., 2014, we sought to further characterize the effects of different
incidental emotional face primes on intertemporal choice. There were two key differences
between this study and Luo et al 2014. First, while the prior study utilized computer-
generated faces (all male faces and, due to a limitation of the software, all without hair),
here we used real faces of mixed gender to reduce any potential differential responses to
emotional faces determined by congruency to sexual preference (Kranz & Ishai, 2006).
In addition to concerns regarding possible “bald male” idiosyncrasies, we reasoned that
computer generated faces may have been less evocative. Neural differences in response
to computer-generated and human faces indicate that despite similar network
involvement, computer-generated faces lead to reduced activity in key emotion and face-
relevant brain regions (Moser et al., 2007). Finally, in Luo et al., 2014 working memory
was manipulated by having participants remember the level of emotion in the prime
throughout the intertemporal choice set, and correctly answer whether a second face
image presented after their choice matched the emotion displayed in the prime. Though
this is a clever way to ensure processing of the emotion displayed in the face, subjects
were forced to engage in a working memory task during the decision making process,
which may have been implicitly easier for neutral trials, and may have dampened any
implicit effects of the primes. Indeed, working memory manipulations during
intertemporal choice have been associated with increased preference for the smaller
sooner option (Hinson, Jameson, & Whitney, 2003). If the individual is being pushed
towards the smaller sooner option in all conditions of the task via the working memory
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 10
component, it may obscure effects that emotional primes have, independent of memory
load. These alterations considered, hypothesis for Experiment 1 were as follows:
Hypothesis 1. Replication of behavioral effects in Luo et al. 2014, with fearful
faces increasing LL preference, and happy faces increasing SS preference.
Hypothesis 2. The simplified design will result in both emotion-cue conditions
differing from neutral, in addition to differing from each other.
Methods
Participants. Participants included a total of 38 undergraduate and graduate
students attending the University of Southern California (27 female). Of the originally
recruited 38 subjects, 8 were excluded for either having a lack of discounting (choosing
the delayed option when it had the same dollar value as the immediate option at least
once), or not maintaining a relative proportion of immediate/delayed choices between 25-
75% during the neutral condition (indicating improper titration of intertemporal choice
offers). Of the retained 30 subjects, 23 were female with a mean age in the entire sample
of 21.5±4.25. Informed written consent was obtained from all participants before
completion of the experiment. Participants completed the task in exchange for both
course credit and the carrying out of a randomly selected intertemporal choice trial to
ensure choices were truly representative of the preferences of the subjects.
Monetary Choice Questionnaire. Prior to completion of the experimental task of
interest, all participants were required to complete the Monetary Choice Questionnaire
(Kirby et al., 1999) in order to ascertain an individualized estimate of delay-discounting.
This questionnaire has a fixed set of 27 choices between immediate and temporally
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 11
delayed but larger monetary rewards. Responses to the set of 27 options were fit to the
following hyperbolic function:
Equation 1: 𝑉=
!
!!!"

Where V represents the current value of the option, A is the delayed monetary amount
offered on the screen, D is the temporal delay, and k is a free parameter representing level
of hyperbolic discounting. As k increases, the individual discounts the future more
steeply, and this parameter is fit based on the pattern of responses in the Monetary Choice
Questionnaire.
Face Prime Intertemporal Choice Paradigm. Following completion of the
Monetary Choice Questionnaire, participants completed a face-prime intertemporal
choice paradigm similar to that used in prior research (Luo et al., 2014). The estimated k-
value from the Monetary Choice Questionnaire informed the presentation of alternatives
in which the SS and LL options were similarly attractive. This greatly increases
sensitivity to hypothesized experimental effects (e.g., if a participant is presented with an
alternative pair in which, for him, the LL option is far more appealing, then even if an
emotion prime increases his discounting it is unlikely to cause a change in choice.)
As shown in the visual representation of the experimental paradigm in Figure 1.,
all trials began with a white fixation cross at the center of a dark gray screen lasting
between 1 and 5 seconds with a mean of 2. Next, either a neutral, happy or fearful
emotional face from the NimStim facial stimulus set (Tottenham et al., 2009) was
displayed for 2 seconds. The NimStim facial stimulus set is comprised of 17 actors from
4 different races (African American, Asian, Hispanic, & White) who were trained to
perform 3 different facial expressions (Fear - open mouth, Neutral - closed mouth, Happy
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 12
- slightly open mouth). Face primes were again followed by a fixation cross lasting
between 1.5-3 seconds with a mean of 2 seconds. Participants were then given 15 seconds
to respond to an intertemporal choice set that was designed to have both the immediate
and delayed option roughly equal in present value based on pre-determined k-values.
Values of the SS option (which were always offered today) were calculated based on
previously determined hyperbolic discounting, with delay varying between 20-66 days,
and delayed value varying between $14-57. As the task progressed, individualized k-
values were titrated by 1/8 log steps based on the revealed preferences similar to previous
studies (Luo, Ainslie, Giragosian, & Monterosso, 2009) following only the neutral trials
to avoid systematic biases in preference stemming from emotional face primes. This
ensured that participants would be near 50% in their distribution of choices during the
neutral condition (though not necessarily in the other conditions). Participants completed
48 total trials, with happy, fearful and neutral face primes all presented 16 times each in a
pseudorandomized order.

Figure 1. Visual depiction of a single trial of the face prime intertemporal choice task.
Following a brief fixation cross of 1-5 seconds, either a neutral, fearful or happy face was
displayed. Following another brief fixation cross of 1.5-3 seconds, individuals were
presented with a ‘hard’ intertemporal choice set. Following choice, the selected option
was highlighted for 0.5 seconds and the next trial began. Individuals were given a
maximum of 15 seconds to respond to the choice set.
1.5$3&sec&
2&sec&
0.5&sec&
RT&
1$5&sec&
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 13
Behavioral Analysis
Behavioral data was analyzed using multilevel generalized linear mixed effects
models with the lme4 package in R statistical software via the glmer and anova functions.
This multilevel structure takes a ‘leave one out’ strategy in which two generalized linear
mixed models are created, one without the predictor of interest, the other with the
predictor of interest (one model fully nested within the other). An ANOVA is then
preformed on the log likelihood estimate of each regression, using a 𝜒
!
distribution with
degrees of freedom (DOF) equal to the number of increased DOF in the expanded model
minus one. Due to the uneven gender distribution within the sample (of participants, not
of affective stimuli), participant gender was controlled in all analysis by including gender
as a predictor in both models to at least attempt to capture gender related variance in
choice. In addition, due to the necessity of using a titrating procedure during the task, the
1/8 log step alterations in k-values may have created systematic variations in preference
due to these slight alterations in presented offers. To account for this potential bias, the
log of each trial’s k-value was also controlled for and was allowed to vary as a random
slope. All multilevel glmer models used an Adaptive Gauss-Hermite Quadrature to
estimate the integral.
Results
Emotional face prime condition affected intertemporal preference (𝜒
!
(2) = 7.62,
p=0.022) with a regression coefficient of -0.369 ± 0.145 (standard error) for happy face
primes (z=-2.55, p=0.011) corresponding to an odds ratio for choosing the LL option
following happy primes of 0.692 when compared to neutral face primes, and a regression
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 14
coefficient of -0.317 ± 0.143 (standard error) for fearful face primes (z=-2.21,p=0.027)
corresponding to an odds ratio of 0.729 when compared to neutral faces (Figure 2).

Figure 2. Happy and fearful face primes both significantly increased impulsive
intertemporal choices. a. Beta values for emotional prime conditions with their estimated
95% confidence intervals with the zero point representing neutral face primes indicating
that both happy and fearful primes significantly differed from neutral face primes. b. The
predicted probability of choosing the larger later option was 0.464, 0.387, 0.398 for the
neutral, happy and fearful face primes respectively. Error bars shown correspond to upper
and lower quartiles with stars indicating significance.

−0.50
−0.25
0.00
Fear Happy
Beta Value
Happy and Fearful Primes Significantly Increase Impulsive Choice
0.3
0.4
0.5
0.6
fear happy neutral
Predicted Probability
Predicted Probability of choosing the LL option following face primes
*"
*"
a.
b.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 15
Discussion
Partially in line with Hypothesis 1, the happy face primes resulted in increased
choice of the SS option. However, contrary to this hypothesis, fearful face primes
(relative to neutral primes) also increased SS choice (similar to the happy face primes) in
contrast to increased LL choice as seen in Luo et al., 2014. In line with Hypothesis 2, the
preference shifts were increased to a point where happy and fearful conditions were
significantly different from neutral. Contrary to Hypothesis 2, the two prime conditions
were not significantly different from one another as was seen in Luo et al., 2014, likely
due to both affective conditions leading to similar preference shifts towards the impulsive
option. These results, though divergent from previous findings, are directly in line with
the predictions made by the Somatic Marker Hypothesis in that the affective cues
instigated a valence-independent shift towards preference for short-term gains.
Experiment 2
Building on Experiment 1, Experiment 2 utilized positively valenced, negative
valenced, and neutral affective sound primes prior to intertemporal choice on a trial-by-
trail basis. In Experiment 1, as in Luo et al., 2014, intertemporal preference shifts were
experimentally manipulated by the presentation of emotional face stimuli. By instead
utilizing affective sound stimuli, Experiment 2 sought to address whether increased
impulsive choice following facial affective primes seen in Experiment 1 can generalize to
other affective cues, or if the preference shifts are specific to face stimuli. In addition,
Experiment 2 was conducted during the collection of eye-tracking data so gaze patterns
during the decision making process and sympathetic arousal to the prime (as measured by
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 16
pupillary dilation) could be tested as mediating variables in the relationship between
affective primes and choice.
Prior work has indicated that pupillary response to affective stimuli likely
represents increases in sympathetic nervous system activity despite the pupil being
innervated by both sympathetic and parasympathetic nervous systems, making the
collection of pupillary response during affective stimuli likely a measure of sympathetic
responsivity (Bradley, Miccoli, Escrig, & Lang, 2008). Considering fixations were also of
interest, for simplicity sympathetic response to affective stimuli was acquired via
pupillometry despite other psychophysiological measures being considered ‘cleaner’
measures of sympathetic response (i.e. skin conductance). Past work combining auditory
affective primes with pupillometry have shown increased sympathetic response to
affective sounds, regardless of emotional valence (Partala & Surakka, 2003). Despite an
inherent assumption in prior works that the presented affective cues in some way alter
emotion and/or the autonomic arousal state of the participant, this has yet to be verified
on a trial-by-trial basis. In addition, in line with the Somatic Marker Hypothesis it may be
the case that the preference shifts resulting from incidental affective states is related not
to any valence specific factors, but to the arousal state induced by the prime. To test this
prediction, Experiment 2 investigated whether the sympathetic response to the affective
prime (measured via pupillary response) was a mediating variable in affective-cue based
preference shifts.
As explained in the Arousal Biased Competition model, both physiological
arousal and top-down goals have the ability to alter attention, heightening perception of
high-priority information and weakening perception of low-priority information (Mather
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 17
& Sutherland, 2011). Should the incidental affective status of the participant lead to a
bias towards short-termed goals as predicted by the Somatic Marker Hypothesis, one
might predict that incidental arousal or affective states may bias visual attention towards
the immediate option in an intertemporal choice setting. In addition, past work using
value-based binary choice has implicated that visual attention may play a role (potentially
even a causal one) in the value comparison process when making a value-based decision
(Krajbich, Armel, & Rangel, 2010). This line of work revolves around attentional drift
diffusion models of choice, which have provided evidence that the option last fixated, the
duration of fixation on a particular option, and finally the duration of first fixation may all
be associated with fixation-based biases in choice. The predictions of these drift diffusion
models even go as far as to predict a potentially causal role for fixations in value
computations, and implicate that systematic biases in fixations may lead to decision-
making alterations or deficits (Krajbich et al., 2010). As such, if visual attention is to be
pushed towards the immediate option as the Somatic Marker and Arousal Biased
competition models may predict, this increased attention may be a key mediating factor
in previously observed intertemporal preference shifts. Following this logic, relative
attention to the immediate option, duration of first fixation, and option last fixated were
all utilized in the current work in an attempt to seek out any potential causal mediation
effects on prime-induced preference shifts. My hypothesis for Experiment 2 were as
follows:
Hypothesis 1: Similar to past work, we anticipated increased pupil dilation during
affective primes relative to neutral primes.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 18
Hypothesis 2: Affective audio primes would behave similar to emotional face
primes, and would result in steeper delay discounting, regardless of the
valence of the prime.
Hypothesis 3: Sympathetic arousal during the auditory prime (measured via
pupillary response) will mediate the preference shifts resulting from the
affective primes.
Hypothesis 4: Fixation variables will be altered by affective primes and will be
predictive of choice, and affective cue-based preference shifts will be
mediated by these fixation variables.
Methods
Participants. Participants included a total of 41 undergraduate and graduate
students attending the University of Southern California (21 female) with no overlap in
subjects from Experiment 1. Of the originally recruited 41 subjects, 11 were excluded for
either displaying a lack of discounting or not maintaining a relative proportion of
immediate/delayed choices between 25-75% during the neutral condition similar to
Experiment 1. Of the retained 30 subjects, 14 were female with a mean age of 20.9±2.12.
Informed written consent was obtained from all participants before completion of the
experiment. Participants completed this experiment in exchange for both course credit
and bonus compensation based on the payout of one randomly selected trial to ensure
choices represented the true preferences of the participant.
Monetary Choice Questionnaire. Using the same questionnaire described in
Experiment 1 (Kirby et al., 1999), participants answered the same set of 27 intertemporal
choice questions prior to completion of the experimental task in order to ascertain an
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 19
estimate of delay discounting by fitting the free-parameter k using a hyperbolic model of
delay discounting (see Equation 1).
Sound Prime Intertemporal Choice Task. Similar to the emotional face prime
task used in Experiment 1, each participant’s k-value parameter best fit (see Equation 1)
from the Monetary Choice Questionnaire was used to generate trials in the affective
audio prime task. As in Experiment 1, this was done to increase sensitivity to
hypothesized experimental effects. The timeline of the task is shown in Figure 3. All
trials began with a white fixation cross at the center of a dark gray screen lasting between
2 and 4 seconds with a mean of 3. Next a neutral, pleasant, or unpleasant emotional
sound from the IADS-2 audio stimulus set (Bradley & Lang, 2007) was played through
headphones (for individual sound descriptions see Table 1). Sounds were classified into
these 3 conditions based on the mean ratings from likert scales from a prior study (1-9
scale) (Bradley & Lang 2007). Sounds were deemed pleasant, unpleasant and neutral if
mean pleasantness in the prior sample was between 6.5-9, 1-3.5 or 4.5-5.5 respectively.
Emotional sounds were then thresholded to only include sounds with arousal ratings
greater than 5 in an attempt to remove non-arousing stimuli in the emotional conditions
(figure 4). This resulted in the retention of 21 neutral sounds, 23 pleasant sounds, and 47
unpleasant sounds with the unpleasant sounds being slightly more arousing (M= 6.86,
SD=0.722) than pleasant sounds (M=6.36, SD=0.694), and both emotional conditions
being more arousing than neutral (M=4.51, SD=0.475). Sounds were randomly chosen
and were played to subjects for a total of 6 seconds, while a fixation cross was displayed
at the center of the screen. No image was presented during the sounds to keep luminance
constant and optimize tracking of pupil-based sympathetic arousal. Following sounds, a
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 20
fixation cross was displayed for 2-3 seconds, with a mean of 2.5 seconds. Participants
were then given 15 seconds to respond to an intertemporal choice set that was designed to
have both the immediate and delayed option roughly equal in present value based on pre-
determined k-values. Values of the SS option (which were always offered today) were
calculated based on previously determined hyperbolic discounting, with delay varying
between 20-66 days, and delayed value varying between $14-57. As the task progressed,
individualized k-values were titrated by 1/8 log steps following only neutral trials similar
to experiment 1. Participants completed 48 total trials, with 16 randomly selected
pleasant, unpleasant and neutral sound primes presented in a pseudorandomized order.

Figure 3. Visual depiction of a single trial of the sound prime intertemporal choice task.
Following a brief fixation cross of 2-4 seconds a pleasant, unpleasant or neutral sound
was played. Following another brief fixation of 2-3 seconds, individuals were presented
with a ‘hard’ intertemporal choice set. Following choice, the selected option was
highlighted for 0.5 seconds before the next trial began. Individuals were given a
maximum of 15 seconds to respond to the choice set.
2"3$sec$
6$sec$
0.5$sec$
RT$
2"4$sec$
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 21

Figure 4. Distribution of Arousal and Pleasantness means for the sounds used in the audio
prime intertemporal choice task. There were a total of 23 pleasant, 21 neutral and 47
unpleasant sounds used. For each participant, 16 sounds from each category were
randomly chosen for use during the task.

Eye Tracking. Subject’s pupil dilation and fixations were tracked during the
audio prime intertemporal choice task using an SMI RED250 computer mounted eye-
tracker sampling at 120Hz. Participants were calibrated with the eyetracker prior to the
task enabling successful tracking to be done without a chin rest.
Pupil Preprocessing. In order to preprocess pupil data similar to prior work
(specifically, de Gee, Knapen, & Donner, 2014), I developed a series of R functions
taking a convolution approach to the preprocessing of the pupil data. On a person-by-
person basis, the time series for the pupil dilation data for each eye was first linearly
detrended via the zoo and lm functions in R to remove any slow linear drift not of interest
to the current study. Pupil data was then subjected to a 3
rd
-order low-pass butterworth
filter of 4Hz, which was chosen to represent a threshold unlikely to contain actual
physiological variations in dilation of the pupil, following the methodology used in de
Gee et al., 2014. Data was both forward and reverse filtered via the filtfilt function to
3
4
5
6
7
8
2 4 6 8
Pleasantness Mean
Arousal Mean
condition
Neutral
Pleasant
Unpleasant
Arousal and Pleasantness Ratings for Used Sounds
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 22
prevent the phase delay characteristic of a unidirectional filter. Following detrending and
low-pass filtering, task unrelated pupil data was removed prior to the start of the task
(reading instructions) and following the task (please wait screen). Following the sample
labeling of the SMI eyetracking software, all samples not deemed either a fixation or a
saccade indicating the eye-tracker lost track of the pupil were labeled NA so as to not be
included in the estimation of dilation beta coefficients. Any trial in which 1/3 of the pupil
data was labeled NA was excluded from subsequent analysis. Since we had no
physiological basis for assuming our task would differentially affect the left or right
pupil, pupil data was averaged across eyes to further reduce noise introduced by the eye-
tracking system. In cases where one eye was labeled NA and the other eye was not, only
the non-NA eye was used for the average. Next, a canonical pupil impulse response
function (PIRF) was built using parameters mirroring those used in de Gee et al., 2013,
based on the pupil response to psychological events laid out in Hoeks & Levelt, 1993 (see
equation 2 below). The mathematical formulation of the pupil response function is as
follows:
Equation 2: ℎ 𝑡 = 𝑡
!
𝑒
!!∗!/!
!"#

Where w is the width, and t
max
is the time-to-peak of the function, and each value of t
represents a point in time (4 total seconds for the function, or 480 time points due to the
120Hz sampling rate). In line with the PIRF optimized in de Gee et al. 2014, w was set to
10.1 and t
max
was set to be 930ms. A box-car function was then created for each event of
interest for each trial, which was convolved with the PIRF. All excluded events
(including all decision timecourses and excluded sounds) were included in a single set of
box-cars which was also convolved with the PIRF. Each included sound was given a
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 23
regressor, which was then entered into a multiple linear regression analysis so each sound
could be given a beta weight that could be used for group analysis.
Fixation Preprocessing. Fixation data was preprocessed using R in a manner
similar to prior research on attentional models of decision-making (Krajbich et al., 2010).
Fixations were categorized based on which option was the target of fixation. If an option
was fixated and gaze was then averted and fixated back on the same option, all samples
were coded as viewing the same option. If, however, an option was fixated, gaze was
averted off either option before fixating the other option, then the samples between option
fixations were labeled NA and were excluded from subsequent analysis (similar to
Krajbich et al., 2010). If half or more of the decision time-course was categorized as NA
as just defined it was excluded from further analysis. The mean number of exclusions
was 6.07±12.71 (out of 48 total trials per participant). For each trial, the duration of first
fixation, the relative attention to each option, and the option last fixated were all extracted
for group analysis.
Data Analysis
Pupil Dilation and Valence of Primes. In an attempt to address Hypothesis 1
and replicate prior work using the IADS-2 in conjunction with pupil dilation (Partala &
Surakka 2003), multilevel linear mixed effect models via the lmer function in R were run
using the pupil dilation betas that were extracted from the generalized linear model used
to preprocess the time-course of pupillary dilation during the auditory prime as the
dependent variable and using the valenced condition (pleasant, unpleasant or neutral) as
the predictor variable taking the ‘leave-one-out’ strategy described in detail in
Experiment 1. This analysis was done to both validate that pupil dilation in our sample
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 24
was not abnormal during the primes, and to validate that our developed pupil processing
pathway would replicate prior results (Partala & Surakka, 2003).
Primes and Choice. The effect of prime condition on intertemporal choice was
analyzed using multilevel generalized linear mixed effects models. Similar to Experiment
1, the log of k and gender were controlled for, and the log of k was allowed to vary as a
random slope. The valenced condition (i.e., pleasant, unpleasant and neutral audio
primes) was used as the independent variable of interest in an attempt to address
Hypothesis 2.
Pupil Dilation and Choice. A mixed-effect logistic regression was run using
intertemporal choice as the dependent variable, and the extracted audio prime pupil
dilation beta value as the predictor of interest controlling for gender, and the log of k with
the log of k varying as a random slope.
Dilation, Primes and Choice Mediation. Due to the relationship seen between
prime dilation, affective condition and choice, a causal mediation analysis was run via the
mediate function in the mediation package in R. The unpleasant and neutral affective
conditions were used as the treatment conditions due to only unpleasant primes reaching
significance in the behavioral analysis, and pupil dilation betas during exposure to the
prime were tested as the mediator.
Fixation and Choice. A mixed effects logisitic regression model was run using
the different fixation variables as predictors of interest in order to elucidate which aspects
of fixation were influencing choice. Gender and the log of k were controlled for in the
model and the log of k was allowed to vary as a random slope in the model.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 25
Fixation and Primes. Three separate multilevel models were run on percent
fixation on the SS option, duration of first fixation and the option fixated last all using
affective prime condition as the independent variable controlling for gender and the log
of k and allowing the log of k to vary as a random slope.
Exploratory Analysis. As discussed in the introduction, several papers have
found evidence not for valence effects, but for emotion specific effects on intertemporal
choice preferences (e.g. in Lerner, Li, & Weber, 2012 sadness increased impulsive choice
whereas there was no preference shift following disgust despite a similar valence).
Following suit with studies of this kind, in an exploratory manner each specific audio
prime that was heard by the participant was tested as a predictor of interest in a single
mixed effects logistic regression model. Due to the large number of sounds available for
presentation, this exploratory analysis was only done with sounds that at least half the
sample (n=15) had responded to in an attempt to remove low powered primes, resulting
in a total of 48 audio primes being used. This prime-specific analysis was first done with
all sounds included in the model in an attempt to answer if it improves estimation of
choice, then all neutral primes were treated as the comparison condition to see which
emotional primes differed significantly in either direction.
Results
Pupil Dilation and Valence of Primes. Consistent with past work and
Hypothesis 1, affective-cue conditions were associated with increased pupil dilation beta
values when compared with neutral condition dilatory response (𝜒
!
(2) = 10.83, p=0.004)
(see Figure 5.a. for dilation betas, and Figure 5.b. for auditory prime dilation time-
courses).
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 26

Figure 5. Pupil dilation during the audio prime was significantly increased during the
pleasant and unpleasant sounds when compared with the neutral primes (𝜒
!
(2) = 10.83,
p=0.004). a. The affective prime dilation betas as outputted from multiple linear
regression model averaged within affective condition for each subject with error bars
representing the standard error of the mean. b. Mean time course for each affective
condition averaged across all trials and subjects including 1 second following the prime
to illustrate the full range of data included in the PIRF which has a 930ms rise and fall.

0
1
2
Neutral Pleasant Unpleasant
Dilation Beta
Valenced Conditions Increase Pupil Dilation
−0.1
0.0
0.1
0.2
0 2 4 6
Sound Timecourse
Pupil Dilation
Neutral
Pleasant
Unpleasant
Pupil Dilation Time−courses Differ During Valenced Sounds
a.
b.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 27
Primes and Choice. Contrary to Hypothesis 2, there was not a significant overall
main effect of cue condition on choice despite both affective prime conditions tending to
push people towards the SS option (𝜒
!
(2) = 4.146, p=0.126). However, intertemporal
choice following unpleasant primes was associated with significantly increased impulsive
choice when compared to neutral with a regression coefficient of -0.28 ± 0.14
corresponding to an odds ratio of 0.75 (z= -1.97, p = 0.0484) (Figure 6).

Figure 6. Unpleasant affective cue-primes were associated with increased impulsive
choice. a. Beta values for prime conditions with their 95% confidence intervals when
compared with neutral. b. The predicted probability of choosing the larger later option
−0.6
−0.4
−0.2
0.0
Pleasant Unpleasant
Beta Value
Unpleasant Primes Significantly Increase Impulsive Choice
0.3
0.4
0.5
0.6
neutral pleasant unpleasant
Predicted Probability
Predicted Probability of choosing the LL option following auditory primes
*"
a.
b.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 28
was 0.44, 0.40, 0.384 for the neutral, pleasant and unpleasant audio primes respectively
with only the unpleasant condition being significantly different from neutral. Error bars
correspond to upper and lower quartiles with stars indicating significance.

Pupil Dilation, Primes and Choice. Contrary to Hypothesis 3, pupil dilation
during the affective primes was not significantly related to choice (z = -1.47, p = 0.14),
though this was trending in the predicted direction (increased dilation when an individual
went on to choose the SS option). In addition, pupillary dilation during the prime
comparing the unpleasant and neutral conditions did not significantly mediate the effects
of affective condition on choice (Total Effect = -0.013, ACME = -0.002 (p = 0.17)).
Fixation and Choice. Consistent with Hypothesis 4, the option that was last
fixated was significantly predictive of choosing the larger later option with a regression
coefficient of -2.13±0.16 (standard error) for the smaller sooner option being fixated last
(z=-13.01, p<0.001) corresponding to an odds ratio of 0.118 (Figure 6.). In general,
participants tended to fixate more on the smaller sooner option last (percent,
M=0.53,SD=0.499). Similarly, the attention paid to the smaller sooner option was
significantly predictive of choosing the larger later option with a regression coefficient of
-3.46±0.74 (standard error) (z=-4.69, p<0.001) corresponding to an odds ratio of 0.031
(Figure 7.). In general, participants tended to spend a greater percentage of time fixating
on the larger later option (M=0.626, SD=0.139). However, contrary to Hypothesis 4, the
duration of first fixation was not predictive of choice though in general the larger later
option was fixated longer (M=0.713s, SD=0.458) than the smaller sooner option
(M=0.444s, SD=0.248) when first fixated.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 29

Figure 6. The probability of choosing the larger later option is significantly related to
which option was fixated last in the decision-making process (z=-13.01, p<0.001)
corresponding to an odds ratio of 0.118. Error bars shown correspond to upper and lower
quantiles with stars indicating significance.

Figure 7. The probability of choosing the larger later option goes down as the amount of
time spent fixating on the smaller sooner option is increased (z=-4.69, p<0.001)
corresponding to an odds ratio of 0.031.

Fixation and Primes. Contrary to Hypothesis 4, percent fixation on the SS
option, duration of first fixation, and option last fixated were not significantly altered as a
function of affective prime condition.
0.2
0.4
0.6
LL SS
Option Last Fixated
Predicted Probability
Predicted Probability of choosing the LL option dependent on last option fixated
*"
0.00
0.25
0.50
0.75
1.00
0.25 0.50 0.75
Percentage of Fixation on SS
Probability of LL Choice
Attention to Options Predicts Choice
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 30
Exploratory Analysis. There was a significant main effect of which prime was
heard on intertemporal choice preferences (𝜒
!
(47) = 112.99, p<0.001). Following up on
this significant main effect of prime, the mixed effects logistic regression model was run
again with all neutral sounds dummy coded to be the baseline, this new model indicated
that both pleasant and unpleasant sounds were capable of both increasing and decreasing
impulsive choice though affective primes generally tended to encourage impulsive choice
(22 of 32 affective primes had negative regression coefficients representing increased
impulsive choice) indicating the behavioral effects described here are likely prime-
specific and not valence-specific (Table 1).
Table 1. Summary statistics for the behavioral effects of different valenced primes on intertemporal
choice. Significant or trending primes were highlighted with blue indicating increased SS choice, and
red indicating increased LL choice. Both pleasant and unpleasant primes pushed people towards both
the smaller sooner and larger later choice depending on prime. Arousal ratings are from a prior
sample.
Audio Prime Valence Arousal Regression
Coefficient
Log
Odds
Z-score p-value
Erotic Female (201) Pleasant 7.31 -1.11 0.33 -2.02 0.04*
Erotic Female (202) Pleasant 7.13 -0.796 0.45 -1.6 0.11
Boy Laughing (220) Pleasant 6 -0.8 0.45 -1.64 0.1
Male Laugh (221) Pleasant 5.05 0.037 1.04 0.08 0.93
Laughing (226) Pleasant 5.42 -0.27 0.77 -0.54 0.59
Crowd (311) Pleasant 7.12 -0.31 0.73 -0.69 0.49
Baseball (353) Pleasant 6.62 0.84 2.32 1.86 0.06
Party (365) Pleasant 6.32 -0.46 0.63 -0.95 0.35
Colonial Music (601) Pleasant 5.84 -0.08 0.92 -0.19 0.85
Slot Machine (716) Pleasant 6.44 -1.31 0.27 -2.38 0.02*
Slot Machine (717) Pleasant 6.56 0.29 1.34 0.56 0.57
Wedding (813) Pleasant 5.89 -0.46 0.63 -0.9 0.37
Rock N Roll (815) Pleasant 6.85 -0.69 0.5 -1.55 0.12
Guitar (816) Pleasant 5.23 0.17 1.19 0.38 0.7
Bongos (817) Pleasant 7.15 1.22 3.4 2.62 0.009**
Funk Music (820) Pleasant 5.87 -0.74 0.48 -1.52 0.13
Buzzing (116) Unpleasant 6.51 1.43 4.17 2.3 0.02*
Babies Cry (260) Unpleasant 6.87 -0.39 0.68 -0.74 0.45
Scream (275) Unpleasant 8.16 0.34 1.41 0.71 0.48
Female Scream (276) Unpleasant 7.77 -0.43 0.65 -0.79 0.43
Fight (283) Unpleasant 6.2 -1.0 0.37 -1.79 0.07
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 31
Creep (288) Unpleasant 6.82 0.08 1.09 0.15 0.88
Fight (290) Unpleasant 7.61 0.50 1.65 1.02 0.31
Tire Skids (422) Unpleasant 7.52 -0.04 1.04 -0.07 0.94
Car Wreck (424) Unpleasant 7.99 -1.02 0.36 -1.53 0.13
Plane Crash (501) Unpleasant 6.93 0.004 1.0 0.01 0.99
Air Raid (624) Unpleasant 7.1 -1.29 0.27 -2.14 0.03*
Busy Signal (703) Unpleasant 5.68 -0.71 0.49 -1.11 0.27
Siren (711) Unpleasant 7.39 -1.87 0.15 -2.63 0.009**
Buzzer (712) Unpleasant 7.98 -0.7 0.5 -1.38 0.17
Siren (714) Unpleasant 6.94 -1.09 0.33 -1.78 0.07
Crash (732) Unpleasant 6.98 -0.42 0.66 -0.81 0.42
Note: All statistics are based on comparisons with the average choice behavior following
neutral primes.

Discussion
Consistent with prior work and Hypothesis 1, the positively and negatively
valenced primes were both associated with increases in pupillary dilation indicative of an
increased sympathetic arousal during the prime. In confirmation of Hypothesis 2, results
evidenced increased impulsive choice following both pleasant and unpleasant primes,
though only unpleasant sounds reached significance. Both Hypothesis 3 and 4 were only
partially confirmed. Despite sympathetic arousal being successfully manipulated by the
primes, this sympathetic response did not statistically mediate the preference shifts
resulting from the affective primes. Similarly, despite both the option last fixated, and
relative fixation to each option being predictive of choice, neither of these variables was
affected by affective prime condition, making causal mediation analysis unwarranted.
General Discussion
Across both experiments, there was evidence that both positively and negatively
valenced primes have an overall tendency to push people towards impulsive choice as
predicted by the Somatic Marker Hypothesis framework described in the introduction.
Despite this general trend for affective primes to indiscriminately encourage impulsive
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 32
choice, behavioral results of Experiment 2 (specifically the exploratory analysis) implies
that being deemed as a pleasant or unpleasant affective prime may not be enough to
accurately predict directionality of preference shifts, with our results implicating that
future works should perhaps consider using arousal states to categorize affective primes.
The current work is the first of its kind to verify that trial-by-trial affective-cues
prior to intertemporal choice lead to sympathetic changes within the decision-maker.
Pupillometry data indicated that there is a valence independent increase in sympathetic
response to both pleasant and unpleasant stimuli when presented prior to intertemporal
choice, consistent with prior work (Partala & Surakka, 2003). These trial-by-trial
variations in sympathetic response were used to address whether alterations in
sympathetic response can lead to systematic shifts in intertemporal choice preferences.
However, contrary to the Hypothesis of Experiment 2, sympathetic arousal status did not
significantly predict SS choice (though the data was trending in the predicted direction),
and the sympathetic response to the affective prime was not seen to mediate the
relationship seen between the affective categorization of the prime, and resulting
systematic preference shifts towards the SS option as predicted.
Though the current work has provided early evidence that arousal may not be a
mediating variable in the observed effects, patterns within the data point to some
interesting opportunities for future works. The mean arousal ratings in Experiment 2
given from a prior sample for the unpleasant affective cues had a slightly higher mean
value than did the pleasant affective cues. In addition, the unpleasant cues were
associated with slightly increased sympathetic responses in the current work (as measure
via pupillary dilation). If these attributes are taken into consideration in conjunction with
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 33
the behavioral findings of Experiment 2 in which both pleasantly and unpleasantly
valenced primes increased relative SS choice, though this was only statistically
significant during the unpleasant primes, this may implicate that should the sounds be
categorized not by valence, but by either self-report or autonomic arousal, this may lead
to predictable systematic preference shifts towards the SS option. The current work did
not collect self-reported arousal ratings, and is underpowered for ranking the primes
based on mean sympathetic responses, but this may be a fruitful area for future works.
1

In addition to investigating the role of sympathetic arousal, eye gaze was also
monitored in Experiment 2 in an attempt to determine whether visual attention to the
intertemporal choice set was altered as a result of the affective primes. Results indicated
that in confirmation of past formulations of visual attention during binary choice (i.e.
attentional Drift Diffusion Models), both relative attention and the option last fixated
were predictive of intertemporal choice. However, our Hypothesis that the affective
status of the prime may be altering visual attention during choice was not confirmed,
which provides some initial evidence that alterations in intertemporal choice preference
may not be mediated by an increased attention to the SS option. This does, however, still
leave as an open question whether visual attention to the specific attributes (i.e., dollar
amount, or temporal delay) during choice are altered by affective primes. The current
work was not designed to disentangle which attribute of the option was being fixated,
something that future works may consider addressing.
Overall, the current work adds a new perspective to the growing literature on how
emotions and affective states can alter intertemporal choice preferences. Results from
both Experiments seem to support predictions that might be made by the Somatic Marker
1. Since the original formulation of this manuscript, 23 additional subjects have been added to Experiment 2 (total n = 64) and with the addition of more subjects
ranking of the affective primes based on their sympathetic response becomes predictive of impulsive choice.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 34
Hypothesis in that the induction of an affective state seems to generally increase short-
term goal drives which are manifested by increased preference for the immediate option
in an intertemporal choice set. The current work also implies that future experimental
investigations seeking out the mechanisms underlying the manner in which affective
states can influence intertemporal choice preferences might consider focusing on the
arousal status of the participant, instead of continuing to build on prior works which
primarily focus on either the valence of affective-cues or specific emotions.

PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 35
References
Ainslie, G. (1975). Specious reward: A behavioral theory of impulsiveness and impulse
control. Psychological Bulletin, 82(4), 463–496.
Augustine, A., & Larsen, R. (2011). Affect Regulation and Temporal Discounting:
Interactions Between Primed, State, and Trait Affect. Emotion, 11(2), 403–412.
Bechara, A., & Damasio, A. (2005). The somatic marker hypothesis: A neural theory of
economic decisions. Games and Economic Behavior, 52, 336–372.
Bradley, M., & Lang, P. (2007). The International Affective Digitized Sounds (2nd
Editions; IADS-2): Affective ratings of sounds and instruction manual.
Gainselville, FL: University of Florida.
Bradley, M., Miccoli, L., Escrig, M., & Lang, P. (2008). The pupil as a measure of
emotional arousal and autonomic activation. Psychophysiology, 45, 602–607.
Clore, G., & Huntsinger, J. (2007). How emotions inform judgment and regulate thought.
Trends in Cognitive Sciences, 11(9), 393–399.
De Gee, J. W., Knapen, T., & Donner, T. (n.d.). Decision-related pupil dilation reflects
upcoming choice and individual bias. PNAS. Retrieved from
www.pnas.org/cgi/doi/10.1073/pnas.1317111
Doyle, J. (2013). Survey of time preference, delay discounting models. Judgment and
Decision Making, 8(2), 116–135.
Gray, J. (1999). A Bias Towards Short-term Thinking in Threat Related Negative
Emotional States. Personality and Social Psychology Bulletin, 25(1), 65–75.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 36
Hinson, J. M., Jameson, T. L., & Whitney, P. (2003). Impulsive decision making and
working memory. Journal of Experimental Psychology: Learning, Memory, and
Cognition, 29(2), 298–306.
Hirsch, J., Guindon, A., Marisano, D., & Peterson, J. (2010). Positive Mood Effects on
Delay Discounting. Emotion, 10(5), 717–721.
Hoeks, B., & Levelt, W. (1993). Pupillary dilation as a measure of attention: A
quantitative system analysis. Behavior Research Methods, Instruments, &
Computers, 25(1), 16–26.
Ifcher, J., & Zarghamee, H. (2011). Happiness and Time Preference: The Effect of
Positive Affect in a Random-Assignement Experiment. The American Economic
Review, 101(7), 3109–3129.
Kirby, K. N. (2009). One-year temporal stability of delay-discount rates. Psychonomic
Bulletin & Review, 16(3), 457–462. http://doi.org/10.3758/PBR.16.3.457
Kirby, K. N., Petry, N. M., & Bickel, W. K. (1999). Heroin addicts have higher discount
rates for delayed rewards than non-drug-using controls. Journal of Experimental
Psychology: General, 128(1), 78–87. http://doi.org/10.1037/0096-3445.128.1.78
Krajbich, I., Armel, C., & Rangel, A. (2010). Visual fixations and the computation and
comparison of value in simple choice. Nature Neuroscience, 13(10), 1292–1298.
Kranz, F., & Ishai, A. (2006). Face Perception Is Modulated by Sexual Preference.
Current Biology, 16, 63–68.
Lerner, J., Li, Y., & Weber, E. (2012). The Financial Costs of Sadness. Psychological
Science, 10.1177(0956797612450302), 1–8.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 37
Li, X. (2008). The Effects of Appetitive Stimuli on Out-of-Domain Consumption
Impatience. Journal of Consumer Research, 34(5), 649–656.
Liu, L., Feng, T., Chen, J., & Li, H. (2013). The Value of Emotion: How Does Episodic
Prospection Modulate Delay Discounting. PLoS ONE, 8(11), e81717.
Loewenstein, G. (1996). Out of control: Visceral influences on behavior. Organizational
Behavior and Human Decision Processes, 65(3), 272–292.
Loewenstein, G., & Lerner, J. (2003). The Role of Affect in Decision Making. In
Handbook of Affective Sciences (pp. 619–642). Oxford University Press.
Luo, S., Ainslie, G., Giragosian, L., & Monterosso, J. R. (2009). Behavioral and neural
evidence of incentive bias for immediate rewards relative to preference-matched
delayed rewards. The Journal of Neuroscience, 29(47), 14820.
Luo, S., Ainslie, G., & Monterosso, J. (2014). The behavioral and neural effect of
emotional primes on intertemporal decisions. Social Cognitive and Affective
Neuroscience, 9, 283–291.
Mather, M., & Sutherland, M. (2011). Arousal-biased competition in perception and
memory. Perspectives on Psychological Science, 6(2), 114–133.
Moser, E., Derntl, B., Robinson, S., Fink, B., Gur, R., & Grammer, K. (2007). Amygdala
activation at 3T in response to human and avatar facial expressions of emotions.
Journal of Neuroscience Methods, 161, 126–133.
Murawski, C., Harris, P., Bode, S., Dominguez, J., & Egan, G. (2012). Led into
Temptation? Rewarding Brand Logos Bias the Neural Encoding of Incidental
Economic Decisions. PLoS ONE, 7(3), e34155.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 38
Partala, T., & Surakka, V. (2003). Pupil size variation as an indication of affective
processing. International Journal of Human-Computer Studies, 59, 185–198.
Patak, M., & Reynolds, B. (2007). Question-based assessments of delay discounting: Do
respondents spontaneously incorporate uncertainty into their valuations for
delayed rewards? Addictive Behaviors, 32, 351–357.
Pyone, J., & Isen, A. (2011). Positive Affect, Intertemporal Choice, and Levels of
Thinking: Increasing Consumers’ Willingness to Wait. Journal of Marketing
Research, XLVIII, 523–543.
Raeva, D., Mittone, L., & Scharzbach, J. (2010). Regret no, take it now: On the role of
experienced regret on intertemporal choice. Journal of Economic Psychology, 31,
634–642.
Samuelson, P. (1937). A Note on Measurement of Utility. The Review of Economic
Studies, 4(2), 155–161.
Schwarz, N., & Clore, G. (1983). Mood, Misattribution, and Judgments of Well-Being:
Informative and Directive Functions of Affective States. Journal of Personality
and Social Psychology, 45(3), 513–523.
Tottenham, N., Tanaka, J., Leon, A. C., McCarry, T., Nurse, M., Hare, T. A., … Nelson,
C. A. (2009). The NimStim set of facial expressions: judgments from untrained
research participants. Psychiatry Research, 168(3), 242–249.
Tuk, M., Trampe, D., & Warlop, L. (2011). Inhibition Spillover: Increased Urination
Urgency Facilitates Impulse Control in Unrelated Domains. Psychological
Science, 22(5), 627–633.
PRIMING AND INTERTEMPORAL PREFERENCE SHIFTS 39
Van den Bergh, B., Dewitte, S., & Warlop, L. (2008). Bikinis Instigate Generalize
Impatience in Intertemporal Choice. Journal of Consumer Research, 35(1), 85–
97.
Wilson, M., & Daly, M. (2004). Do pretty women inspire men to discount the future.
Proceedings of the Royal Society of London., 271, S177–S179.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Homeostatic imbalance and monetary delay discounting: effects of feeding on RT, choice, and brain response

PDF

Behabioral and neural evidence of state-like variance in intertemporal decisions

PDF

Sequential decisions on time preference: evidence for non-independence

PDF

Aging and emotion regulation in the judgment of facial emotion

PDF

Behavioral and neural evidence of incentive bias for immediate rewards relative to preference-matched delayed rewards

PDF

Choice biases in making decisions for oneself vs. others

PDF

Emotion, attention and cognitive aging: the effects of emotional arousal on subsequent visual processing

PDF

Reward immediacy and subjective stress modulate anticipation of primary and secondary rewards in temporarily-abstinent cigarette smokers

PDF

Heart, brain, and breath: studies on the neuromodulation of interoceptive systems

PDF

Spatial and temporal patterns of brain activity associated with emotions in music

PDF

Age differences in jealousy: an ATSS examination of the role of attributions in the emotional reactions of older and younger adults in relationship-threatening situations

PDF

Effect of mindfulness training on attention, emotion-regulation and risk-taking in adolescence

PDF

Influence of age and anxiety on recognition of facial expressions of emotion: exploring the role of attentional processes

PDF

Reacting to the negative consequences of your decisions: an ATSS comparison of cognitive and affective reactions in older versus younger adults

PDF

The acute impact of glucose and sucralose on food decisions and brain responses to visual food cues

PDF

Deconstructing the psychological components of emotional decision making and their relation to the suicide continuum

PDF

Emotional arousal amplifies the selectivity of visual selective attention

PDF

When it's good to feel bad: how responses to virtual environments predict real-life sexual risk-reduction

PDF

Taking the temperature of the Columbia Card Task

PDF

A neuropsychological exploration of low-SES adolescents’ life goals and their motives

Asset Metadata

Creator Melrose, Andrew James (author)

Core Title Visual and audio priming of emotional stimuli and their relationship to intertemporal preference shifts

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Psychology

Publication Date 07/06/2015

Defense Date 05/30/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag affect,delay discounting,emotion,eye-tracking,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Monterosso, John R. (committee chair), Bechara, Antoine (committee member), Mather, Mara (committee member)

Creator Email amelrose@usc.edu,andrewjamesmelrose@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-585820

Unique identifier UC11299533

Identifier etd-MelroseAnd-3545.pdf (filename),usctheses-c3-585820 (legacy record id)

Legacy Identifier etd-MelroseAnd-3545.pdf

Dmrecord 585820

Document Type Thesis

Format application/pdf (imt)

Rights Melrose, Andrew James

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA