University of Southern California Dissertations and Theses

Examining Pavlovian-to-instrumental transfer effect in human mate selection behavior

(USC Thesis Other)

Examining Pavlovian-to-instrumental transfer effect in human mate selection behavior

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

EXAMINING PAVLOVIAN-TO-INSTRUMENTAL TRANSFER EFFECT IN HUMAN
MATE SELECTION BEHAVIOR

by

Aili Qiao

A Thesis Presented to the
FACULTY OF THE USC DORNSIFE COLLEGE OF LETTERS, ARTS AND SCIENCES
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTERS OF ARTS
(PSYCHOLOGY)

December 2021

Copyright 2021 Aili Qiao

ii
Table of Contents
List of Tables iii

List of Figures iv

Abstract v

Introduction 1

Experiment 1 4
Methods 4
Results 7
Conclusion 8

Experiment 2 9
Methods 9
Results 11
Conclusion 13

Experiment 3 14
Methods 14
Results 16
Conclusion 18

General Discussion 18

References 22

Appendices 29

iii
List of Tables
Table 1: List of Variables and Measurements for Experiment 1. 31
Table 2: Experiment 1 Model Outputs. Binary Logistic Multilevel Modeling on Swipe. 32
Table 3: List of Variables and Measurements for Experiment 2. 33
Table 4: Experiment 2 Model Outputs. Binary Logistic Multilevel Modeling on Swipe. 34
Table 5: List of Variables and Measurements for Experiment 3. 35
Table 6: Experiment 3 Model Outputs. Binary Logistic Multilevel Modeling on Swipe. 36

iv
List of Figures
Figure 1: Gifs Used for Pavlovian Training. 38
Figure 2: Fractals Used for Training and Testing. 39
Figure 3: Process of One Training Trial. 40
Figure 4: Process of Testing Trials. 41
Figure 5: Likelihood of Swiping “Connect” by Participants’ SAQ Score Moderated by
Cue Condition. 42
Figure 6: Likelihood of Swiping “Connect” by Participants’ Anxious Subscale Score of
AAS Moderated by Cue Condition. 43
Figure 7: Likelihood of Swiping “Connect” by Participants’ Motivation to Seek New
Partners Moderated by Cue Condition. 44
Figure 8: Likelihood of Swiping “Connect” by Participants’ BAS Drive Score
Moderated by Cue Condition. 45

v
Abstract
Pavlovian-to-instrumental transfer (PIT) effect suggests that external cues can signal
potential reward/punishment values, subsequently affecting decision-making and behavior.
Despite extensive research on PIT effect in humans, there is still a lack of research examining
PIT in social contexts. Therefore, the current paper aims to develop a social behavioral PIT task
in three experiments, in which participants are trained with social outcomes and tested with a
social task. Participants first learned the association between neutral cues and facial expressions,
and then they completed a dating task where participants read dating profiles of potential mates
with previously trained cues as profile backgrounds. Although the results from the three
experiments did not sufficiently replicate one another, our findings suggest that participants’
seeking-motivation played a crucial role in how they utilize Pavlovian cues in their mate
selection behavior. Future direction and further improvement needed for the social behavioral
PIT task are discussed.

1
Introduction
Being in the movie theatre might make us want to purchase popcorn, even if we do not
usually eat theatre snacks. Hearing the emergency siren might stop us from crossing the street,
even if we are unsure as to where the ambulance is coming from or going to. Instances like these
show that external cues can motivate and inhibit our actions, thus are crucial in explaining
human and animal behavior. One paradigm that addresses the relationship between predictive
cues and behavior is the Pavlovian-to-instrumental transfer (PIT) effect.
A PIT effect is generally established in three steps. First, Pavlovian cues are associated
with the presence of rewards/punishments. Second, instrumental actions are trained that achieve
the same rewards/punishments. Finally, a transfer effect is observed when the mere presence of
the Pavlovian cue can elicit the corresponding instrumental action, even without the outcomes. In
other words, Pavlovian cues that signal the value and probability of a reward/punishment can
subsequently affect behavior.
The PIT effect was initially studied with non-human animals (e.g., Colwill & Roscorla,
1988; Estes, 1943; Holland, 2004; Lovibond, 1981, 1983; Rescorla & Solomon, 1967; Roscorla,
1994; Solomon & Turner, 1962; Knowlton & Yin, 2006; for review, see Dickinson & Balleine,
2002; and Holmes et al., 2010). In the past decade, PIT research extended to human behaviors
with studies examining either appetitive PIT (see Cartoni et al., 2015, and Quail et al., 2017a,
2017b for food; see Hogarth et al., 2007 for tobacco; see Talmi et al., 2008, and Eder & Dignath,
2016 for money; for review, see Cartoni et al., 2016) or aversive PIT (see Trick et al., 2011, and
Garofalo & Robbins, 2017 for aversive noises; see Rigoli et al., 2012, and Claes et al., 2016 for
electric shocks), and in rare occasions, both appetitive PIT and aversive PIT simultaneously
(Huys et al., 2011; Huys et al., 2016).

2
Within human PIT research, most paradigms involve money (e.g., the stock market
paradigm, Allman et al., 2010), defensive games (e.g., Paredes-Olay et al., 2002; Garofalo &
Robbins, 2017), food (e.g., the vending machine paradigm, Quail et al., 2017a, 2017b), or simple
go/no-go (e.g., the mushroom paradigm, Huys et al., 2011; Huys et al., 2016). While human PIT
research is abundant, there is still a lack of research examining PIT in social contexts. Only a few
studies so far include stimuli and outcomes that can be considered social. One is a study by
Lehner and colleagues (2017) where the authors used a social reward – i.e., a picture of a person
making the “thumbs-up” gesture – among many other rewards, then examined how different
types of rewards affect the PIT effect. Another is a study by Mueller and colleagues (2019)
where the authors used pictures of faces as neutral cues and paired those with aversive outcomes.
Both studies used human faces as a component of their studies: the prior used them as an
outcome while the latter used them as neutral cues. However, to our knowledge, no prior
research had employed social tasks that examine social PIT effects.
We believe that PIT effect is a crucial piece in understanding social behavior. For
example, learned external cues could affect whom we choose to approach and befriend in social
situations and whom we avoid, or determine when we speak up in class and when we choose to
stay silent. Since prior research has already found strong evidence that learned external cues can
affect human and non-human non-social behavior, one can certainly speculate that a similar
effect exists in social behavior: cues that signal the possibility of social reward/punishment will
affect one’s social behavior.
Therefore, we aim to develop a novel PIT paradigm in which participants are trained with
social outcomes and tested with a social task. Participants first learned the association between
neutral cues and positive/negative/neutral facial expressions, and then they were told to complete

3
a dating task. The dating task is modeled after modern-day dating apps, where users encounter
profiles of potential mates one by one and make dichotomous decisions via swiping right vs. left
to either connect or reject the potential mate. Each profile was paired with a previously trained
cue for us to examine whether the cues affect participants’ mate-selection behavior during the
dating task.
A primary hypothesis is that the presentation of a positively trained Pavlovian cue (i.e.,
the positive facial expression) would increase participants’ likelihood to swipe-connect during
the dating task. In contrast, the presentation of a negatively trained Pavlovian cue (i.e., the
negative facial expression) would lead to a decrease in the likelihood of a swipe-connect. Our
reasoning is that the trained Pavlovian cues would signal the probability of the other person’s
responses (e.g., welcoming or disapproving) and thus affect participants’ social behavior to
either approach or avoid a potential mate. Building upon the primary hypothesis, we further
speculate on the effects of individual differences such as gender, relationship status, and current
mate-seeking motivation on the PIT effect. Given that mate selection is a goal-directed behavior,
we have highlighted participants’ mate-seeking motivation as a predictor of interest. To those
who have little to no motivation in mate-seeking, learned Pavlovian cues that signal a new
potential mate’s response should have little to no effect on their behavior. Whereas those who
have high mate-seeking motivation might heavily utilize the Pavlovian cues in their mate
selection decision-making.
Furthermore, we selected several personality measures that we believed to be related to
our particular social PIT paradigm. One of which is the Social Anxiety Questionnaire (SAQ;
Caballo et al., 2010), as individuals with social anxiety may have heightened sensitivity to social
reward/punishment cues compared to those who are not socially anxious. Another is the Adult

4
Attachment Scale (AAS; Collins, 1996), as individuals with different attachment styles respond
differently to possible romantic rejections. A final one is the Behavioral Approach/Inhibition
Systems Scales (BIS/BAS; Carver & White, 1994), as it measures participants’ sensitivity to
reward/punishment and their approach/avoidance motivations.
We tested the social PIT paradigm in three experiments. Experiment 1 served as an initial
trial of the social training and testing tasks. Experiment 2 offered more targeted training to
attempt to maximize the potential PIT effect. Experiment 3 included primes that aimed to
manipulate participants' mate-seeking motivation as a means to examine seeking motivation’s
effect on PIT.
Experiment 1
Methods
Participants
A total of 120 participants was recruited from Prolific with the criteria that a) they are
U.S. citizens and b) they are between the ages 18 and 29 inclusive. Each participant was
rewarded $3.00 for completing the experiment. Some participants were removed from analysis
due to failed attention checks, leaving a final sample size of 113 (mean age = 23.42, 49.6%
females, 53.1% Non-Hispanic White).
Materials
Expressions' gifs. Twelve facial expression gifs were used in this experiment as social
reward and punishment. The gifs were obtained from the Ryerson Audio-Visual Database of
Emotional Speech and Song (RAVDESS; Livingstone & Russo, 2018). The selected gifs are
from four different actors (2 males and 2 females), with each actor exhibiting three different
expressions (welcoming, neutral, and disapproving) in front of a white background. The

5
welcoming gifs were used as social reward, the disapproving gifs were used as social
punishment, and the neutral gifs were used as controls. Each gif is around three seconds in
duration. The expression gifs were pilot tested prior to the experiment, to ensure that they elicit
the target emotions and evaluations. See Figure 1 for the gifs used for Pavlovian training.
Fractals. Three black-and-white fractals were used in this experiment as neutral
Pavlovian cues. The fractals were obtained from a previous PIT study by Watson and colleagues
(2014). The patterns are relatively neutral – one of a polka-dot pattern, one of a zebra print, and
one of a fingerprint pattern. See Figure 2 for the fractals used for instrumental testing.
Dating Profiles. Twenty-one simulated dating profiles were used in this experiment as a
testing measure for the PIT effect. Each profile contained a 250-word gender-neutral self-
introduction, and none of the profiles had pictures. The dating profiles were pilot tested prior to
the experiment, to ensure that they are sufficiently gender-ambiguous and are not overly
attractive or unattractive. See appendix A for full list of profiles used for the transfer test.
Individual Differences. Three scales were used in the current experiment. The Social
Anxiety Questionnaire (SAQ; Caballo et al., 2010) was used to measure social anxiety. The
Adult Attachment Scale (AAS; Collins, 1996) was used to measure attachment style. The
Behavioral Approach/Inhibition Systems scales were used to measure approach/avoidant
motivations (BIS/BAS; Carver & White, 1994).
Procedure
After providing informed consent, participants underwent a training phase and a testing
phase. Finally, demographic information and individual differences measures were obtained.
Training Phase. Before training started, participants were asked to pay close attention in
the following sections. During training, participants viewed facial expression gifs with black-

6
and-white fractals as the backdrop. The training consisted of 5 blocks, each with 12 expression-
fractal training trials (four actors doing three expressions each), totaling 60 trials during the
training phase. For each trial, one fractal appeared full screen for 2.5s with the words
“Loading… Please wait.” displayed in the center. Then, a corresponding facial expression gif
appeared for 3s on top of the existing fractal, taking up approximately a quarter of the screen
space while still allowing the fractal to be visible. See Figure 3 for an example a training trial.
The order of the gifs and the expression-fractal pairings were pseudo-randomized. Participants
were trained on all three expressions (welcoming, neutral, and disapproving) and their
corresponding fractals. Attention checks were placed in between blocks, asking the participants
to recall the last actor they saw in the gif.
Testing Phase. Immediately after training, participants were informed that they will be
completing a dating task, where they will read each dating profile and choose to either show an
interest to connect or reject the person. Participants were instructed to imagine the profiles are
written by people of the same gender/age/ethnicity of their dating preference. The testing
consisted of 21 trials – 7 trials in each of the three fractal conditions. Each trial consisted of a
dating profile displayed on top of a previously trained fractal. See Figure 4 for a visual
representation of the testing process. The order of the dating profiles and the profile-fractal
pairings were pseudo-randomized.
Analysis
A series of generalized linear multilevel analyses for a binomial distribution was
performed in R using the “glmer” function from “lme4” package. Level 1 consisted of the
measurement of swipes (binary: connect or reject), with a sample size of 2,373. Participants were
required to make a decision before proceeding to the next trial, therefore no missing trials were

7
recorded. Level 2 consisted of (a) participants, with a sample size of 113, and of (b) testing
stimuli (i.e., which profile they read), with a sample size of 21.
Our main predictors of interest were cue condition (Level 1) and participants’ seeking
motivation (Level 2a). Additional predictors we explored include gender, relationship status,
SAQ scores, BIS/BAS scores, and AAS scores (all Level 2a). Finally, participants’ awareness of
the expression-fractal relationship was examined for awareness’ effect on PIT. All continuous
variables were mean-centered. Categorical predictors were treated as factors.
Results
Null Model
A null model was constructed with the following parameters: a fixed-effect intercept and
two parameters expressing the variance that the participant and stimulus levels accounted for.
Participant-level ICC was 0.868, with a design effect of 18.354. Stimulus-level ICC was 0.132,
with a design effect of 15.819. This indicates that both participant- and stimulus- levels are
needed for the multilevel analyses.
Predictors of Interest
Participants reported low-to-moderate desire to seek new partners (M = 2.05 on a 5-point
scale; SD = 1.25). However, there were more connect-swipes than reject-swipes in general (64%
connect-swipes).
Cue Condition. Outcomes of this model indicated that cue condition had no main effect
on participants’ likelihood to swipe connect. In other words, the likelihood of swiping to connect
did not differ significantly across the three cue conditions. See Table 2(A) for the full results of
this model.

8
Seeking Motivation. Outcomes of this model indicated that participants’ desire to seek
new partners also had no main effect on their likelihood to swipe-connect. See Table 2(B) for the
full results of this model.
Cue Condition × Seeking Motivation. Outcomes of this interaction model indicated that
cue condition did not moderate the effects of seeking motivation on participants’ likelihood to
swipe-connect. See Table 2(C) for the full results of this model.
Exploratory Models
Of the exploratory models examined, the SAQ scores and the Anxious subscale of the
AAS exhibited a PIT effect on participants’ likelihood to swipe-connect between positive and
neutral conditions. In other words, when in the positive condition, participants’ likelihood to
swipe-connect decreased as SAQ score increased; whereas the reverse is true in the neutral cue
condition. A similar trend is found in the Anxious subscale of the AAS, where participants’
likelihood to swipe connect decreased as Anxious subscale score increased in the positive cue
condition, while the reverse effect is found in the neutral condition. See Figure 5 and Figure 6 for
the plotted interactions, and Table 2(D) and Table 2(E) for the full results of these models.
Awareness
Participants generally reported low awareness of the relationship between the expressions
and the black-and-white fractals (M = 0.87 on a 0-3 scale; SD = 0.86). Awareness did not have a
main or moderating effect on PIT. See Table 2(F) for the full results of this model.
Conclusion
In Experiment 1, we did not detect any main effects of the predictors of interest on
participant swipes. We did find that aspects of anxiety (social anxiety and anxious attachment)

9
interacted with cue conditions. However, this effect was not observed in the two subsequent
experiments, and thus we will not be further discussing the possible implications.
Upon reflection, we speculate that the Pavlovian training did not successfully signal the
correct reward and punishment values – i.e., we were training participants that Pavlovian cues
indicated other people’s responses but not specifically potential mates’ responses. In other
words, during Experiment 1, participants viewed facial expression gifs of both male and female
actors, instead of actors of the gender that participants are attracted to. We recognize that it is
more common for individuals to be attracted to only one gender and not the others. Therefore, a
second experiment is designed to maximize training success and to correct cue signaling. Instead
of training Pavlovian cues with male and female actors, participants would indicate the preferred
gender of their potential mates, and we would train participants only with gifs that contain actors
of participants’ dating-preference gender.
Experiment 2
Following up on Experiment 1, we designed and carried out an experiment that matched
the gif actors’ gender to the participants’ dating-preference gender. In other words, participants
who preferred to date men were trained with male gif expressions, and participants who
preferred to date women were trained with female gif expressions. This modification to the
paradigm was included specifically to improve the participants’ learning of the Pavlovian cues
and signaled values, such that the Pavlovian cues would signal a potential mate’s response, and
not just social responses in general.
Method
Participants

10
A total of 120 participants was recruited from Prolific with the same recruitment criteria
as the previous experiment – U.S. citizens between ages 18-29 inclusive. Each participant was
rewarded $3.00 for completing the experiment. Some participants were removed from analysis
due to failed attention checks, leaving a final sample size of 105 (mean age = 23.43, 47.6%
females, 51.5% Non-Hispanic White).
Materials
The same expression gifs, fractals, dating profiles, and scales were used in the current
experiment as Experiment 1.
Procedure
After providing informed consent, participants underwent a preference-targeted training
phase and a testing phase. Demographic information and individual difference measures were
obtained at the end of the experiment.
Training Phase. Before training started, participants were asked which gender they
primarily preferred to date – male or female. During training, participants viewed facial
expression gifs of either male or female actors, consistent with their dating preference gender
choice. The remaining paradigm is consistent with that of Experiment 1. Participants underwent
5 blocks of training with 12 trials per block, comprising of 60 total trials. The “loading” sign was
presented for 2.5s on top of a fractal, followed by a 3s display of gif. Attention checks were
placed in between blocks to ensure data quality.
Testing Phase. No paradigm changes were made to the testing phase in the current
experiment. Participants were asked to read dating profiles and to either choose “connect” or
“reject”. The testing consisted of 21 trials. Each trial included one dating profile displayed on top
of a previously trained fractal.

11
Analysis
The method of analyses was kept identical to that of Experiment 1. A series of
generalized linear multilevel analyses with binomial distribution was performed in R using the
“glmer” function from “lme4” package. Level 1 consisted of binary swipes with a sample size of
2,205. No missing trials were recorded. Level 2 consisted of (a) participants, with a sample size
of 105, and (b) testing stimuli, with a sample size of 21.
Our predictors of interest and exploratory predictors were also kept identical to that of
Experiment 1. Cue condition (Level 1) and participants’ desire to seek partners (Level 2a) were
our main predictors of interest. Gender, relationship status, SAQ scores, BIS/BAS scores, and
AAS scores (all Level 2a) remain our exploratory predictors, with the addition of preference-
gender (the gender that participants prefer to date). Participants’ awareness of the expression
fractal was also examined for awareness’ effect on PIT. All continuous variables were mean-
centered. Categorical predictors were treated as factors.
Results
Null Model
A null model was constructed with the following parameters: a fixed-effect intercept and
two parameters expressing the variance that the participant and stimulus levels accounted for.
Participant-level ICC was 0.853, with a design effect of 18.065. Stimulus-level ICC was 0.147,
with a design effect of 16.407. This indicates that both participant- and stimulus- levels are
needed for the multilevel analyses.
Predictors of Interest

12
Participants reported low-to-moderate desire to seek new partners (M = 2.04 on a 5-point
scale; SD = 1.20). However, there were more connect-swipes than reject-swipes in general (66%
connect-swipes).
Cue Condition. Consistent with findings from Experiment 1, outcomes of this model
indicated that cue condition had no main effect on participants’ likelihood to swipe connect. See
Table 4(A) for the full results of this model.
Seeking Motivation. Outcomes of this model indicated that participants’ desire to seek
new partners also had no main effect on their likelihood to swipe-connect. See Table 4(B) for the
full results of this model.
Cue Condition × Seeking Motivation. Outcomes of this model indicated that there is an
interaction effect – cue condition moderated the effects of seeking motivation on participants’
likelihood to swipe-connect. Specifically, when in the negative condition, participants had
increasing likelihood to connect-swipe as their seeking motivations increased; whereas the
reverse is true in the positive cue condition (when contrasting Positive vs. Negative: Log
Likelihood = –.205, SE = .100, OR = 0.815, p = .041). See Figure 7 for the plotted interaction,
and Table 4(C) for the full results of this model.
Exploratory Models
Of the exploratory models examined, the BISBAS model showed that the BAS Fun
Seeking scores had a main effect on participants’ likelihood to swipe – such that with every one-
point increase in BAS Fun Seeking score, the likelihood of swiping to connect is 1.145 times
higher (p < .001). See Table 4(D) for the full results of the BISBAS model.

13
We did not detect the same effect with SAQ and anxious attachment that were observed
in Experiment 1. In the current experiment, social anxiety and anxious attachment did not
significantly moderate the PIT effect nor did they produce main effects.
Awareness
Participants generally reported low awareness of the relationship between the expressions
and the black-and-white fractals (M = 0.95 on a 0-3 scale; SD = 0.89). Awareness did not have a
main or moderating effect on PIT. See Table 4(E) for the full results of this model.
Conclusion
In Experiment 2, we found a cue condition × seeking motivation interaction effect on
swipes. The likelihood of a connect-swipe increases as participants’ seeking motivation increases
but only in the negative cue condition. This goes against our intuition that participants are more
likely to connect-swipe in the positive condition and more likely to reject-swipe in the negative
condition, especially when their seeking motivation is high. The findings of this experiment
indicate the contrary – if a participant’s seeking motivation is low, they are more aversive of
negative cues; but if their seeking motivation is high, they are more likely to approach those
associated with negative cues. Furthermore, we did not find evidence that support findings from
Experiment 1, as both SAQ and Anxious attachment were insignificant in the current
experiment, but we did find that BAS fun-seeking score is associated with higher likelihood of a
connect-swipe. These findings highlight the importance of seeking motivation’s role on PIT
effect. However, participants’ seeking motivation remains relatively low for this experiment (M
= 2.04 on a 5-point scale; SD = 1.20). Thus, a third experiment was designed to manipulate
participants’ seeking motivation using video clips that prime positive, negative, or neutral
perceptions of romantic relationships.

14
Experiment 3
In order to further investigate the relationship between mate-seeking motivation and PIT
effect, we carried out an experiment that included a priming phase before the testing phase.
Participants were randomly assigned to either watch movie clips of couples being intimate,
neutral, or hostile towards each other. This modification to the paradigm was included in hopes
to manipulate participants’ mate-seeking motivation.
Method
Participants
A total of 120 participants was recruited from Prolific with the same recruitment criteria
as the two previous studies – U.S. citizens between ages 18-29 inclusive. Each participant was
rewarded $3.00 for completing the experiment. Some participants were removed from analysis
due to failed attention checks, leaving a final sample size of 113 (mean age = 23.19, 55.7%
females, 51.3% Non-Hispanic White).
Materials
The same expression gifs, fractals, dating profiles, and scales were used in the current
experiment as the previous studies.
Movie Clips. Six movie clips were used in this experiment to prime mate-seeking
motivation. Two romantically intimate clips were used (Doremus & Jones, 2011, 0:10:01; Webb
et al., 2009, 0:27:23), with each clip depicting a couple on a romantic date(s). Two neutral clips
were used (Krieger et al., 2012, 0:44:21; Linklater et al., 2013, 0:47:19), with one clip depicting
two people having a neutral conversation and another clip depicting two people ordering food at
a restaurant. Two hostile clips were used (Doremus & Jones, 2011, 1:04:02; Reed et al., 2006,
0:22:03), with each clip depicting a couple in a heated argument, saying hurtful things to each

15
other. Each clip lasts approximately 1-2min. Audio and subtitles were both provided. These
movie clips were pilot tested prior to the experiment to ensure that they are recognized as being
intimate, neutral, or hostile.
Procedure
After providing informed consent, participants first underwent the preference-targeted
training phase, then they were randomly assigned to one of three prime conditions – intimate,
neutral, or hostile – finally, they complete the testing phase. Demographic information and
individual difference measures were obtained at the end of the experiment.
Training Phase. The training phase was kept identical to that of Experiment 2.
Participants were shown expression gifs of either male or female actors consistent to the gender
they indicated as their dating preference. Participants underwent a total of 5 blocks of training,
12 trails per block.
Priming Phase. After training, participants were told that they will watch two movie
clips and report on how they feel towards the clips. Participants were then randomly assigned to
either intimate, neutral, or hostile prime conditions. In the intimate condition, participants
watched romantic movie clips; in the neutral condition, participants watched clips of couples
being neither intimate nor hostile; and in the hostile condition, participants watched clips of
couples fighting with each other. For each movie clip, participants were asked to answer a simple
evaluation question (e.g., “How would you describe the content of this clip?”). The purpose of
participants’ evaluation was mainly to ensure that they had paid attention and registered the
content of the prime. After the priming is complete, participants were asked to report their level
of mate-seeking motivation.

16
Testing Phase. No changes were made to the testing phase in the current experiment.
Participants read dating profiles and chose either to “connect” or “reject” the individual. Testing
consisted of 21 trials. Each trial included one dating profile displayed on top of a previously
trained Pavlovian cue.
Analysis
The method of analyses was kept identical to that of the previous two studies. A series of
generalized linear multilevel analyses with binomial distribution was performed in R using the
“glmer” function from “lme4” package. Level 1 consisted of binary swipes with a sample size of
2,373. No missing trials were recorded. Level 2 consisted of (a) participants, with a sample size
of 113, and (b) testing stimuli, with a sample size of 21.
Our predictors of interest and exploratory predictors were also kept similar to that of the
previous two studies. Cue condition (Level 1) and participants’ desire to seek partners (Level 2a)
were our main predictors of interest with the addition of prime condition (level 2a). Gender,
preference-gender, relationship status, SAQ scores, BIS/BAS scores, and AAS scores (all Level
2a) remain our exploratory predictors. Participants’ awareness of the expression fractal was also
examined for awareness’ effect on PIT. All continuous variables were mean-centered.
Categorical predictors were treated as factors.
Results
Null Model
A null model was constructed with the following parameters: a fixed-effect intercept and
two parameters expressing the variance that the participant and stimulus levels accounted for.
Participant-level ICC was 0.924, with a design effect of 19.482. Stimulus-level ICC was 0.076,

17
with a design effect of 9.653. This indicates that both participant- and stimulus- levels are needed
for the multilevel analyses.
Predictors of Interest
Participants reported moderate desire to seek new partners (M = 2.43 on a 5-point scale;
SD = 1.29). However, video clip primes did not significantly alter participants’ seeking
motivations (F = .579, p = .562). There were slightly more connect-swipes than reject-swipes in
general (60% connect-swipes).
Cue Condition. Consistent with findings from previous two studies, outcomes of this
model indicated that cue condition had no main effect on participants’ likelihood to swipe
connect. See Table 6(A) for the full results of this model.
Seeking Motivation. Unlike findings from the previous two studies, outcomes of this
model indicated that there was a main effect of participants’ desire to seek new partners on their
likelihood to swipe-connect. With each point increase (on a five-point scale) in participants’
reported desire to seek new partners, their likelihood of swiping to connect is 1.206 times higher
(p = .004). See Table 6(B) for the full results of this model.
Prime Condition. Outcomes of this model indicated that the priming video clips had no
main effect on participants’ likelihood to swipe connect. See Table 6(C) for the full results of
this model.
Cue Condition × Seeking Motivation. Outcomes of this model indicated that there was
no interaction effect. Cue condition did not moderate the effects of seeking motivation on
participants’ likelihood to swipe-connect. See Table 6(D) for the full results of this model.

18
Cue Condition × Prime Condition. Outcome of this model indicated that there was no
interaction effect. Cue condition did not moderate the effects of video priming on participants’
likelihood to swipe-connect. See Table 6(E) for the full results of this model.
Exploratory Models
Of the exploratory models examined, BAS Drive scores showed a main and moderating
effect with cue conditions on participants’ likelihood to swipe-connect. Specifically, participants’
likelihood to swipe-connect increased as their BAS Drive scores increased but only in the neutral
cue condition; in positive and negative conditions, the reverse was true. See Figure 8 for the
plotted interaction, and Table 6(F) for the full results of the model.
Awareness
Participants generally reported low awareness of the relationship between the expressions
and the black-and-white fractals (M = 0.89 on a 0-3 scale; SD = 0.88). Awareness did not have a
main or moderating effect on PIT. See Table 6(G) for the full results of this model.
Conclusion
In Experiment 3, we found a seeking motivation main effect on swipes. The likelihood of
a connect-swipe increases as participants’ overall seeking motivation increases, despite cue and
prime conditions. Even though findings of this experiment did not replicate Experiment 2, the
main finding aligns with our intuition – participants with higher mate-seeking motivation are
more likely to approach potential mates than those who have lower mate-seeking motivation.
General Discussion
In three experiments, we tested a novel social Pavlovian-to-instrumental transfer (PIT)
paradigm and examined PIT effect on participants’ mate-selection behavior: In Experiment 1, we
trained participants using black-and-white Pavlovian cues with facial expression gifs and tested

19
PIT effect using a dating task, but no main or interaction effects were detected; in Experiment 2,
we gender-matched gif actors’ genders to that of participants’ dating-preference and detected a
PIT effect moderated by seeking motivation – those who are low in seeking motivation were
more aversive of negative cues while those who are high in seeking motivation were more likely
to approach mates associated with negative cues; in Experiment 3, we attempted to manipulate
participants’ mate-seeking motivation by using video clips and found a main effect of seeking
motivation on swipes, but no PIT effect was detected.
Inconsistent findings of secondary predictors were also present in the three experiments:
In Experiment 1, the Anxious sub-scores of the AAS and the SAQ scores presented a moderating
effect – participants’ likelihood to swipe-connect decreased as SAQ and AAS subscale scores
increased in the positive condition when compared to the neutral condition; in Experiment 2,
BAS fun seeking score was found to associate with swiping tendency – participants’ likelihood
to swipe-connect increased as BAS fun seeking scores increased; in Experiment 3, BAS Drive
scores showed both main and moderating effects – participants’ likelihood to swipe-connect
increased as BAS Drive scores increased but only in the neutral condition.
The findings of these experiments do not fully support our previously posited hypotheses.
We did not detect a main PIT effect of the trained Pavlovian cues on participants’ swiping
tendencies – the cues alone did not associate with the likelihood of swipe-connects. However, in
Experiment 2, we found evidence that participants with higher seeking motivation were more
likely to approach potential mates paired with negative cues than those with lower seeking
motivation. This goes against our intuition, as we have speculated that a higher seeking
motivation would cause participants to be more aversive of the potential mates paired with

20
negative cues. Yet, when the Pavlovian cues signaled potential rejection and disapproval, they
had made participants of high seeking motivation more approach-oriented.
One possible explanation of this observed pattern stems from the rejection literature.
Rejection and ostracism research demonstrated that after experiencing social rejection,
participants exerted more effort to connect with novel individuals (e.g., Lakin et al., 2008, and
Maner et al., 2007). In this case of the current paper, one might argue that the negatively trained
Pavlovian cue reminded the participants of the disapproving (or rejecting) responses from the
training gifs and signaled a possible upcoming rejection, therefore prompting participants to
exert more effort to preemptively approach the potential mate in order to avoid the rejection.
Furthermore, those with high seeking motivation would be more driven to establish connection
and avoid rejection. Needless to say, this explanation is mere speculation so far, and would need
extensive replication and examination in upcoming studies.
In terms of the development of the social PIT training and testing paradigm, we
acknowledge that there still are many improvements to be made. One of which is the lack of
demonstration in conscious cue-outcome awareness. Despite undergoing 60 training trials (20
trials per cue condition) with multiple attention checks placed, participants’ awareness of the
cue-outcome relationship remained low throughout all three experiments (M = 0.87 for
Experiment 1, M = 0.95 for Experiment 2, M = 0.89 for Experiment 3, all on a 0-3 scale).
Perhaps it is due to this low awareness that no effects, main or moderating, were found of
awareness on PIT; as it has been previously argued (Lovibond & Shanks, 2002) and
demonstrated (Hogarth et al., 2007; Jeffs & Duka, 2017; Talmi et al., 2008; Trick et al., 2011)
that cue-outcome awareness is necessary for a PIT effect to emerge. Thus, we are unable to
provide much insight on how cue awareness affects PIT effect in participants mate-selection

21
decision-making. We are currently attempting to rectify this issue for upcoming studies by
providing more noticeable Pavlovian cues (i.e., incorporating colors and more distinct patterns).
Another caveat is that the video clip primes did not manipulate participants’ seeking
motivation as we had anticipated. We recognize that we did not take into account the nature of
relationships individuals were seeking, and thus it is possible that while the romantic and hostile
primes were effective with those who are seeking more serious relationships, they did not affect
individuals who are only seeking casual relationships. The subsequent studies we conduct will
include measurements of relationship-types that participants are seeking after, as well as provide
primes that touch on both serious and casual relationships.
One important caveat that is unrelated to the paradigm design is that participants’ seeking
motivation was relatively low with little variance during the three experiments (M = 2.05 for
Experiment 1, M = 2.04 for Experiment 2, M = 2.43 for Experiment 3, all on a 1-5 scale), and
the distribution heavily skewed right (i.e., most participants reported below-mean level of mate-
seeking motivation, with only a few reported high level of mate-seeking motivation). We
speculate that this trend of low seeking motivation is due to the COVID-19 pandemic. The three
experiments were conducted in February and March of 2021, when stay-at-home orders were in
effect and the CDC strongly recommended against socializing with those outside of one’s
household. This understandably affected the whole of U.S. population, and most individuals
were wary of social activities like meeting new people and going out on dates. For upcoming
studies that we conduct as the U.S. loosen its restrictions, we expect to see seeking motivation
increase.

22
References
Allman, M. J., DeLeon, I. G., Cataldo, M. F., Holland, P. C., & Johnson, A. W. (2010). Learning
processes affecting human decision making: An assessment of reinforcer-selective
Pavlovian-to-instrumental transfer following reinforcer devaluation. Journal of
Experimental Psychology. Animal Behavior Processes, 36(3), 402–408. DOI:
10.1037/a0017876
Caballo, V. E., Salazar, I. C., Irurtia, M. J., Arias, B., & Hofmann, S. G. (2010). Measuring
social anxiety in 11 countries: Development and validation of the social anxiety
questionnaire for adults. European Journal of Psychological Assessment, 26(2), 95-107.
DOI: 10.1027/1015-5759/a000014
Cartoni, E., Balleine, B., & Baldassarre, G. (2016). Appetitive Pavlovian-instrumental transfer:
A review. Neuroscience and Biobehavioral Reviews, 71, 829–848. DOI:
10.1016/j.neubiorev.2016.09.020
Cartoni, E., Moretta, T., Puglisi-Allegra, S., Cabib, S., & Baldassarre, G. (2015). The
relationship between specific Pavlovian instrumental transfer and instrumental reward
probability. Frontiers in Psychology, 6, 1697–1697. DOI: 10.3389/fpsyg.2015.01697
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective
responses to impending reward and punishment: The BIS/BAS scales. Journal of
Personality and Social Psychology, 67, 319-333.

23
Claes, N., Crombez, G., Franssen, M., & Vlaeyen, J. W. S. (2016). The impact of Pavlovian cues
on pain avoidance: A behavioral study. Learning and Motivation, 56, 73-83. DOI:
10.1016/j.lmot.2016.10.001
Collins, N. L. (1996). Working models of attachment: Implications for explanation, emotion, and
behavior. Journal of Personality and Social Psychology, 71(4), 810-832. DOI:
10.1037/0022-3514.71.4.810
Colwill, R. M., & Rescorla, R. A. (1988). Associations between the discriminative stimulus and
the reinforcer in instrumental learning. Journal of Experimental Psychology. Animal
Behavior Processes, 14(2), 155–164. DOI: 10.1037/0097-7403.14.2.155
Dickinson, A. & Balleine, B. (2002). The role of learning in the operation of motivational
systems. In Stevens' Handbook of Experimental Psychology, H. Pashler (Ed.). DOI:
10.1002/0471214426.pas0312
Doremus, D. (Director, Writer), & Jones, B. Y. (Writer). (2011). Like crazy [Motion picture].
USA: Paramount Pictures.
Eder, A. B., & Dignath, D. (2016). Asymmetrical effects of posttraining outcome revaluation on
outcome-selective Pavlovian-to-instrumental transfer of control in human adults.
Learning and Motivation, 54, 12-21. DOI: 10.1016/j.lmot.2016.05.002
Estes, W. K. (1943). Discriminative conditioning. I. A discriminative property of conditioned
anticipation. Journal of Experimental Psychology, 32(2), 150–155. DOI:
10.1037/h0058316

24
Garofalo, S., & Robbins, T. W. (2017). Triggering avoidance: Dissociable influences of aversive
Pavlovian conditioned stimuli on human instrumental behavior. Frontiers in Behavioral
Neuroscience, 11. DOI: 10.3389/fnbeh.2017.00063
Hogarth, L., Dickinson, A., Wright, A., Kouvaraki, M., & Duka, T. (2007). The role of drug
expectancy in the control of human drug seeking. Journal of Experimental Psychology.
Animal Behavior Processes, 33(4), 484–496. DOI: 10.1037/0097-7403.33.4.484
Holland, P. C. (2004). Relations between Pavlovian-instrumental transfer and reinforcer
devaluation. Journal of Experimental Psychology. Animal Behavior Processes, 30(2),
104–117. DOI: 10.1037/0097-7403.30.2.104
Holmes, N. M., Marchand, A. R., & Coutureau, E. (2010). Pavlovian to instrumental transfer: A
neurobehavioral perspective. Neuroscience & Behavioral Reviews, 34(8), 1277-1295.
DOI: 10.1016/j.neubiorev.2010.03.007
Huys, Q. J. M., Cools, R., Golzer, M., Friedel, E., Heinz, A., Dolan, R. J., & Dayan, P. (2011).
Disentangling the roles of approach, activation and valence in instrumental and Pavlovian
responding. PLoS Computational Biology, 7(4). DOI: 10.1371/journal.pcbi.1002028
Huys, Q. J. M., Gölzer, M., Friedel, E., Heinz, A., Cools, R., Dayan, P., & Dolan, R. J. (2016).
The specificity of Pavlovian regulation is associated with recovery from
depression. Psychological Medicine, 46(5), 1027–1035. DOI:
10.1017/S0033291715002597

25
Jeffs, S., & Duka, T. (2017). Predictive but not emotional value of Pavlovian stimuli leads to
Pavlovian-to-instrumental transfer. Behavioural Brain Research, 321(15), 214-222. DOI:
10.1016/j.bbr.2016.12.022
Knowlton, B. J., & Yin, H. H. (2006). The role of the basal ganglia in habit formation. Nature
Reviews. Neuroscience, 7(6), 464–476. DOI: 10.1038/nrn1919
Krieger, L. T. (Director), Jones, R. (Writer), & McCormack, W. (Writer). (2012). Celeste &
Jesse forever [Motion picture]. USA: Envision Media Arts.
Lakin, J. L., Chartrand, T. L., & Arkin, R. M. (2008). I am too just like you: Nonconscious
mimicry as an automatic behavioral response to social exclusion. Psychological
Science, 19(8), 816–822. DOI: 10.1111/j.1467-9280.2008.02162.x
Lehner, R., Balsters, J. H., Herger, A., Hare, T. A., & Wenderoth, N. (2017). Monetary, food,
and social rewards induce similar Pavlovian-to-instrumental transfer effects. Frontiers in
Behavioral Neuroscience, 10, 247–247. DOI: 10.3389/fnbeh.2016.00247
Linklater, R. (Director, Writer), Delpy, J. (Writer), & Hawke, E. (Writer). (2013). Before
midnight [Motion picture]. USA: Faliro House Prodcutions.
Livingstone, S.R., Russo, F.A. (2018). The Ryerson Audio-Visual Database of Emotional
Speech and Song (RAVDESS): A dynamic, multimodal set of facial and vocal
expressions in North American English. PLoS ONE 13(5): e0196391. DOI:
10.1371/journal.pone.0196391

26
Lovibond, P. F. (1981). Appetitive Pavlovian-instrumental interactions: Effects of inter-stimulus
interval and baseline reinforcement conditions. The Quarterly Journal of Experimental
Psychology. B, Comparative and Physiological Psychology, 33(4), 257–269. DOI:
10.1080/14640748108400811
Lovibond, P. F. (1983). Facilitation of instrumental behavior by a Pavlovian appetitive
conditioned stimulus. Correction to Lovibond. Journal of Experimental Psychology.
Animal Behavior Processes, 9(4), 389–389. DOI: 10.1037/0097-7403.9.4.389
Lovibond, P. F., & Shanks, D. R. (2002). The role of awareness in Pavlovian conditioning:
Empirical evidence and theoretical implications. Journal of Experimental Psychology.
Animal Behavior Processes, 28(1), 3–26. DOI: 10.1037/0097-7403.28.1.3
Maner, J. K., DeWall, C. N., Baumeister, R. F., & Schaller, M. (2007). Does social exclusion
motivate interpersonal reconnection? Resolving the "porcupine problem". Journal of
Personality and Social Psychology, 92(1), 42-55. DOI: 10.1037/0022-3514.92.1.42
Mueller, E. M., Sperl, M. F. J., & Panitz, C. (2019). Aversive imagery causes de novo fear
conditioning. Psychological Science, 30(7), 1001–1015. DOI:
10.1177/0956797619842261
Paredes-Olay, C., Abad, M. J. F., Gamez, M., & Rosas, J. M. (2002). Transfer of control
between causal predictive judgments and instrumental responding. Learning & Behavior,
30(3).

27
Quail, S. L., Laurent, V., & Balleine, B. W. (2017a). Inhibitory Pavlovian-instrumental transfer
in humans. Journal of Experimental Psychology. Animal Learning and Cognition, 43(4),
315–324. DOI: 10.1037/xan0000148
Quail, S. L., Morris, R. W., & Balleine, B. W. (2017b). Stress associated changes in Pavlovian-
instrumental transfer in humans. Quarterly Journal of Experimental Psychology, 70(4),
675–685. DOI: 10.1080/17470218.2016.1149198
Reed, P. (Director), Garelick, J. (Writer), & Lavender, J. (Writer). The break-up [Motion
picture]. USA: Universal Pictures.
Rescorla, R. A. (1994). Transfer of instrumental control mediated by a devalued
outcome. Animal Learning & Behavior, 22, 27–33. DOI: 10.3758/BF03199953
Rescorla, R. A., & Solomon, R. L. (1967). Two-process learning theory: Relationships between
Pavlovian conditioning and instrumental learning. Psychological Review, 74(3), 151–182.
DOI: 10.1037/h0024475
Rigoli, F., Pavone, E. F., & Pezzulo, G. (2012). Aversive Pavlovian responses affect human
instrumental motor performance. Frontiers in Neuroscience, 6, 134–134. DOI:
10.3389/fnins.2012.00134
Solomon, R. L., & Turner, L. H. (1962). Discriminative classical conditioning in dogs paralyzed
by curare can later control discriminative avoidance responses in the normal
state. Psychological Review, 69(3), 202–218. DOI: 10.1037/h0049385

28
Talmi, D., Seymour, B., Dayan, P., & Dolan, R. J. (2008). Human Pavlovian instrumental
transfer. The Journal of Neuroscience, 28(2), 360–368. DOI:
10.1523/JNEUROSCI.4028-07.2008
Trick, L., Hogarth, L., & Duka, T. (2011). Prediction and uncertainty in human Pavlovian to
instrumental transfer. Journal of Experimental Psychology. Learning, Memory, and
Cognition, 37(3), 757–765. DOI: 10.1037/a0022310
Watson, P., Wiers, R. W., Hommel, B., & de Wit, S. (2014). Working for food you don't desire:
Cues interfere with goal-directed food-seeking. Appetite, 79, 139–148. DOI:
10.1016/j.appet.2014.04.005
Webb, M. (Director), Neustadter, S. (Writer), & Weber, M. H. (Writer). (2009). 500 days of
summer [Motion picture]. USA: Fox Searchlight Pictures.

29
Appendix A.
Full list of dating profiles used in dating task (transfer test).

My friends tell me that I cheer up a room and have a knack for making people laugh. I’m a very
down to earth person who can have a good time doing almost anything. But I mostly love trying
out new things.

I live in a small apartment with two pups. I like to go bike riding and occasionally go camping
with my dogs, but I haven’t been very many times recently. I am a bit of a nerd and a fun person
to be around.

I am flexible, social, and very easy to get along with, but I also know what I want out of myself
and of others and life. Some of my hobbies include reading, writing, movies, hiking, and
cooking.

I love traveling and exploring new places but I also love to sit on the beach and relax. I enjoy
reading, art, photography, stand-up comedy/Improv, and playing video games with friends. I'm
also a bit of a Disney nerd.

I am obsessed with food, memes, movies, and just going out to try new things. I also enjoy being
outdoors, although between school and my part-time job, I don’t get to spend as much time
outdoors as I’d like.

Currently a student working way too hard. But I’m energetic, positive-minded, driven, passionate
about my talents. I love animals and nature! In my spare time I like to write and go on
adventures.

By nature I am a hard working yet laid back, simple but honest person whose friends would say
is especially terrible at jigsaw puzzles. I enjoy almost everything - including puzzles.

I am easygoing, friendly, caring, optimistic, and compassionate. I like spending time with
friends/family, reading books, exercising, and getting to know new people.

I’m a shy, introverted extrovert but open up if I get to know someone. I try to stay active by
playing soccer, hiking, and surfing on the weekends. I also love animals and giving back to my
community.

I will take any opportunity to be in/with nature, and I adore spending time with animals. I am a
generally more reserved person, but I do love adventures when I'm with people I'm comfortable
with.

I just started working out again and hopefully one day I'll actually enjoy running… but mostly I
like to spend my lazy days binge watching Parks and Rec while eating sushi. I love to travel, so
I'm always planning my next vacation.

30
I’m laid back, a great listener, kind, a bit sarcastic, I’ve been told I give good advice, and want to
make a difference for the time that I am here on earth. I like music, photography, and traveling.

I'm a lover of history, music, art, and music. I also find puns hilarious. Nothing makes me
happier than coming back from my morning run in the park. I will try just about anything once,
and I thrive on meeting new people.

I spend most of my time drawing, woodworking, writing, or playing cards with friends. I work
on exhibits as my job and am always interested in museums, art exhibitions, or individual artists.

My interests include book clubs, board game nights, working out, cooking, and volunteering. I
really enjoy going to concerts when I have the time, and I try to explore as many local musicians
as possible.

I love music, working out, eating too much carbs, going on adventures, learning new things, and
having engaging conversations! Also, I'm a little bit sarcastic and tend to laugh out loud at bad
jokes.

I enjoy nice dinners, viewing art galleries and gardens, grabbing a nice coffee, volunteering in
my community, and reading classic literature. I am quite an old soul actually, but still humorous
and love to joke around!

I am passionate about life, my family, and my career. More than that though, I try to enjoy the
simple things in life. I like to keep myself busy, but there's nothing more exciting than having a
whole day to do nothing!

I'm always down for anything whether it's going on a hike or staying in all day to binge Netflix.
I'm very easy to get along with. One fun fact about me is that I'm good with numbers, but can't
spell to save my life.

I have a great sense of humor and a big heart. I joke around a lot but I can get quite serious when
it comes to doing work. My friends would describe me as goofy and clumsy.

I am a middle child of a big family of 7. Growing up with my siblings made me outgoing and
very social. I'm almost done with school and am ready for my next adventure, whatever it may
be.

31
Table 1.

List of Variables and Measurements for Experiment 1.

Variable Name Level Measurement and Coding
Dependent Variable
Swipe Level 1 Binary 1/0 for connect/reject
Predictors of Interest
Cue Condition Level 1 Positive, neutral, and negative
Seeking Motivation Level 2a From 1-5; mean-centered
Secondary Predictors
Gender Level 2a Female and Male
Relationship Status Level 2a Single and Coupled
Awareness Level 2a From 0-4 items correct; mean-centered
SAQ score Level 2a Continuous; mean-centered
BIS/BAS scores Level 2a Continuous; mean-centered
AAS scores Level 2a Continuous; mean-centered
Note. Level 1 consisted of swipes. Level 2a consisted of participants. Level 2b consisted of the
testing stimuli (i.e., dating profiles participants read).

32
Table 2.

Experiment 1 Model Outputs. Binary Logistic Multilevel Modeling on Swipe.

Parameter Estimate SE Exp(B) p
A. Cue Condition model
Negative vs. Neutral Cue .069 .216 1.072 .748
Positive vs. Neutral Cue .084 .216 1.087 .699
B. Seeking Motivation model
Seeking Motivation –.020 .077 0.980 .792
C. Cue Condition × Seeking
Negative vs. Neutral Cue .069 .216 1.072 .748
Positive vs. Neutral Cue .082 .216 1.086 .704
Seeking Motivation –.035 .923 0.966 .704
Negative vs. Neutral : Seeking <.001 .900 1.000 .999
Positive vs. Neutral : Seeking .045 .902 1.046 .617
D. SAQ model
Negative vs. Neutral Cue .070 .217 1.072 .748
Positive vs. Neutral Cue .087 .217 1.091 .690
SAQ score .007 .005 1.007 .175
Negative vs. Neutral : SAQ –.008 .005 0.992 .115
Positive vs. Neutral : SAQ –.013 .005 0.987 .012*
E. AAS model
Negative vs. Neutral Cue .066 .217 1.068 .761
Positive vs. Neutral Cue .075 .217 1.078 .728
AAS Anxious Score .033 .022 1.033 .138
AAS Avoidant Score –.027 .014 0.973 .058
Negative vs. Neutral : Anxious –.026 .022 0.975 .245
Positive vs. Neutral : Anxious –.044 .022 0.957 .047*
Negative vs. Neutral : Avoidant .003 .014 1.003 .854
Positive vs. Neutral : Avoidant .018 .014 1.018 .212
F. Awareness model
Negative vs. Neutral Cue .070 .216 1.072 .746
Positive vs. Neutral Cue .086 .216 1.089 .692
Awareness –.004 .134 0.996 .974
Negative vs. Neutral : Awareness –.025 .130 0.975 .847
Positive vs. Neutral : Awareness –.067 .130 0.936 .609
Note. *p<.05, **p<.01, ***p<.001.

33
Table 3.

List of Variables and Measurements for Experiment 2.

Variable Name Level Measurement and Coding
Dependent Variable
Swipe Level 1 Binary 1/0 for connect/reject
Predictors of Interest
Cue Condition Level 1 Positive, neutral, and negative
Seeking Motivation Level 2a From 1-5; mean-centered
Secondary Predictors
Gender Level 2a Female and Male
Preferred Gender Level 2a Female and Male
Relationship Status Level 2a Single and Coupled
Awareness Level 2a From 0-4 items correct; mean-centered
SAQ score Level 2a Continuous; mean-centered
BIS/BAS scores Level 2a Continuous; mean-centered
AAS scores Level 2a Continuous; mean-centered
Note. Level 1 consisted of swipes. Level 2a consisted of participants. Level 2b consisted of the
testing stimuli (i.e., dating profiles participants read).

34
Table 4.

Experiment 2 Model Outputs. Binary Logistic Multilevel Modeling on Swipe.

Parameter Estimate SE Exp(B) p
A. Cue Condition model
Negative vs. Neutral Cue –.193 .179 0.824 .280
Positive vs. Neutral Cue –.323 .178 0.724 .070
B. Seeking Motivation model
Seeking Motivation .058 .072 1.060 .416
C. Cue Condition × Seeking
Negative vs. Neutral Cue –.187 .179 0.829 .297
Positive vs. Neutral Cue –.324 .179 0.723 .069
Seeking Motivation .055 .094 1.057 .558
Negative vs. Neutral : Seeking .112 .103 1.119 .277
Positive vs. Neutral : Seeking –.092 .100 0.912 .357
D. BIS/BAS model
BAS Drive –.014 .038 0.986 .709
BAS Fun .142 .041 1.152 <.001***
BAS Reward Responsiveness –.054 .043 0.947 .209
BIS .003 .020 1.003 .895
E. Awareness model
Neutral vs. Negative Cue –.192 .179 0.825 .283
Positive vs. Negative Cue –.319 .178 0.727 .073
Awareness –.025 .122 0.975 .835
Neutral vs. Negative : Awareness –.045 .130 0.956 .728
Positive vs. Negative : Awareness –.110 .129 0.896 .393
Note. *p<.05, **p<.01, ***p<.001.

35
Table 5.

List of Variables and Measurements for Experiment 3.

Variable Name Level Measurement and Coding
Dependent Variable
Swipe Level 1 Binary 1/0 for connect/reject
Predictors of Interest
Cue Condition Level 1 Positive, neutral, and negative
Seeking Motivation Level 2a From 1-5; mean-centered
Prime Condition Level 2a Intimate, neutral, and hostile
Secondary Predictors
Gender Level 2a Female and Male
Preferred Gender Level 2a Female and Male
Relationship Status Level 2a Single and Coupled
Awareness Level 2a From 0-4 items correct; mean-centered
SAQ score Level 2a Continuous; mean-centered
BIS/BAS scores Level 2a Continuous; mean-centered
AAS scores Level 2a Continuous; mean-centered
Note. Level 1 consisted of swipes. Level 2a consisted of participants. Level 2b consisted of the
testing stimuli (i.e., dating profiles participants read).

36
Table 6.

Experiment 3 Model Outputs. Binary Logistic Multilevel Modeling on Swipe.

Parameter Estimate SE Exp(B) p
A. Cue Condition model
Negative vs. Neutral Cue .111 .164 1.118 .499
Positive vs. Neutral Cue .061 .164 1.063 .711
B. Seeking Motivation model
Seeking Motivation .189 .066 1.208 .004**
C. Prime Condition model
Intimate vs. Neutral Prime –.157 .213 0.855 .460
Hostile vs. Neutral Prime –.151 .217 0.860 .486
D. Cue Condition × Seeking Motivation
Negative vs. Neutral Cue .114 .165 1.120 .490
Positive vs. Neutral Cue .065 .165 1.068 .691
Seeking Motivation .110 .082 1.116 .182
Negative vs. Neutral : Seeking .100 .086 1.105 .246
Positive vs. Neutral : Seeking .141 .086 1.152 .100
E. Cue Condition × Prime Condition
Negative vs. Neutral Cue –.015 .228 0.985 .946
Positive vs. Neutral Cue –.109 .227 0.897 .630
Intimate vs. Neutral Prime –.234 .262 0.791 .373
Hostile vs. Neutral Prime –.375 .268 0.687 .162
Negative vs. Neutral : Intimate vs. Neutral .069 .266 1.071 .796
Positive vs. Neutral : Intimate vs. Neutral .161 .265 1.175 .544
Negative vs. Neutral : Hostile vs. Neutral .320 .273 1.377 .241
Positive vs. Neutral : Hostile vs. Neutral .355 .272 1.426 .192
F. Cue Condition × BIS/BAS model
Negative vs. Neutral Cue .113 .165 1.119 .496
Positive vs. Neutral Cue .059 .165 1.063 .719
BAS Drive .117 .052 1.124 .024*
BAS Fun –.025 .050 0.976 .622
BAS Reward Responsiveness .016 .052 1.016 .760
BIS .023 .031 1.023 .469
Negative vs. Neutral : BAS Drive –.136 .053 0.873 .011*
Positive vs. Neutral : BAS Drive –.147 .053 0.863 .005**
Negative vs. Neutral : BAS Fun .025 .051 1.025 .623
Positive vs. Neutral : BAS Fun .036 .050 1.036 .478
Negative vs. Neutral : BAS RR .080 .054 1.084 .134
Positive vs. Neutral : BAS RR .064 .053 1.066 .231
Negative vs. Neutral : BIS –.057 .032 0.945 .077
Positive vs. Neutral : BIS –.041 .032 0.960 .202
G. Awareness model
Negative vs. Neutral Cue .111 .165 1.118 .499
Positive vs. Neutral Cue .062 .164 1.064 .707

37
Parameter Estimate SE Exp(B) p
Awareness .110 .128 1.116 .392
Negative vs. Neutral : Awareness .041 .130 1.041 .755
Positive vs. Neutral : Awareness –.097 .129 0.907 .452
Note. RR = Reward Responsiveness. *p<.05, **p<.01, ***p<.001.

38
Figure 1.

Gifs Used for Pavlovian Training.

Note. The twelve gifs used for training. Selected gifs were from two male actors (left two
columns) and two female actors (right two columns). The gifs exhibited three expressions:
welcoming (top row), neutral (middle row), and disapproving (bottom row).

Male Actor 1 Male Actor 2 Female Actor 1 Female Actor 2
Welcoming
Neutral
Disapproving

39
Figure 2.
Fractals Used for Training and Testing.

Note. Black-and-white fractals used for the experiment. Panel (a) shows the polka dot fractal;
panel (b) shows the zebra print fractal; panel (c) shows the fingerprint fractal.

(a)
(b)
(c)

40
Figure 3.
Process of One Training Trial.

Note. Visual representation of one training trial. A fractal appears for 2.5s, then the facial
expression gif appears for approximately 3s.

Time (s)
0 2.5 5.5

41
Figure 4.
Process of Testing Trials.

Note. Visual representation of the testing phase. Previously trained black-and-white fractals were
randomly paired with dating profiles. Dating profiles’ order of presentation was pseudo-
randomized.

. . .
Profiles (N=21)

42
Figure 5.
Likelihood of Swiping “Connect” by Participants’ SAQ Score Moderated by Cue Condition.

43
Figure 6.
Likelihood of Swiping “Connect” by Participants’ Anxious Subscale Score of AAS Moderated by
Cue Condition.

44
Figure 7.

Likelihood of Swiping “Connect” by Participants’ Motivation to Seek New Partners Moderated
by Cue Condition.

45
Figure 8.

Likelihood of Swiping “Connect” by Participants’ BAS Drive Score Moderated by Cue
Condition.

Abstract (if available)

Abstract Pavlovian-to-instrumental transfer (PIT) effect suggests that external cues can signal potential reward/punishment values, subsequently affecting decision-making and behavior. Despite extensive research on PIT effect in humans, there is still a lack of research examining PIT in social contexts. Therefore, the current paper aims to develop a social behavioral PIT task in three experiments, in which participants are trained with social outcomes and tested with a social task. Participants first learned the association between neutral cues and facial expressions, and then they completed a dating task where participants read dating profiles of potential mates with previously trained cues as profile backgrounds. Although the results from the three experiments did not sufficiently replicate one another, our findings suggest that participants’ seeking-motivation played a crucial role in how they utilize Pavlovian cues in their mate selection behavior. Future direction and further improvement needed for the social behavioral PIT task are discussed.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Investigating the effects of Pavlovian cues on social behavior

PDF

Using Pavlovian cues to boost self-control

PDF

Dopamine dependent: examining the link between learning and treatment-resistant depression

PDF

Jilted and tilted: an exploration of post-rejection response and introduction of a novel experimental paradigm

PDF

Rewards bias self-evaluations of ability

PDF

Social exclusion decreases risk-taking

PDF

How misinformation exploits moral values and framing: insights from social media platforms and behavioral experiments

PDF

Generalization of social rejection in social networks

PDF

On second thought, people choose to be prosocial

PDF

Insula activity during safe-sex decision-making in sexually risky men suggests negative urgency and fear of rejection drives risky sexual behavior

PDF

The cost of missing objectives in multiattribute decision modeling

PDF

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

PDF

#BLM or #ALM: accessible perspective shapes downstream judgment even among people high in social dominance

PDF

The effect of present moment awareness and value intervention of ACT on impulsive decision-making and impulsive disinhibition

PDF

Taking the temperature of the Columbia Card Task

PDF

Are jokes funnier when they’re easier to process?

PDF

The role of music training on behavioral and neurophysiological indices of speech-in-noise perception: a meta-analysis and randomized-control trial

PDF

Facebook habits: rewards, cues, and automaticity

PDF

Evaluating standard error estimators for multilevel models on small samples with heteroscedasticity and unbalanced cluster sizes

PDF

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

Asset Metadata

Creator Qiao, Aili (author)

Core Title Examining Pavlovian-to-instrumental transfer effect in human mate selection behavior

Contributor Electronically uploaded by the author (provenance)

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Psychology

Degree Conferral Date 2021-12

Publication Date 09/29/2021

Defense Date 09/28/2021

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

Tag decision-making,Learning,multilevel modeling,OAI-PMH Harvest,Pavlovian conditioning

Format application/pdf (imt)

Language English

Advisor Read, Stephen (committee chair), Hackel, Leor (committee member), Monterosso, John (committee member)

Creator Email aqiao@usc.edu,qiaoalice@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15961508

Unique identifier UC15961508

Legacy Identifier etd-QiaoAili-10110

Document Type Thesis

Format application/pdf (imt)

Rights Qiao, Aili

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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

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

Repository Email cisadmin@lib.usc.edu