ArticlePDF Available

Sadism and Aggressive Behavior: Inflicting Pain to Feel Pleasure

Authors:

Abstract

Sadism is a “dark” trait that involves the experience of pleasure from others’ pain, yet much is unknown about its link to aggression. Across eight studies (total N = 2,255), sadism predicted greater aggression against both innocent targets and provocateurs. These associations occurred above-and-beyond general aggressiveness, impulsivity, and other “dark” traits. Sadism was associated with greater positive affect during aggression, which accounted for much of the variance in the sadism–aggression link. This aggressive pleasure was contingent on sadists’ perceptions that their target suffered due to their aggressive act. After aggression, sadism was associated with increases in negative affect. Sadism thus appears to be a potent predictor of aggression that is motivated by the pleasure of causing pain. Such sadistic aggression ultimately backfires, resulting in greater negative affect. More generally, our results support the crucial role of anticipated and positive forms of affect in motivating aggression. © 2018 by the Society for Personality and Social Psychology, Inc.
Running head: SADISM AND AGGRESSION 1
Sadism and Aggressive Behavior: Inflicting Pain to Feel Pleasure
David S. Chester1*, C. Nathan DeWall2, Brian Enjaian2
1Department of Psychology, Virginia Commonwealth University, USA
2Department of Psychology, University of Kentucky, USA
in press at Personality and Social Psychology Bulletin
Main Text Word Count (minus Figures and Tables): 7,928
Abstract Word Count: 143
*Correspondence should be addressed to:
David S. Chester
302 Thurston House
Virginia Commonwealth University
Richmond, VA, 23284, USA
1-804-828-7624
dschester@vcu.edu
SADISM AND AGGRESSION 2
Abstract
Sadism is a ‘dark’ trait that involves the experience of pleasure from others’ pain, yet
much is unknown about its link to aggression. Across eight studies (total N=2,255),
sadism predicted greater aggression against both innocent targets and provocateurs.
These associations occurred above-and-beyond general aggressiveness, impulsivity,
and other ‘dark’ traits. Sadism was associated with greater positive affect during
aggression, which accounted for much of the variance in the sadism-aggression link.
This aggressive pleasure was contingent on sadists’ perceptions that their target
suffered due to their aggressive act. After aggression, sadism was associated with
increases in negative affect. Sadism thus appears to be a potent predictor of aggression
that is motivated by the pleasure of causing pain. Such sadistic aggression ultimately
backfires, resulting in greater negative affect. More generally, our results support the
crucial role of anticipated and positive forms of affect in motivating aggression.
Keywords: sadism, aggression, dark tetrad, positive affect, emotion
SADISM AND AGGRESSION 3
Introduction
The wholesome pleasure of prosocial acts must contend with humankind’s
darker delights. Some people exhibit sadism, which involves “the deliberate infliction of
pain for the sake of enjoyment” (pp. 227; Nell, 2006). Sadistic tendencies are not purely
the domain of violent criminals but appear among non-clinical and non-criminal
populations (Buckels, Jones, & Paulhus, 2013). Preliminary evidence links sadism to
aggression, yet many aspects of sadistic aggression remain incompletely understood.
This investigation examined the robustness of the sadism-aggression link across a
variety of contexts, targets, and operationalizations of aggression. Further, we
examined whether sadists enjoyed the aggressive act and whether the victim’s pain
caused this aggressive pleasure.
Sadism: The Pleasure of Inflicting Pain
Sadism is a constellation of personality traits that are characterized by the
tendency to enjoy the suffering of others (Baumeister, 1997; Nell, 2006). Rather than
passively taking pleasure in others’ pain, sadists actively perpetrate harm, motivated by
the enjoyment of the aggressive act and the painful outcome (O’Meara, Davies, &
Hammond, 2011). In the past, sadism was as a clinically-diagnosable form of
psychopathology, yet such diagnoses have now changed (e.g., sexual sadism disorder;
Krueger, 2010). More contemporary approaches to sadism conceptualize it as a
continuously-distributed facet of ‘dark’ personality that extends beyond forensic and
clinical samples into the broader distribution of humankind (Buckels et al., 2013;
Chabrol, Van Leeuwen, Rodgers, & Sejourne, 2009; O’Meara et al., 2011).
The ‘Dark Triad’ and Forms of Aggression
SADISM AND AGGRESSION 4
Clues about sadism’s link to aggression are likely to come from the research on
the ‘dark triad’: Machiavellianism (manipulating others to fulfill selfish goals), Narcissism
(holding grandiose and vulnerable views of the self), and psychopathy (sensation-
seeking and callous disregard for others; Furnham, Richards, & Paulhus, 2013; Paulhus
& Williams, 2002). The dark triad often predict aggressive traits (Jonason & Webster,
2010) and acts (e.g., bullying; Baughman, Dearing, Giammarco & Vernon, 2012).
However, the type of aggression is critical to understanding the links between the dark
triad and harm-doing. Harming innocent targets in the absence of provocation is
deemed proactive aggression, which can be juxtaposed against reactive aggression
that takes the form of retaliation against perceived provocateurs (Raine et al., 2006).
These forms of aggression are not mutually exclusive and correlate strongly= (Miller &
Lynam, 2006). Narcissism is associated with greater reactive aggression, particularly
retaliatory acts in response to ego threats (Bushman & Baumeister, 1998; c.f.
Kirkpatrick, Waugh, Valencia, & Webster, 2002). Psychopathy is linked to both proactive
(Porter & Woodworth, 2006; Raine et al., 2006) and (to a lesser extent) reactive
aggression (Reidy, Zeichner, & Martinez, 2008; Reidy, Zeichner, Miller, & Martinez,
2007). When compared in a meta-analytic framework, psychopathy had the strongest
association with proactive aggression followed by Machiavellianism, whereas
Narcissism was unassociated with proactive aggression (Webster et al., 2014). Some
scholars argue that the dark triad form the latent basis of aggressive dispositions and
replace the unitary construct of ‘trait aggression’ (Paulhus, Curtis, & Jones, 2018).
Sadism is considered part of these dark traits, forming a ‘dark tetrad’ (Paulhus, 2014).
SADISM AND AGGRESSION 5
Although the links between the dark triad and aggression have been well-established,
the potential role of sadism in aggression is less understood.
Preliminary Evidence for the Sadism-Aggression Link
Nascent research has hinted at a link between sadism and aggression. For
instance, sadism correlates positively with trait physical aggression and is a core feature
of trait revenge-seeking (Chester & DeWall, 2018). However, such correlations use self-
report instead of overt behavior. Initial research on the relationship between sadism and
aggressive behavior focused on the infliction of harm upon innocent targets. For
example, sadism is linked to harming insects and innocent humans who refuse to
retaliate (Buckels et al., 2013). Implicit sadism was associated with greater electric
shocks administered to an innocent target (Reidy, Zeichner, & Seibert, 2011). Sadism is
also uniquely associated with self-reported acts of sexual violence (Russell & King,
2016), tendencies towards antisocial vices (Jonason, Zeigler-Hill, & Okan, 2017), as
well as self-reports of conventional and online forms of bullying and ‘trolling’ behavior
(Buckels, Trapnell, & Paulhus, 2014; ; March, Grieve, Marrington, & Jonason, 2017).
Despite this array of research, further confirmation is necessary to establish
sadism’s link to aggression. One reason for this uncertainty is that only a handful of
studies have assessed sadism’s link to actual behavioral measures of aggression.
Among these few studies, aggression was operationalized proactively, as the targets of
aggression were innocent victims (e.g., Buckels et al., 2013). Aggression is more often
reactive than proactive because provocation is the most reliable situational predictor of
aggression (Anderson & Bushman, 2002). Sadism’s link to reactive aggression remains
uncertain, and therefore, the sadism-aggression link is uncertain. Despite this lack of
SADISM AND AGGRESSION 6
evidence, there are good theoretical reasons to expect that sadism would be associated
with reactive aggression, largely centering on the role of positive affect.
The Role of Positive Affect in Aggression
Traditionally, aggression was theorized to arise from negatively-valenced
affective states such as frustration and pain (Berkowitz, 1989). More recently, a wealth
of evidence has arisen to support the role of positively-valenced affect in motivating
revenge and retaliatory aggressive behavior (Chester, 2017). For example, reading
about acts of retaliatory aggression induces positive affect (Eadeh, Peak, & Lambert,
2016). Actual acts of retaliatory aggression are associated with activity in the brain’s
reward network (Chester & DeWall, 2016) and genetic profiles that modulate pleasure-
seeking (Chester et al., 2015, 2016). This hedonic reward appears to form a positive
feedback loop in which acts of violence beget even more acts of violence (Martens,
Kosloff, Greenberg, Landau, & Schmader, 2007). The perceived and ephemeral ability
of aggression to regulate and improve aversive affective states further fuels this cyclical
aggression (Bushman, Baumeister, & Phillips, 2001; Chester & DeWall, 2017;
Gollwitzer & Bushman, 2012). This positive feedback loop may even explain the
development of stable, sadistic tendencies.
Such a positive feedback loop meshes well with the General Aggression Model
(Anderson & Bushman, 2002; DeWall, Anderson, & Bushman, 2011). Specifically,
sadism acts as a personality input variable that increases the likelihood of aggression
through the internal route of positive affect. Such a positive, rewarding experience
informs appraisal and decision processes when sadists encounter potential victims,
making them more likely to engage in impulsive acts. Those impulsive acts, in turn,
SADISM AND AGGRESSION 7
influence how sadists approach their future social encounters, strengthening the
knowledge structures that form the basis of how they interpret and react to events in
their social world. The next section fleshes out this potential tendency for sadism to
experience the pleasure of aggression.
Sadism and the Pleasure of Aggression
Some individuals are more prone than others to experience the pleasure of
aggression and sadism measures are designed to capture this variability (Chester &
DeWall, 2018). Only preliminary evidence exists for sadism’s link to aggression-related
positive affect (i.e., aggressive pleasure). Sadism was positively-correlated with more
enjoyment of killing insects (Buckels et al., 2013) and ‘trolling’ others online (Buckels,
Trapnell, Andjelovic, & Paulhus, in press; Buckels et al., 2014). However, little evidence
links sadism to the pleasure of harming other individuals across proactive and reactive
forms of aggression. It also remains uncertain from what aspect of the aggressive act
do sadists derive pleasure. Theoretical accounts of sadism invoke the suffering of the
victim as the source of aggressive pleasure (Baumeister, 1997; Nell, 2006), yet there is
no evidence for this proposal. Such evidence is necessary to determine the very nature
of the sadism construct. Further, the timecourse of aggressive pleasure remains
uncertain, whether it arises during or after the aggressive act, and how long the feeling
lasts after the aggressive act.
Present Research
To fill these gaps in the literature, the present research tested the over-arching
hypotheses that (A & B) sadism would be associated with greater proactive and reactive
aggressive behavior, (C & D) sadism would be associated with greater positive affect
SADISM AND AGGRESSION 8
during and after the aggressive act, and (E) such aggressive pleasure would be
contingent upon the actual suffering of the intended target. In an exploratory fashion, we
also examined the role of sadism in negative affect during and after aggressive acts.
To test these hypotheses, we conducted eight studies in which we measured
participants’ dispositional sadism and gave them an opportunity to act aggressively,
either in response to or in the absence of provocation. Many of these studies included
measures of negative and positive affect during and after aggression, experimental
manipulations of interpersonal provocation, measures of crucial covariates to ensure the
specificity of the sadism-aggression link, and variations in the victim’s level of suffering
due to the aggressive act.
Statistical Power Statement
Meta-analytic estimates of the mean effect size for aggression studies in
personality and social psychology yield r=.24 (k=3,323; Richard, Bond, & Stokes-Zoota,
2003). Therefore, samples of 130 or more participants have at least 80% power to
detect main effects of this magnitude or larger. A priori power analyses were not used to
determine the sample sizes in any of the eight studies, though they surpass this sample
size threshold (excepting Study 3, N=126). Some of our more complex inferential tests
(e.g., moderation, indirect effects) may be powered below the 80%.
Open Science Statement
All data files needed to reproduce these results can be publicly-accessed
(https://osf.io/fjwhc/files/) and all research materials are available in the associated
Methodology Attachment document.
Study 1
SADISM AND AGGRESSION 9
Study 1 tested the hypothesis that sadism would predict greater retaliatory
aggression. Participants were provoked and reported their sadism, as well as their trait
self-control and impulsivity in order to test whether sadistic aggression is driven by
general impulsivity. Psychopathy was also measured in order to assess sadism’s link to
aggression, above-and-beyond the dark triad (Buckels et al., 2013). To test whether
individuals enjoyed the aggressive act, we measured participants positive and negative
affect after the aggression measure.
Methods
Participants
Participants were 162 undergraduates (116 females, 42 males, 4 missing gender
data).
Materials
Participants completed the Brief Self-Control Scale (BSCS; Tangney,
Baumeister, & Boone, 2004), positive and negative affect items from the Need Threat
Scale (NTS; Williams, 2009), Levenson Self-Report Psychopathy Scale (LSRP;
Levenson, Kiehl, & Fitzpatrick, 1995), Short Sadistic Impulse Scale (SSIS; O'Meara et
al., 2011), and the UPPS-P Impulsivity Scale (Lynam, Smith, Whiteside, & Cyders,
2006; Whiteside & Lynam, 2001).
Procedure
This study was part of a larger project on the role of psychostimulants on
aggression. Participants arrived at our laboratory where they were randomly assigned to
receive either a capsule containing 100mg of caffeine, a placebo capsule, or no
capsule. Participants who received a capsule were blind to its contents. Participants
SADISM AND AGGRESSION 10
reported their baseline affective state and watched nature videos for 30 minutes.
Participants were then experimentally provoked (as in Pedersen, Gonzales, & Miller,
2000). To do so, participants were asked to complete a list of difficult and impossible
anagrams. The experimenter repeatedly interrupted the participant and expressed
frustration with their poor performance, eventually ending the task prematurely.
Participants then completed the Taylor Aggression Paradigm (TAP) against another
same-gender undergraduate student.
The TAP is a well-validated measure of behavioral, retaliatory aggression framed
as a competitive reaction time game played over the internet with a fictitious opponent
(Anderson & Bushman, 1997; Chester & Lasko, in press; Giancola & Chermack, 1998;
Taylor, 1967). For each of the 17 trials of the task, participants set the volume (60 105
decibels) and duration (0 5 seconds) of an aversive noise blast that their opponent
ostensibly heard if participants won the competition (i.e., press a button faster). Within
the volume and duration settings, responses were coded along a 1 (lowest volume,
shortest duration) - 10 (highest volume, longest duration) score gradient. A non-
aggression option was also provided (coded as 0). The order of participant wins and
losses were randomized and then held constant across all participants, excepting when
participants failed to respond in time (in which they automatically lost the trial to ensure
believability). Participants’ opponents always selected the loudest and longest noise
blast on the first trial in order to provoke participants. This approach to setting the wins,
losses, and opponent’s noise blast settings was used in all subsequent studies that
employed the TAP. Finally, participants reported their affective state again and then
reported their sadism and psychopathy.
SADISM AND AGGRESSION 11
Results
Descriptive Statistics
All self-report measures exhibited sufficient internal consistency except the SSIS
(Supplemental Table 1), which was largely driven by a single, reverse-scored item “I
wouldn’t intentionally hurt anyone.We re-calculated the SSIS without this item and
performed analyses using the new, 9-item scale (Supplemental Table 1). We averaged
volume and duration levels across all 17 trials (as recommended by Chester & Lasko, in
press). Descriptive statistics are summarized in Supplemental Table 1. Zero-order
correlations between all study variables are summarized in Supplemental Table 2.
Correlations with Aggressive Behavior
Sadism was associated with greater aggressive behavior, r(145)=.25, p=.003,
and this association remained after controlling for trait self-control, primary and
secondary psychopathy, and all five facets of impulsivity (Table 1). We used two
contrast codes to examine potential effects of the pill condition. The first code
contrasted the effect of the caffeine condition (contrast weight: 1) against the placebo
condition (contrast weight: -1), while not modeling the no pill condition (contrast weight:
0). The second code contrasted the effect of taking either the caffeine (contrast weight:
1) or placebo (contrast weight: 1) pill against the no pill condition (contrast weight: -2).
Table 1. Sadism’s association with aggressive behavior on the Taylor Aggression
Paradigm in Study 1. Gender is coded: male=1, female=-1.
Model
Predictor
β
t
df
p
ΔR2
1
Sadism
.25
3.05
145
.003
2
Caffeine vs. Placebo Pill
.05
0.58
143
.562
SADISM AND AGGRESSION 12
Pill vs. No Pill
-.07
-0.79
143
.432
Sadism
.23
2.85
143
.005
.05
3
Gender
.13
1.44
144
.152
Sadism
.19
2.08
144
.039
.03
4
Caffeine vs. Placebo Pill
.05
0.58
142
.564
Pill vs. No Pill
-.06
-0.78
142
.437
Self-Control
.00
0.05
142
.962
Sadism
.23
2.78
142
.006
.05
5
Caffeine vs. Placebo Pill
.00
0.01
141
.993
Pill vs. No Pill
-.04
-0.48
141
.632
Psychopathy - Primary
.19
1.86
141
.065
Psychopathy - Secondary
.04
0.43
141
.669
Sadism
.18
2.02
141
.045
.03
6
Caffeine vs. Placebo Pill
.03
0.30
138
.768
Pill vs. No Pill
-.04
-0.53
138
.596
Lack of Perseverance
-.18
-1.92
138
.057
Lack of Premeditation
.00
0.01
138
.994
Negative Urgency
.11
0.83
138
.411
Positive Urgency
.02
0.17
138
.864
Sensation Seeking
.01
0.08
138
.935
Sadism
.25
2.97
138
.004
.06
Correlations with Post-Aggression Affect
SADISM AND AGGRESSION 13
Sadism was associated with greater post-aggression negative and positive affect
(Supplemental Table 2). After controlling for baseline negative affect, post-aggression
negative affect was no longer associated with sadism, r(125)=.14, p=.126. After
controlling for baseline positive affect, post-aggression positive affect was no longer
associated with sadism, r(125)=-.13, p=.136.
Study 2
Study 2 sought to replicate Study 1 using a different aggression measure: the hot
sauce paradigm. Study 2 also tested whether sadism was associated with either
proactive or reactive aggression or both, and whether sadism’s link to such aggression
would occur even if participants had to experience the same suffering as their victim.
Methods
Participants
Participants were 168 undergraduates (118 females, 47 males, 3 missing gender
data). Participants were excluded if they had a relevant food allergy.
Procedure
Participants arrived at the laboratory where they were randomly assigned to be
either socially rejected or accepted via two same-gender students in the Cyberball
paradigm (version 4.0; Williams et al., 2012). Next, participants retrospectively reported
their negative and positive affect during the Cyberball task using the Need Threat Scale,
completed the Short Sadistic Impulse Scale and Self-Report Psychopathy Scale, and
then completed the hot sauce aggression paradigm (Lieberman, Solomon, Greenberg,
& McGregor, 1999) against one of their Cyberball partners. Participants were told that
they were going to complete a separate study on taste preferences. Each participant
SADISM AND AGGRESSION 14
completed a short questionnaire about their own food preferences, which were
supposedly exchanged with one of their Cyberball partners. Participants tasted the hot
sauce and then assisted the experimenters by measuring the same hot sauce for one of
their Cyberball partners, using the partner’s food questionnaire to guide their decision.
Just prior to administering the hot sauce to their partner, participants were randomly
assigned to be told that they would either have to eat as much hot sauce as they
allocated to their Cyberball partner, or not. After allocating the hot sauce to their partner,
participants again reported their current negative and positive affect using the Need
Threat Scale.
Results
Descriptive Statistics
Nineteen participants were excluded from analyses because they previously
completed the Cyberball paradigm. All self-report measures exhibited sufficient internal
consistency, except the Secondary Psychopathy subscale of the LSRP and the SSIS
(Supplemental Table 3), which was again largely driven by the single, reverse-scored
item. As in Study 1, we used the SSIS without this item. Hot sauce allocations were
positively skewed (skew=3.21) but not zero-inflated (5.4% zeroes). Therefore we
conducted base 10 logarithmic transformations of these values (adding 1 beforehand to
ensure that 0 values would still be transformed; as in DeWall, Twenge, Bushman, Im, &
Williams, 2010; Webster & Kirkpatrick, 2006). This approach reduces problems with
skew and kurtosis in aggression data (Chester & Lasko, in press). Descriptive statistics
are summarized in Supplemental Table 3 and zero-order correlations between all study
variables are summarized in Supplemental Table 4.
SADISM AND AGGRESSION 15
Correlations with Aggressive Behavior
Sadism was unassociated with greater hot sauce allocations, r(147)=.16, p=.053.
This null effect that was not moderated by the Cyberball manipulation or whether
participants believed that they would have to consume as much hot sauce as they
allocated to their partner (Table 2). Sadism was unassociated with aggression after
controlling for primary and secondary psychopathy (Table 2).
Table 2. Sadism’s association with aggressive behavior on the Hot Sauce
Aggression Paradigm in Study 2. Gender is coded: male=1, female=-1. Rejection
condition is coded: rejection=1, acceptance=-1.
Predictor
β
t
df
p
ΔR2
Sadism
.16
1.95
147
.053
Gender
.19
2.27
146
.025
Sadism
.12
1.48
146
.141
.01
Rejection Condition
.17
0.65
145
.519
Sadism
.13
1.38
145
.169
.02
Rejection x Sadism
-.22
-0.82
145
.416
.00
Self-Harm Condition
-.04
-0.17
145
.865
Sadism
.16
1.90
145
.060
.03
Self-Harm x Sadism
.02
0.09
145
.925
.00
Rejection Condition
-.06
-0.77
143
.442
Self-Harm Condition
-.02
-0.27
143
.790
Psychopathy - Primary
.12
1.27
143
.205
Psychopathy - Secondary
.11
1.19
143
.236
SADISM AND AGGRESSION 16
Sadism
.07
0.69
143
.489
.00
Correlations with Post-Aggression Affect
Sadism was significantly associated with greater negative but not positive affect
after aggression (Supplemental Table 4). Sadism remained associated with greater
negative affect after aggression after controlling for prior negative affect, r(145)=.20,
p=.015. Conversely, sadism remained unassociated with positive affect after aggression
after controlling for prior positive affect, r(145)=-.07, p=.403.
Study 3
Study 3 sought to replicate Study 2 using an online aggression measure.
Methods
Participants
Participants were 126 undergraduates (62 females, 64 males).
Procedure
Participants completed the study online, in which they were randomly assigned to
be either socially rejected or accepted via two same-gender students in the Cyberball
paradigm (version 4.0; Williams et al., 2012). Afterwards, participants retrospectively
reported their negative and positive affect during Cyberball via the Need Threat Scale
and then completed an image-assignment aggression measure (as in Gollwitzer &
Bushman, 2012). To do so, participants chose the number of “images that depict
aversive scenes such as homicide crime scenes, rotting animal carcasses, and
traumatic injuries” for one of their Cyberball partners to see (between 0 and 9 images).
Participants were randomly assigned to be told that the person who would view these
images was either one of their Cyberball partners or a new person (i.e., an innocent
SADISM AND AGGRESSION 17
target). Participants again reported their levels of negative and positive affect via the
Need Threat Scale and completed the Short Sadistic Impulse Scale.
Results
Descriptive Statistics
Descriptive statistics are summarized in Supplemental Table 5 and zero-order
correlations between all study variables are summarized in Supplemental Table 6.
Correlations with Aggressive Behavior
Sadism was associated with greater numbers of gruesome images assigned to
be viewed by another person, r(123)=.18, p=.049. The association between sadism and
gruesome image allocation was not moderated by whether participants had been
experimentally rejected or whether their aggression was directed at their Cyberball
partners or a new person (Table 3).
Table 3. Sadism’s association with aggressive behavior in Study 3. Gender is
coded: male=1, female=-1. Rejection condition is coded: rejection=1,
acceptance=-1; Retaliatory condition is coded: retaliatory=1, non-retaliatory=-1.
Predictor
β
t
df
p
ΔR2
Sadism
.18
1.99
123
.049
Gender
.03
0.30
122
.533
Sadism
.17
1.90
122
.060
.03
Rejection Condition
.40
1.93
121
.056
Sadism
.16
1.80
121
.074
.03
Rejection x Sadism
-.35
-1.70
121
.091
.02
Retaliatory Condition
-.59
-2.76
121
.007
SADISM AND AGGRESSION 18
Sadism
.03
0.31
121
.761
.02
Retaliatory x Sadism
.43
2.01
121
.047
.03
Rejection Condition
.44
1.63
117
.105
Retaliatory Condition
-.56
-2.58
117
.011
Sadism
.03
0.30
117
.767
Rejection x Sadism
-.36
-1.48
117
.142
Retaliatory x Sadism
.42
1.93
117
.056
Rejection x Retaliatory
-.19
-0.62
117
.535
Rejection x Retaliatory x
Sadism
.10
0.35
117
.730
.00
Correlations with Post-Aggression Affect
Sadism was associated with greater negative affect after aggression
(Supplemental Table 6), even after controlling for prior negative affect, r(118)=.24,
p=.008. Sadism was not associated with less positive affect after aggression,
(Supplemental Table 6), even after controlling for prior positive affect, r(118)=-.11,
p=.232.
Study 4
Study 4 sought to replicate the previous studies with an array of aggression
measures, while also including a measure of affect during the aggressive act.
Methods
Participants
SADISM AND AGGRESSION 19
Participants were 211 undergraduates (132 females, 73 males, 6 missing gender
data
1
).
Measures
Participants completed the History of Physical Fights Scale (Chester & Lasko, in
press) and an ad hoc questionnaire in which participants retrospectively reported
various positively-valenced feelings they experienced during an act of aggression. Such
retrospective affect reports are quite accurate (Harmon-Jones, Bastian, & Harmon-
Jones, 2016) and allowed us to test whether sadism was linked to positive affect in the
midst aggression, not just afterwards. Thirty six, positively-valenced items were
acquired from the NTS, PANAS, Profile of Mood States (Curran, Andrykowski, & Studts,
1995), Discrete Emotions Questionnaire (Harmon-Jones et al., 2016), and generated
independently by the authors (list of items available in Supplemental Table 7).
Participants retrospectively reported whether they experienced these 36 affective states
during the Taylor Aggression Paradigm (“indicate the extent to which this [affective
state] described how you felt when your opponent received the noise blasts that you
picked in the competitive reaction-time task”). An exploratory factor analysis (EFA)
detailed in Study 7 produced a 9-item measure of Aggressive Pleasure from this larger
item-set (final items available in Supplemental Table 8). Study 7 was used for this EFA
instead of Study 4 given its much larger sample size.
Procedure
1
Aggression data from these participants has been reported elsewhere (Chester & Lasko, in press), but
not in the context of sadism.
SADISM AND AGGRESSION 20
Participants arrived individually to the laboratory. In order to experimentally-
induce retaliatory aggression, this study used an essay evaluation paradigm in which
participants received harsh or positive feedback on an essay (Bushman & Baumeister,
1998; Chester & DeWall, 2017). The essay evaluation contained either negative (8/35
points, “One of the WORST essays I’ve EVER read!”) or positive (33/35 points, “Great
essay!”) feedback, as determined by random assignment. Participants then completed a
25-trial version the Taylor Aggression Paradigm, which was otherwise identical to the
task used in Study 1, the Hot Sauce Aggression Task (as in Study 2), and the Voodoo
Doll Aggression Task (VDAT). The VDAT presents participants with a virtual
representation of a human target that participants are given an opportunity to
symbolically harm that person by harming the doll (DeWall et al., 2013). Participants
selected the number of virtual, sharp pins that they wished to stick into a plush human
doll as a symbolic representation of their actual essay evaluator (from 0 to 51 pins).
Finally, participants completed a battery of questionnaires, which included the Positive
Affect During Aggression Scale and the Short Sadistic Impulse Scale
Results
Descriptive Statistics
Hot sauce allocations were leptokurtic (kurtosis=2.77) but not zero-inflated
(10.3% zeroes). Voodoo doll pin counts were positively skewed (skew=2.18), leptokurtic
(kurtosis=4.37), and zero-inflated (36.9% zeroes). Therefore we conducted logarithmic
transformations of these values (as in Study 2). Because of the extensive zero-inflation
in Voodoo Doll pin counts, we adopted generalized linear modeling that specified a
Poisson distribution (as recommended by DeWall et al., 2013). This Poisson approach
SADISM AND AGGRESSION 21
was also adopted for reports of violence over the past year and past 5 years from the
History of Physical Fights Scale, due to their extensive zero inflation: 65.5% zeroes and
86.8% zeroes, respectively. Descriptive statistics are summarized in Supplemental
Table 9. Zero-order correlations between all study variables are summarized in
Supplemental Table 10. The nine-item version of the SSIS was adopted due to internal
consistency issues, as in Studies 1 and 2.
Correlations with Aggressive Behavior and Pleasure
Sadism was associated with administering louder and longer noise blasts to
essay evaluators, r(180)=.19, p=.011, though not with greater hot sauce allocations,
r(175)=.06, p=.442. Using Poisson modeling, sadism was associated with greater
voodoo doll pin counts, B=0.35, Χ2(1,170)=69.95, p<.001, and greater frequencies of
physical fights over the past five years, B=1.10, Χ2(1,180)=54.20, p<.001, and past
year, B=0.76, Χ2(1,177)=61.18, p<.001. The sadism-aggression link was not moderated
by prior provocation, excepting the case of voodoo doll pin counts, in which provocation
attenuated the effect of sadism on pin counts (Table 4). As evidence for the construct
validity of our sadism measure, sadism was associated with greater reports of pleasure
during noise blast administration, r(180)=.19, p=.009.
Table 4. Sadism’s association with aggressive behavior in Study 4. Gender is
coded: male=1, female=-1. Provocation condition is coded: provoked=1,
unprovoked=-1.
Aggression
Measure
Predictor
β
t
df
p
ΔR2
Hot Sauce
Sadism
.06
0.77
175
.442
SADISM AND AGGRESSION 22
Gender
.20
2.63
174
.009
Sadism
.01
0.18
174
.861
.00
Provocation Condition
-.05
-0.21
173
.833
Sadism
.05
0.65
173
.519
.00
Provocation x Sadism
.05
0.23
173
.817
.00
TAP
Sadism
.19
2.58
180
.011
Gender
-.11
-1.43
179
.154
Sadism
.21
2.84
179
.005
.04
Provocation Condition
.29
1.41
178
.160
Sadism
.18
2.34
178
.020
.03
Provocation x Sadism
-.12
-0.58
178
.586
.00
Voodoo Doll
Sadism
.14
1.83
170
.069
Gender
-.22
-2.86
169
.005
Sadism
.19
2.44
169
.016
.03
Provocation Condition
.66
3.19
168
.002
Sadism
.16
2.03
168
.044
.01
Provocation x Sadism
-.44
-2.08
168
.039
.02
Structural Equation Modeling: Estimating Direct and Indirect Effects of Sadism
and Aggressive Pleasure on a Latent Form of Aggressive Behavior
We used structural equation modeling to examine whether the pleasure of the
aggressive act accounted for the sadism-aggression link (Figure 1). All three aggression
measures were modeled onto a latent aggression factor. The two physical fight
variables were too zero-inflated to warrant inclusion in this parametric model. Bias-
SADISM AND AGGRESSION 23
corrected bootstrapping was used to estimate the indirect effect and its associated 95%
confidence interval (500 bootstrap samples; via AMOS 24.0 software). The variance of
the latent aggression factor was pre-set to 1 to allow for the estimation of all paths.
Overall the model fit the data well, Χ2(4)=3.64, p=.458; CFI=1.00; NFI=0.94;
RMSEA=.00 (90% CI=.00, .11); TLI=1.02, though hot sauce allocations failed to
significantly load onto the latent aggression factor. Sadism was linked to greater
aggressive behavior through greater pleasure during the aggressive act [indirect effect:
B=.05, SE=.03, 95% CI=.01, .14, p=.013]. Thus, sadism’s link to aggression is
explained, in part, by the enjoyment of the aggressive act.
Figure 1. Structural equation model from Study 4 modeling the indirect effect of
sadism on aggression via greater aggressive pleasure. Values above paths
represent standardized coefficients and values attached to variables represent
residual, unstandardized variances. Dashed paths are non-significant. *p<.05,
**p<.01, ***p<.001.
SADISM AND AGGRESSION 24
Study 5
Study 5 sought to replicate the previous studies while including a baseline affect
measure that was missing from Study 4. Without a baseline affect estimate, the results
of Study 4 could be due to underlying differences in affect levels between individuals
high and low in sadism. Study 5 included measures of negative and positive affect
before, during, and after aggression. Study 5 also included a measure of trait
aggression as an additional control variable.
Methods
Participants
Participants were 156 adult participants recruited from Amazon’s Mechanical
Turk subject pool (75 females, 80 males, 1 missing gender data). Participants were
compensated with $0.50
2
.
Materials
Participants completed the 12-item Brief Aggression Questionnaire, a short-form
of the 29-item Buss-Perry (1982) Aggression Questionnaire (BAQ; Webster et al.,
2013).
Procedure
Participants were randomly assigned to be provoked or not through an online
version of the essay-evaluation paradigm employed in Study 4. After the essay task,
participants reported their current negative and positive affect using the Need Threat
2
Aggression and affect data from these participants has been reported elsewhere (Chester & DeWall,
2017), but not in the context of sadism.
SADISM AND AGGRESSION 25
Scale, and then completed the Voodoo Doll Aggression Task (target was ‘your essay
evaluator’). After confirming their pin count, participants retrospectively reported their
negative and positive affect experienced during the Voodoo Doll Aggression Task and
then reported their current negative and positive affect, in both cases using variants of
the Need Threat Scale. Finally, participants completed the Short Sadistic Impulse Scale
and Brief Aggression Questionnaire. Embedded in this battery of questionnaires was a
single quality check item that asked participants to pick a specific number from a
number array.
Results
Descriptive Statistics
Three participants failed the quality check and were removed from all subsequent
analyses. Voodoo doll pin counts were not excessively skewed (skew=1.54) or kurtotic
(kurtosis=1.02) but they were zero-inflated (50.3% zeroes). To address this zero-
inflation, we adopted a Poisson analytic approach. Descriptive statistics are
summarized in Supplemental Table 11. Zero-order correlations between all study
variables are summarized in Supplemental Table 12.
Correlations with Aggressive Behavior
Sadism was associated with a greater number of pins stuck in the voodoo doll,
which remained statistically significant after controlling for gender and the four facets of
trait aggression (Table 5). As in Study 4, the effect of sadism on aggression was
attenuated by the provocation manipulation (Table 5).
SADISM AND AGGRESSION 26
Table 5. Sadism’s association with aggressive behavior on the Voodoo Doll
Aggression Task in Study 5. Gender is coded: male=1, female=-1. Provocation
condition is coded: provoked=1, unprovoked=-1.
Model
Predictor
B
Χ2
p
1
Sadism
.37
497.89
<.001
2
Gender
-.30
33.63
<.001
Sadism
.38
520.53
<.001
3
Provocation Condition
2.28
311.20
<.001
Sadism
0.22
131.07
<.001
Provocation x Sadism
-0.35
82.87
<.001
4
Provocation Condition
1.28
364.94
<.001
Anger
0.05
4.01
.045
Hostility
0.25
109.24
<.001
Physical Aggression
0.14
66.98
<.001
Verbal Aggression
-0.19
57.42
<.001
Sadism
0.14
39.01
<.001
Correlations with Affect During Aggression
An indirect effect analysis (using 5,000 bias-corrected and accelerated bootstrap
samples via the PROCESS macro for SPSS v.3.1, model 4; Hayes, 2012) showed that
the direct effect of sadism on log-transformed aggression scores was explained, in part,
by positive affect experienced during the aggressive act [indirect effect: B=0.04,
SE=0.02, 95% CI=0.01, 0.08] but not through negative affect experienced during the
aggressive act [B=-0.00, SE=0.01, 95% CI=-0.01, 0.01] (difference between these
SADISM AND AGGRESSION 27
indirect effects, B=0.04, SE=0.02, 95% CI=0.01, 0.08), controlling for post essay
feedback negative and positive affect (Figure 2).
Figure 2. Statistical model from Study 5, modeling the indirect effect of sadism on
aggression through greater positive affect during aggression (key paths
highlighted in black). Values represent unstandardized regression coefficients,
dashed lines represent non-significant effects, value in parentheses is direct
effect after controlling for indirect effect, *p<.05, **p<.01, ***p<.001.
Correlations with Post-Aggression Affect
Sadism was associated with greater negative affect after aggression
(Supplemental Table 12), even after controlling for prior negative affect, r(149)=.26,
p=.001. Sadism was unassociated with positive affect after aggression (Supplemental
Table 12), even after controlling for prior positive affect, r(149)=.06, p=.456.
Study 6
Study 6 sought to replicate the previous studies using a different affect measure.
Further, Study 6 manipulated self-regulatory fatigue, given evidence linking this
experience to heightened aggression (Denson, DeWall, & Finkel, 2012).
SADISM AND AGGRESSION 28
Methods
Participants
Participants were 238 undergraduates (176 females, 62 males).
Materials
Participants completed the Positive Affect Negative Affect Schedule (PANAS), a
validated measure of current levels of positive and negative affect (i.e., mood; Watson,
Clark, & Tellegen, 1988).
Procedure
Participants arrived at the laboratory where they were randomly assigned to write
an essay for 5 minutes about a personally meaningful event. Participants were
randomly assigned to not use the letters X or Z (control condition) or A or N (fatigue
condition; as in Mead, Baumeister, Gino, Schweitzer, & Ariely, 2009). After doing so,
participants completed the PANAS and then completed a series of benign, cognitive
tasks that were related to a larger project on self-regulation (e.g., Stroop Task). Then,
participants completed the Voodoo Doll Aggression Task against an imagined person
from participants’ real lives that they “feel a great amount of anger towards.”
Participants again reported their affective state during and after aggression via the
PANAS, and then completed the Short Sadistic Impulse Scale.
Results
Descriptive Statistics
Voodoo doll pin counts were not excessively skewed (skew=1.49) or kurtotic
(kurtosis=0.98) but they were zero-inflated (42.2% zeroes). Because of the extensive
zero-inflation, we adopted a Poisson approach. Descriptive statistics are summarized in
SADISM AND AGGRESSION 29
Supplemental Table 13. Zero-order correlations between all study variables are
summarized in Supplemental Table 14. The nine-item version of the Short Sadistic
Impulse Scale was adopted due to internal consistency issues, as in Studies 1, 2, and
4.
Correlations with Aggressive Behavior
Aggressive behavior was associated with sadism, which was observed after
controlling for gender (Table 6). As in Studies 4 and 5, the sadism-aggression link was
attenuated by the study’s experimental manipulation, in this case, of self-regulatory
fatigue (Table 6).
Table 6. Sadism’s association with aggressive behavior on the Voodoo Doll
Aggression Task in Study 6. Gender is coded: male=1, female=-1. Fatigue
condition is coded: fatigued=1, unfatigued=-1.
Model
Predictor
B
Χ2
p
1
Sadism
0.47
233.33
<.001
2
Gender
-0.15
9.65
.002
Sadism
0.51
235.20
<.001
3
Fatigue Condition
0.67
43.06
<.001
Sadism
0.21
18.46
<.001
Fatigue x Sadism
-0.52
69.83
<.001
Correlations with Affect During Aggression
The effect of sadism on aggression, was explained, in part, by positive affect
experienced during the aggressive act [B=0.04, SE=0.03, 95% CI=0.01, 0.10] but not
through negative affect experienced during the aggressive act [B=-0.00, SE=0.01, 95%
SADISM AND AGGRESSION 30
CI=-0.03, 0.03] (though the difference between these indirect effects was not significant,
B=0.04, SE=0.03, 95% CI=-0.11, 0.01), controlling for baseline affect (Figure 3).
Figure 3. Statistical model from Study 6, modeling the indirect effect of sadism on
aggression through greater positive affect during aggression (key paths
highlighted in black). Values represent unstandardized regression coefficients,
dashed lines represent non-significant effects, value in parentheses is direct
effect after controlling for indirect effect, *p<.05, **p<.01, ***p<.001.
Correlations with Post-Aggression Affect
Sadism was associated with greater negative affect after aggression
(Supplemental Table 14), even after controlling for prior negative affect, r(234)=.16,
p=.016. Sadism was unassociated with positive affect after aggression (Supplemental
Table 14), even after controlling for prior negative affect, r(234)=-.03, p=.602.
Study 7
Study 7 served to replicate the previous studies employing a rejection
manipulation, while also creating a novel state-level measure of pleasure experienced
during aggression that was employed in Studies 4, 7, and 8. This study also combined
SADISM AND AGGRESSION 31
multiple measures of sadism (to ensure that our effects were not specific to an
individual measure of this construct), as well as assessments of the dark triad, to ensure
the existence of the sadism-aggression link above-and-beyond these constructs.
Finally, this study created a novel state-level measure of perceptions that the victim of
an aggressive act suffered, in order to investigate the idea that aggressive pleasure is
derived directly from the pain inflicted on the victim by the aggressive act.
Methods
Participants
Participants were 388 undergraduates (249 females, 134 males).
Materials
Participants completed the Assessment of Sadistic Personality (ASP; Plouffe,
Saklofske, & Smith, 2017), the Comprehensive Assessment of Sadistic Tendencies
scale (CAST; Buckels et al., in press), the Dirty Dozen Questionnaire (DD; Jonason &
Webster, 2010), and the Victim Suffering Scale (VSS), in which participants were
instructed to respond to the extent to which they perceived that the aggression they
inflicted upon their opponent (via the Voodoo Doll Aggression Task) resulted in their
victim’s actual pain and suffering. They did so across 10 ad hoc statements (for list of
items see Supplemental Table 15), to which they responded along a 1 (strongly
disagree) to 7 (strongly agree) scale.
Procedure
Participants completed the study online where they were randomly assigned to
be either socially rejected or accepted via two same-gender students in the Cyberball
paradigm (version 4.0; Williams et al., 2012). After the Cyberball task, participants
SADISM AND AGGRESSION 32
completed the Voodoo Doll Aggression Task, with one of their Cyberball partners as the
target. Participants then completed a battery of questionnaires that also included the
Positive Affect During Aggression Scale (prompt: Indicate the extent to which this
statement described how you felt when you picked how many pins to stick in the doll.”;
e.g., Delighted, Serene), and the Short Sadistic Impulse Scale. Embedded in the battery
of questionnaires were two quality check items, which instructed participants to select a
specific number from a number array.
Results
Descriptive Statistics
Thirty-two participants failed at least one of the two quality checks and were
subsequently excluded from all analyses. A 35-item sadism index was computed by
averaging standardized responses from all 20 items of the ASP, all 10 items of the
SSIS, and the 5-item Direct Physical subscale of the CAST. The Vicarious and Direct
Verbal subscales of the CAST were excluded as they were not relevant to the direct and
physical forms of sadistic aggression examined in this study.
Voodoo doll pin counts were not excessively skewed (skew=1.90) or leptokurtic
(kurtosis=2.71) but they were zero-inflated (47.5% zeroes). Because of the extensive
zero-inflation, we adopted a Poisson approach. Descriptive statistics are summarized in
Supplemental Table 16. Zero-order correlations between all study variables are
summarized in Table 17.
Exploratory Factor Analysis Positive Affect During Aggression Scale
To assess the psychometric properties of the PADAS, we conducted an iterated
EFA (via SAS 9.4), using direct oblimin rotation (δ=0), which allowed for the extraction
SADISM AND AGGRESSION 33
of correlated components. Five factors were retained based on the results of a parallel
analysis (Horn, 1965). An aggressive pleasure’ subscale comprised nine items that
exhibited substantial loadings onto this factor that were equal to or greater than +/-.40
(12.94% variance explained; Supplemental Table 8). None of these items exhibited
problematic cross-factor loadings (+/- .20). The second and fifth factors were labeled as
the ‘calmness’ (6-items [2 excluded due to high cross-factor loadings]; 8.50% variance
explained) and ‘arousal’ (4 items; 7.39% variance explained) subscales, respectively.
The third (5 items [2 excluded due to high cross-factor loadings]; 7.63% variance
explained) and fourth factors (4 items [2 excluded due to high cross-factor loadings];
8.11% variance explained) did not exhibit a coherent or theoretically-sensible
conceptual theme.
Exploratory Factor Analysis Victim Suffering Scale
An identical EFA was applied to the responses to the 10 original items of the
Victim Suffering Scale. Only two factors yielded items with substantial loadings
(Supplemental Table 15). The two reverse-coded items loaded onto one factor that was
discarded (1.10% variance explained), whereas the other eight items loaded onto
another factor that was retained (6.64% variance explained) and exhibited excellent
internal consistency (Supplemental Table 16).
Correlations with Aggressive Behavior
Aggressive behavior on both the VDAT and HPFS were associated with greater
aggression, which was observed after controlling for the dark triad (Table 7).
SADISM AND AGGRESSION 34
Table 7. Sadism’s association with aggressive behavior in Study 7. Rejection
condition is coded: rejection=1, acceptance=-1. Gender is coded: male=1,
female=-1.
Predictor
B
Χ2
df
Sadism
0.38
244.91
1, 354
Gender
0.51
172.79
1, 348
Sadism
0.32
166.40
1, 348
Rejection
0.24
33.23
1, 352
Sadism
0.42
223.56
1, 352
Rejection x
Sadism
0.20
12.27
1, 352
Rejection
0.28
55.40
1, 350
Machiavellianism
0.07
13.32
1, 350
Narcissism
-0.13
76.69
1, 350
Psychopathy
-0.07
7.73
1, 350
Sadism
0.44
204.04
1, 350
Effect of Victim Suffering Manipulation on Aggressive Pleasure
We tested the moderating ability of perceived victim suffering using the same
PROCESS macro as detailed in Study 6; model 1). Sadism’s association with
aggressive pleasure experienced during aggression was magnified by perceived victim
suffering, B=0.17, t(352)=2.04, p=.043, 95% CI=0.01, 0.34 (Figure 4). At relatively high
(+1 SD) levels of perceived victim suffering, sadism was positively associated with
aggressive pleasure, B=0.62, t(352)=3.89, p<.001, 95% CI=0.30, 0.93. However, this
SADISM AND AGGRESSION 35
effect was absent at relatively low (-1 SD) levels of perceived victim suffering, B=0.19,
t(352)=0.90, p=.371, 95% CI=-0.23, 0.61. Similar interactions were not observed with
the calmness, B=0.06, t(352)=0.63, p=.532, 95% CI=-0.13, 0.25, or arousal, B=0.08,
t(352)=1.05, p=.296, 95% CI=-0.07, 0.23, subscales of the PADAS.
Figure 4. Interactive effect from Study 7 whereby the association between sadism
and the aggressive pleasure experienced during aggression is magnified by
perceived victim suffering. Bands around regression lines represent 95%
confidence intervals.
Study 8
Study 8 extended beyond Study 7’s correlational evidence for the critical role of
victim suffering in sadistic aggression by manipulating how much participants perceived
that the victim of their aggression was truly harmed by the aggressive act and then
measuring the extent to which they experienced pleasure in response to the suffering.
Additionally, this study employed different measures of sadism and the dark triad as
control variables.
SADISM AND AGGRESSION 36
Methods
Participants
Participants were 207 undergraduates (166 females, 41 males).
Materials
Participants completed the Short Dark Triad Scale (SD3; Jones & Paulhus, 2014)
and the Varieties of Sadistic Tendencies Scale (VAST; Paulhus & Jones, 2015).
Procedure
Participants arrived at the laboratory where they completed the VAST, Short
Sadistic Impulse Scale, and the SD3 scale. Participants then completed a 25-trial
version of the Taylor Aggression Paradigm (as in Study 4). Participants were randomly
assigned to hear feedback from their opponent in the aggression paradigm that either
indicated they suffered due to the task’s noise blasts (“Those noise blasts were
unbearable! They were so loud they gave me a migraine!”) or that the partner did not
suffer (“Those noise blasts were nothing! Mostly, they were just annoying.”).
Participants completed the aggressive pleasure subscale of the Positive Affect During
Aggression Scale, which was modified to assess levels of currently-felt pleasure that
lingered after the aggressive act. Finally, participants completed a manipulation check
(Supplemental Table 15).
Results
Descriptive Statistics
A 19-item sadism index was computed by averaging standardized responses
from all 10 items of the SSIS and the 9-item Direct Sadism subscale of the VAST. We
excluded the Vicarious Sadism subscale of the VAST because it was irrelevant to the
SADISM AND AGGRESSION 37
direct and physical form of sadistic aggression measured in this study. We averaged
both volume and duration levels across all 25 trials (as recommended by Chester &
Lasko, in press). Descriptive statistics are summarized in Supplemental Table 18. Zero-
order correlations between all study variables are summarized in Supplemental Table
19.
Manipulation Check
As predicted, participants in the victim suffering condition reported more victim
suffering (M=2.51, SD=1.45) than did participants in the no suffering condition (M=1.83,
SD=1.09), t(203)=3.73, p<.001, d=0.52, 95% CI=.25, .80.
Correlations with Aggressive Behavior
Aggressive behavior on the TAP was not associated with sadism, r(202)=.04,
p=.620, even after controlling for gender and the dark triad (Table 8).
Table 8. Sadism’s association with aggressive behavior in Study 8, separated by
measure. Victim pain condition is coded: victim pain=1, no pain=-1. Gender is
coded: male = 1, female=-1.
Predictor
β
t
df
p
ΔR2
Sadism
.04
0.50
202
.620
Gender
.06
0.83
201
.405
Sadism
.02
0.25
201
.802
.00
Victim Pain
.06
0.84
200
.401
Sadism
.04
0.54
200
.588
.00
Victim Pain * Sadism
-.05
-0.65
200
.519
.00
Victim Pain
.02
0.32
198
.749
SADISM AND AGGRESSION 38
Machiavellianism
.06
0.66
198
.511
Narcissism
.23
3.20
198
.002
Psychopathy
.10
1.06
198
.289
Sadism
-.07
-0.81
198
.422
.00
Effect of Victim Suffering Manipulation on Aggressive Pleasure
Sadisms association with post-aggression pleasure was magnified by the victim
suffering manipulation, B=0.47, t(201)=2.43, p=.016, 95% CI=0.09, 0.85 (Figure 5).
Among participants in the suffering condition, sadism was unassociated with aggressive
pleasure, B=0.37, t(201)=1.56, p=.121, 95% CI=-0.10, 0.85. Yet among participants in
the no suffering condition, sadism was negatively associated with aggressive pleasure,
B=-1.51, t(201)=-2.31, p=.022, 95% CI=-2.80, -0.22.
Figure 5. Interactive effect from Study 8 whereby the association between sadism
and the aggressive pleasure experienced during aggression is magnified by an
experimental manipulation of victim suffering. Bands around regression lines
represent 95% confidence intervals.
SADISM AND AGGRESSION 39
Internal Meta-Analysis
An internal, random-effects meta-analysis across the eight studies’ zero-order
correlations between sadism and aggressive behavior was performed using JASP v.9.0
(effect sizes [n=8], study of origin, and corresponding sample sizes are listed in
Supplemental Table 20). Study 4 employed five aggression measures and in order to
avoid issues with dependency between the associations, we selected the correlation
between sadism and the TAP due to this measure’s established validity. Using
restricted maximum likelihood estimation, we observed a modest correlation between
sadism and aggressive behavior, r=.20, SE=0.04, 95% CI=.12, 29, Z=4.70, p<.001
(Figure 6). The effects included in this internal meta-analysis exhibited significant
heterogeneity, Q(7)=20.05, p=.005. An integrative data analysis (Curran & Hussong,
2009) replicated this meta-analytic effect and demonstrated that this effect is curvilinear,
with the sadism-aggression link becoming less positive at higher levels of sadism
(Supplemental Document 1).
SADISM AND AGGRESSION 40
Figure 6. Forest plot of sadism-aggression effects from all eight studies. Numbers
on the left represent the study of origin for each effect and values on the right
represent individual effect sizes and their associated 95% confidence intervals.
Discussion
What makes someone evil? For many, a central feature includes the tendency to
cause harm to others for one’s own enjoyment. These sadistic impulses do not purely
reside in the gray matter of deranged killers, but can be found in the general human
populace (Buckels et al., 2013). We sought to contribute to the understanding of such
‘everyday sadism’, testing whether these tendencies could actually be used to predict
SADISM AND AGGRESSION 41
aggressive behavior. To do so, we examined the robustness of the sadism-aggression
link across myriad aggression measures and towards innocent and provocative targets.
We also tested two core tenets of the sadism construct, that sadism is associated with
subjective pleasure during acts of aggression and that this pleasure is derived from the
suffering inflicted on others (Baumeister, 1999: Chabrol et al., 2009).
Is Sadism Linked to Aggression?
Across eight studies, participants’ self-reported sadism was positively associated
with greater administrations of aversive noise blasts, painfully spicy hot sauce,
gruesome images, and sharp pins administered to other people. Sadism was also linked
to violent acts perpetrated in participants’ real-world recent histories. These
associations between sadism and aggressive behavior were robust, remaining reliable
after controlling for poor self-control, impulsivity, trait aggression, and the dark triad of
Machiavellianism, Narcissism, and psychopathy. This wealth of evidence replicates and
extends upon previous work that highlights the discriminant validity of sadism in its role
as a correlate of greater aggressive behavior (e.g., Buckels et al., 2013; Chester &
DeWall, 2017, 2018; Reidy et al., 2011). Further, sadism was linked to aggression
within both males and females. Given that sadism is higher among males (Buckels et
al., 2013), it was important to rule out this possibility. These results support sadism’s
robust (though modestly-sized) effect on aggressive behavior.
Towards Whom is Sadistic Aggression Directed?
Sadism was associated with both retaliatory aggression towards provocateurs
and also towards innocent targets. In two studies, sadistic aggression was more
strongly directed at innocent individuals. These findings replicate previous work linking
SADISM AND AGGRESSION 42
sadism to aggression against innocent individuals (Buckels et al., 2013), provide a novel
extension by linking sadism to retaliatory forms of aggression, and suggest that sadistic
forms of aggression are largely numb to situational inputs that normally magnify
aggression. The inability of these provocations to amplify sadists’ aggression suggests
a different motivation than revenge, and may be rooted in the anticipated affect
surrounding sadistic acts.
Is the Aggressive Behavior of Sadists Linked to the Experience of Pleasure?
Although we initially expected that sadists would experience greater positive
affect after an aggressive act (as in Buckels et al., 2013), we routinely observed that
sadism was unassociated with such post-aggression positive affect. This discrepancy
may be due to the fact that the targets of aggression in the studies performed by
Buckels and colleagues (2013) were pill bugs and not humans. Conversely, sadism was
most often associated with greater negative affect after aggression. This association
was not simply due to sadists’ general tendency to experience greater negative affect,
as we statistically controlled for baseline affect. It appears that, while sadists appear to
be more aggressive, these aggressive acts seem to have a detrimental impact on their
mood. In line with recent research on aggression’s perceived emotion-regulating
qualities (Chester & DeWall, 2017), sadists may perceive as aggression an effective
means to improve their mood, despite its contrary results.
We developed a new self-report measure of aggressive pleasure, which
demonstrated that sadism was associated with greater pleasure during the aggressive
act. Further, such aggressive pleasure accounted for a significant portion of the effect of
sadism on aggression. Sadism may thus be reinforced by experiences of aggressive
SADISM AND AGGRESSION 43
pleasure and this reinforcement may serve as a proximate mechanism by which
aggressive and sadistic traits and tendencies are formed and reinforced over time
(Chester, Lynam, Milich, & DeWall, 2018). These findings fit within the growing literature
that establishes positively-valenced affective states as a potent motivator of aggression
(Chester, 2017). The findings are also in line with predictions from the General
Aggression Model in terms of how personality input variables can increase the likelihood
of aggression through affect, appraisal and decision-making processes, and feedback
loops (Anderson & Bushman, 2002; DeWall et al., 2011).
Do Sadists Derive Pleasure from Others’ Suffering?
By measuring and manipulating how much the victims of participants’ aggression
were perceived to experience actual suffering, we established that aggressive pleasure
is contingent upon the perceived suffering of sadists’ victims. This core feature of
sadism has been theorized (Baumeister, 1999: Chabrol et al., 2009), but our studies
offer the first definitive evidence. Typically, others’ suffering is automatically met with
empathic concern and shared distress (Preston & De Waal, 2002), yet sadists display
an opposing process in which others’ pain is transmuted into their pleasure. More work
is needed to understand the precise psychological and biological mechanisms that allow
others’ pain to be experienced as pleasant.
Limitations and Future Directions
Our assessments of sadism were exclusively explicit and obtained via self-report.
Sadism is a socially-undesirable trait. As such, participants may have under-reported
their sadism. Future research may use implicit measures of sadism (Reidy et al., 2011)
to outflank this limitation. However, under-reporting is only problematic if certain types of
SADISM AND AGGRESSION 44
individuals differentially under-report sadism, or if certain methodological techniques or
contexts elicit differential responding. We observed no evidence of differential under-
reporting in our data.
Sadism tends to be greater among males than females (Buckels et al., 2013),
and our studies were not conducted to explicitly take gender effects into account.
Entering gender as a covariate did not fundamentally alter sadism’s link to aggression
across the majority of our studies, suggesting that our effects were not artifacts of
males’ greater sadism. However, future research is needed that treats gender as a
variable of interest and not a nuisance factor.
Another effect of the self-report approach we took to measuring sadism was that
our findings were purely correlational. As such, we cannot be sure of the directionality of
our effects or if other variables artificially created them. Such correlations do not allow
for the establishment of a temporal or causal sequences of variables and violate many
of the assumptions of mediation modeling (see Giner-Sorolla, 2016). Experimental
manipulations that increase sadistic states should be developed to allow for causal
inferences and directional statements about the sadism-aggression link. Further,
longitudinal work that is interrogated with cross-lagged analyses would allow for
directional inferences.
In addition to these assessment issues, our findings were obtained with
undergraduate and Mechanical Turk participants who are unlikely to exhibit the violent
and belligerent behavior of forensically- or clinically-aggressive individuals or the larger
proportion of the global human population for that matter (Henrich, Heine, &
Norenzayan, 2010). Undergraduate and Mechanical Turk populations possess
SADISM AND AGGRESSION 45
characteristics (e.g., wealth, education-level, societal structure) that are not replicated in
most of the world. As such, it is crucial for future research to replicate our effects with
diverse populations that better approximate the true range of aggressive tendencies in
the real world. We also frequently used experimentally-manipulated variables as
covariates in many of our analyses. The appropriateness of including experimentally-
manipulated variables, which should already be equivalent across most demographic
and trait domains, is debatable and readers should use caution when interpreting the
results of such covariate analyses.
Conclusion
What people enjoy varies wildly. Some people enjoy hurting others. We found
that these tendencies are not confined to people’s heads and bleed into their actions as
well. Where people fall along the sadistic spectrum seems to predict how aggressively
they act towards others and the pain they inflict promotes a fleeting sense of pleasure
that is soon replaced by affective discomfort. It is our hope that shining light on such
dark features of personality leads to greater understanding and interventions that
manage to break the link between personal pleasure and others’ suffering.
SADISM AND AGGRESSION 46
References
Anderson, C. A., & Bushman, B. J. (1997). External validity of "trivial" experiments: The
case of laboratory aggression. Review of General Psychology, 1, 19-41.
Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of
Psychology, 53, 2751.
Baughman, H. M., Dearing, S., Giammarco, E., & Vernon, P. A. (2012). Relationships
between bullying behaviours and the Dark Triad: A study with adults. Personality
and Individual Differences, 52, 571-575.
Baumeister R. F. (1997). Evil: inside human violence and cruelty. New York, NY: W.H.
Freeman.
Berkowitz, L. (1989). Affective aggression: The role of stress, pain, and negative affect.
In R. G. Geen & E. Donnerstein (Eds.), Human aggression: theories, research,
and implications for social policy (pp. 4972). San Diego, CA: Academic Press.
Buckels, E. E., Jones, D. N., & Paulhus, D. L. (2013). Behavioral confirmation of
everyday sadism. Psychological Science, 24, 2201-2209.
Buckels, E. E., Trapnell, P. D., Andjelovic, T., & Paulhus, D. L. (in press). Internet
trolling and everyday sadism: Parallel effects on pain perception and moral
judgment. Journal of Personality.
Buckels, E. E., Trapnell, P. D., & Paulhus, D. L. (2014). Trolls just want to have
fun. Personality and individual Differences, 67, 97-102.
Bushman, B. J., & Baumeister, R. F. (1998). Threatened egotism, narcissism, self-
esteem, and direct and displaced aggression: Does self-love or self-hate lead to
violence? Journal of Personality and Social Psychology, 75, 219-229.
SADISM AND AGGRESSION 47
Bushman, B. J., Baumeister, R. F., & Phillips, C. M. (2001). Do people aggress to
improve their mood? Catharsis beliefs, affect regulation opportunity, and
aggressive responding. Journal of Personality and Social Psychology, 81, 1732.
Chabrol, H., Van Leeuwen, N., Rodgers, R., & Séjourné, N. (2009). Contributions of
psychopathic, narcissistic, Machiavellian, and sadistic personality traits to
juvenile delinquency. Personality and Individual Differences, 47, 734739.
Chester, D. S. (2017). The role of positive affect in aggression. Current Directions in
Psychological Science, 26, 366-370.
Chester, D. S. & DeWall, C. N. (2016). The pleasure of revenge: Retaliatory aggression
arises from a neural imbalance towards reward. Social Cognitive and Affective
Neuroscience, 11, 1173-1182.
Chester, D. S. & DeWall, C. N. (2017). Combating the sting of rejection with the
pleasure of revenge: A new look at how emotion shapes aggression. Journal of
Personality and Social Psychology, 112, 413-430.
Chester, D. S. & DeWall, C. N. (2018). Personality correlates of revenge-seeking:
Multidimensional links to physical aggression, impulsivity, and aggressive
pleasure. Aggressive Behavior, 44, 235-245.
Chester, D. S., DeWall, C. N., Derefinko, K. J., Estus, S., Peters, J. R., Lynam, D. R., &
Jiang, Y. (2015). Monoamine oxidase A (MAOA) genotype predicts greater
aggression through impulsive reactivity to negative affect. Behavioural Brain
Research, 283, 97-101.
Chester, D. S., DeWall, C. N., Derefinko, K. J., Estus, S., Lynam, D. R., Peters, J. R., &
Jiang, Y. (2016). Looking for reward in all the wrong places: Dopamine receptor
SADISM AND AGGRESSION 48
gene polymorphisms indirectly affect aggression through sensation-seeking.
Social Neuroscience, 11, 487-494.
Chester, D. S. & Lasko, E. N. (in press). Validating a standardized approach to the
Taylor Aggression Paradigm. Social Psychological and Personality Science.
Chester, D. S., Lynam, D. R., Milich, R., & DeWall, C. N. (2018). Neural mechanisms of
the rejection-aggression link. Social Cognitive and Affective Neuroscience, 13,
501-512.
Curran, P. J., & Hussong, A. M. (2009). Integrative data analysis: the simultaneous
analysis of multiple data sets. Psychological Methods, 14, 81-100.
Curran, S. L., Andrykowski, M. A., & Studts, J. L. (1995). Short form of the Profile of
Mood States (POMS-SF): Psychometric information. Psychological
Assessment, 7, 80-83.
Denson, T. F., DeWall, C. N., & Finkel, E. J. (2012). Self-control and
aggression. Current Directions in Psychological Science, 21, 20-25.
DeWall, C. N., Anderson, C. A., & Bushman, B. J. (2011). The General Aggression
Model: Theoretical extensions to violence. Psychology of Violence, 1, 245-258.
DeWall, C. N., Finkel, E. J., Lambert, N. M., Slotter, E. B., Bodenhausen, G. V., Pond,
R. S., … Fincham, F. D. (2013). The voodoo doll task: Introducing and validating
a novel method for studying aggressive inclinations. Aggressive Behavior, 39,
419439.
DeWall, C. N., Twenge, J. M., Bushman, B., Im, C., & Williams, K. (2010). A little
acceptance goes a long way: applying social impact theory to the rejection-
aggression link. Social Psychological and Personality Science, 1, 168-174.
SADISM AND AGGRESSION 49
Eadeh, F. R., Peak, S. A., & Lambert, A. J. (2017). The bittersweet taste of revenge: On
the negative and positive consequences of retaliation. Journal of Experimental
Social Psychology, 68, 2739.Furnham, A., Richards, S. C., & Paulhus, D. L.
(2013). The Dark Triad of personality: A 10 year review. Social and Personality
Psychology Compass, 7, 199-216.
Giancola, P., & Chermack, S. (1998). Construct validity of laboratory aggression
paradigms: A response to Tedeschi and Quigley (1996). Aggression and Violent
Behavior, 3, 237-253.
Giner-Sorolla, R. (2016). Approaching a fair deal for significance and other
concerns. Journal of Experimental Social Psychology, 65, 1-6.
Gollwitzer, M., & Bushman, B. J. (2012). Do Victims of Injustice Punish to Improve Their
Mood? Social Psychological and Personality Science, 3, 572580.Harmon-
Jones, C., Bastian, B., & Harmon-Jones, E. (2016). The Discrete Emotions
Questionnaire: A New Tool for Measuring State Self-Reported Emotions. PLoS
ONE, 11, no pagination specified.
Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable
mediation, moderation, and conditional process modeling. Retrieved from
http://www.afhayes.com/public/process2012.pdf
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the
world?. Behavioral and Brain Sciences, 33, 61-83.
Horn, J. L. (1965). A rationale and test for the number of factors in factor
analysis. Psychometrika, 30, 179-185.
SADISM AND AGGRESSION 50
Jonason, P. K., & Webster, G. D. (2010). The dirty dozen: A concise measure of the
dark triad. Psychological Assessment, 22, 420.
Jonason, P. K., Zeigler-Hill, V., & Okan, C. (2017). Good v. evil: Predicting sinning with
dark personality traits and moral foundations. Personality and Individual
Differences, 104, 180-185.
Jones, D. N., & Paulhus, D. L. (2014). Introducing the short dark triad (SD3) a brief
measure of dark personality traits. Assessment, 21, 28-41.
Kirkpatrick, L. A., Waugh, C. E., Valencia, A., & Webster, G. D. (2002). The functional
domain specificity of self-esteem and the differential prediction of
aggression. Journal of Personality and Social Psychology, 82, 756-767.
Krueger, R. B. (2010). The DSM diagnostic criteria for sexual sadism. Archives of
Sexual Behavior, 39, 325-345.
Levenson, M . R., Kiehl, K. A., & Fitzpatrick, C. M. (1995). Assessing psychopathic
attributes in a noninstitutionalized population. Journal of Personality and Social
Psychology, 68, 151-158.
Lieberman, J. D., Solomon, S., Greenberg, J., & McGregor, H. A. (1999). A hot new way
to measure aggression: Hot sauce allocation. Aggressive Behavior, 25, 331348.
March, E., Grieve, R., Marrington, J., & Jonason, P. K. (2017). Trolling on Tinder®(and
other dating apps): Examining the role of the Dark Tetrad and
impulsivity. Personality and Individual Differences, 110, 139-143.
Martens, A., Kosloff, S., Greenberg, J., Landau, M. J., & Schmader, T. (2007). Killing
begets killing: Evidence from a bug-killing paradigm that initial killing fuels
subsequent killing. Personality and Social Psychology Bulletin, 33, 1251-1264.
SADISM AND AGGRESSION 51
Miller, J. D., & Lynam, D. R. (2006). Reactive and proactive aggression: Similarities and
differences. Personality and Individual Differences, 41, 1469-1480.
O’Meara, A., Davies, J., & Hammond, S. (2011). The psychometric properties and utility
of the Short Sadistic Impulse Scale (SSIS). Psychological Assessment, 23, 523
531.
Paulhus, D. L. (2014). Toward a taxonomy of dark personalities. Current Directions in
Psychological Science, 23, 421-426.
Paulhus, D. L., Curtis, S. R., & Jones, D. N. (2018). Aggression as a trait: the Dark
Tetrad alternative. Current Opinion in Psychology, 19, 88-92.
Paulhus, D. L., & Jones, D. N. (2015). Measuring dark personalities via questionnaire.
In G. J. Boyle, D. H. Saklofske & G. Matthews (Eds.), Measures of personality
and social psychological constructs (pp. 562-594). San Diego, CA: Academic
Press.
Paulhus, D. L., & Williams, K. M. (2002). The dark triad of personality: Narcissism,
Machiavellianism, and psychopathy. Journal of Research in Personality, 36, 556-
563.
Pedersen, W. C., Gonzales, C., & Miller, N. (2000). The moderating effect of trivial
triggering provocation on displaced aggression. Journal of Personality and Social
Psychology, 78, 913927.Plouffe, R. A., Saklofske, D. H., & Smith, M. M. (2017).
The Assessment of Sadistic Personality: Preliminary psychometric evidence for a
new measure. Personality and Individual Differences, 104, 166-171.
SADISM AND AGGRESSION 52
Porter, S., & Woodworth, M. (2006). Psychopathy and aggression. In CJ Patrick
(Ed.), Handbook of psychopathy (pp. 481-494). New York, NY: Guilford Press;
2006.
Preston, S. D., & De Waal, F. B. (2002). Empathy: Its ultimate and proximate
bases. Behavioral and brain sciences, 25, 1-20.
Raine, A., Dodge, K., Loeber, R., GatzkeKopp, L., Lynam, D., Reynolds, C., ... & Liu, J.
(2006). The reactiveproactive aggression questionnaire: Differential correlates
of reactive and proactive aggression in adolescent boys. Aggressive
Behavior, 32, 159-171.
Reidy, D. E., Zeichner, A., & Martinez, M. A. (2008). Effects of psychopathy traits on
unprovoked aggression. Aggressive Behavior, 34, 319-328.
Reidy, D. E., Zeichner, A., Miller, J. D., & Martinez, M. A. (2007). Psychopathy and
aggression: Examining the role of psychopathy factors in predicting laboratory
aggression under hostile and instrumental conditions. Journal of Research in
Personality, 41, 1244-1251.
Reidy, D. E., Zeichner, A., & Seibert, L. A. (2011). Unprovoked aggression: Effects of
psychopathic traits and sadism. Journal of Personality, 79, 75100.
Richard, F. D., Bond Jr, C. F., & Stokes-Zoota, J. J. (2003). One hundred years of social
psychology quantitatively described. Review of General Psychology, 7, 331-363.
Russell, T. D., & King, A. R. (2016). Anxious, hostile, and sadistic: Maternal attachment
and everyday sadism predict hostile masculine beliefs and male sexual
violence. Personality and Individual Differences, 99, 340-345.
SADISM AND AGGRESSION 53
Tangney, J. P., Baumeister, R. F., & Boone, A. L. (2004). High Self-Control Predicts
Good Adjustment, Less Pathology, Better Grades, and Interpersonal Success.
Journal of Personality, 72, 271324.
Taylor, S. (1967). Aggressive behavior and physiological arousal as a function of
provocation and the tendency to inhibit aggression. Journal of Personality, 35,
297-310.Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and
validation of brief measures of positive and negative affect: The PANAS scales.
Journal o Personality and Social Psychology, 54, 10631070.
Webster, G. D., DeWall, C. N., Pond, R. S., Deckman, T., Jonason, P. K., Le, B. M., ...
& Smith, C. V. (2014). The brief aggression questionnaire: Psychometric and
behavioral evidence for an efficient measure of trait aggression. Aggressive
Behavior, 40, 120-139.
Webster, G. D., Gesselman, A. N., Crysel, L. C., Brunell, A. B., Jonason, P. K., Hadden,
B. W., & Smith, C. V. (2016). An actorpartner interdependence model of the
Dark Triad and aggression in couples: Relationship duration moderates the link
between psychopathy and argumentativeness. Personality and Individual
Differences, 101, 196-207.
Webster, G. D., & Kirkpatrick, L. A. (2006). Behavioral and selfreported aggression as
a function of domainspecific selfesteem. Aggressive Behavior, 32, 17-
27.Whiteside, S. P., & Lynam, D. R. (2001). The Five Factor Model and
impulsivity: using a structural model of personality to understand impulsivity.
Personality and Individual Differences, 30, 669689.
SADISM AND AGGRESSION 54
Whiteside, S. P., Lynam, D. R., Miller, J. D., & Reynolds, S. K. (2005). Validation of the
UPPS impulsive behaviour scale: A four-factor model of impulsivity. European
Journal of Personality, 19, 559574.
Williams, K. D. (2009). Ostracism: A temporal needthreat model. In Mark P. Zanna
(Ed.), Advances in experimental social psychology (pp. 275314). Cambridge,
MA: Academic Press.
Williams, K. D., Cheung, C. T, K., & Choi, W. (2000). Cyberostracism: Effects of being
ignored over the Internet. Journal of Personality and Social Psychology, 79, 748
762.
Williams, K.S., Yeager, D.S., Cheung, C.K.T., & Choi, W. (2012). Cyberball (version
4.0) [Software]. Available from https://cyberball.wikispaces.com.
SADISM AND AGGRESSION 55
Supplemental Document 1
Integrative Data Analysis
We performed integrative data analyses by combining the datasets from all eight
studies, standardizing sadism and aggression scores within-study. Because Study 4
employed five aggression measures, we standardized each of them, averaged them
together, and then standardized this aggression index to enter into the integrative data
analysis. The analysis took the form of a multilevel linear model in SAS v.9.4, using
maximum likelihood estimation, specifying the study of origin and intercepts as random,
and modeling the following sources of variance: participant (level 1), study (level 2;).
Sadism exhibited significant linear, β=0.22, SE=0.03, 95% CI=0.17, 0.27, t(1,515)=8.51,
p<.001, and curvilinear, β=-0.19, SE=0.07, 95% CI=-0.05, -0.33, t(918)=-2.68, p=.008
(Supplemental Figure 1), associations with aggression across all eight studies.
SADISM AND AGGRESSION 56
Supplemental Figure 1. Curvilinear association between sadism and aggressive
behavior across all eight studies. Bands around the regression slope depict 95%
confidence intervals.
Running head: SADISM AND AGGRESSION 57
Supplemental Table 1. Descriptive statistics and internal scale consistencies for key variables from Study 1. NTS
= Need Threat Scale, LSRP = Levenson Self-Report Psychopathy Scale, SSIS = Short Sadistic Impulse Scale,
UPPSP = UPPS-P Impulsivity Scale. ‘Pre’ and ‘Post’ refer to before and after participants completed the
aggression measure, respectively.
Measure
M
SD
Observed
Minimum
Observed
Maximum
α
Missing Data
N
Brief Self-Control Scale
3.39
0.67
1.69
5.00
.84
9
NTS - Negative Affect (Pre)
1.38
0.57
1.00
3.33
.73
4
NTS - Positive Affect (Pre)
3.76
0.70
2.00
5.00
.81
4
NTS - Negative Affect (Post)
1.68
0.77
1.00
4.00
.68
28
NTS - Positive Affect (Post)
3.02
0.96
1.00
5.00
.85
28
LSRP - Primary
28.70
8.74
4.00
56.00
.81
17
LSRP - Secondary
18.00
5.11
2.00
35.00
.74
16
SSIS (10 items)
1.48
0.32
1.00
3.30
.55
12
SSIS (9 items)
1.17
0.29
1.00
3.00
.73
12
Taylor Aggression Paradigm
3.80
2.25
0.00
10.00
.98
7
SADISM AND AGGRESSION 58
UPPSP - Lack of Perseverance
1.89
0.47
1.00
3.60
.81
4
UPPSP - Lack of Premeditation
2.01
0.48
1.00
3.36
.83
4
UPPSP - Negative Urgency
2.05
0.60
1.08
3.92
.87
4
UPPSP - Positive Urgency
1.70
0.61
1.00
3.64
.93
4
UPPSP - Sensation-Seeking
2.85
0.62
1.33
4.00
.85
4
SADISM AND AGGRESSION 59
Supplemental Table 2. Zero-order correlations between key variables from Study 1. BSCS = Brief Self-Control Scale, NTS = Need
Threat Scale, LSRP = Levenson Self-Report Psychopathy Scale, SSIS = Short Sadistic Impulse Scale (9 items), TAP = Taylor
Aggression Paradigm, UPPSP = UPPS-P Impulsivity Scale. ‘Pre’ and ‘Post’ refer to before and after participants completed the
aggression measure, respectively. Gender is coded: male = 1, female = -1. *p < .05, **p < .01
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1. BSCS
2. Gender
-.17*
3. NTS - Negative
Affect (Pre)
-.13
-.00
4. NTS - Positive
Affect (Pre)
.14
-.04
-.48**
5. NTS - Negative
Affect (Post)
-.22*
.05
.28**
-.05
SADISM AND AGGRESSION 60
6. NTS - Positive
Affect (Post)
-.00
.02
-.13
.33**
-.44**
7. LSRP - Primary
-.36**
.17*
.10
-.08
.19*
-.02
8. LSRP -
Secondary
-.63**
.07
.23**
-.24**
.20*
-.05
.57**
9. SSIS
-.22**
.45**
.25**
-.25**
.21*
-.22*
.29**
.28**
10. TAP
-.05
.21**
.07
-.08
.17*
-.08
.27**
.19*
.25**
11. UPPSP - Lack
of Perseverance
-.56**
.13
-.02
-.14
.10
-.18*
.09
.34**
.22**
-.08
12. UPPSP - Lack
of Premeditation
-.47**
-.09
.05
-.07
.05
-.18*
.14
.32**
.04
.01
.31**
13. UPPSP -
Negative Urgency
-.76**
.07
.22**
-.20*
.32**
-.18*
.30**
.57**
.18*
.10
.43**
.36**
SADISM AND AGGRESSION 61
14. UPPSP -
Positive Urgency
-.66**
.08
.21*
-.12
.25**
-.09
.33**
.59**
.16
.25
.35**
.36**
.78**
15. UPPSP -
Sensation-
Seeking
-.31**
.09
.07
.12
.11
.08
.25**
.34**
.09
.04
.08
.27**
.20*
.29**
SADISM AND AGGRESSION 62
Supplemental Table 3. Descriptive statistics and internal scale consistencies for key variables from Study 2.
HSAT = Hot Sauce Aggression Task, NTS = Need Threat Scale, LSRP = Levenson Self-Report Psychopathy Scale,
SSIS = Short Sadistic Impulse Scale. ‘Pre’ and ‘Post’ refer to before and after participants completed the
aggression measure, respectively.
Measure
M
SD
Observed
Minimum
Observed
Maximum
α
Missing Data
N
HSAT (Log Transformed)
0.97
0.11
0.85
1.49
n/a
0
NTS - Negative Affect (Pre)
2.17
1.36
1.00
7.00
.84
0
NTS - Positive Affect (Pre)
4.45
1.51
1.00
7.00
.90
0
NTS - Negative Affect (Post)
1.71
1.00
1.00
7.00
.81
1
NTS - Positive Affect (Post)
5.20
1.36
1.00
7.00
.91
1
LSRP - Primary
29.71
6.81
16.00
53.00
.84
0
LSRP - Secondary
19.68
3.66
12.00
29.00
.59
0
SSIS (10 items)
1.42
0.52
1.00
4.10
.65
0
SSIS (9 items)
1.29
0.50
1.00
4.33
.79
0
SADISM AND AGGRESSION 63
Supplemental Table 4. Zero-order correlations between key variables from Study 2. HSAT = Hot Sauce Aggression
Task, NTS = Need Threat Scale, LSRP = Levenson Self-Report Psychopathy Scale. ‘Pre’ and ‘Post’ refer to before and
after participants completed the aggression measure, respectively. Gender is coded: male = 1, female = -1. *p < .05, **p
< .01, ***p < .001
1
2
3
4
5
6
7
8
1. Gender
2. HSAT (Log Transformed)
.21*
3. NTS - Negative Affect (Pre)
-.11
-.01
4. NTS - Positive Affect (Pre)
.06
.05
-.61***
5. NTS - Negative Affect (Post)
-.14
.02
.38***
-.34***
6. NTS - Positive Affect (Post)
.17*
.04
-.14
.51***
-.67***
7. LSRP - Primary
.25**
.18*
-.09
-.06
-.04
.00
8. LSRP - Secondary
.02
.17*
.10
-.15
.15
-.14
.39***
9. Short Sadistic Impulse Scale
.20*
.16
.08
.02
.21**
-.05
.40***
.35***
SADISM AND AGGRESSION 64
Supplemental Table 5. Descriptive statistics and internal scale consistencies for key variables from Study 3. NTS
= Need Threat Scale. ‘Pre’ and ‘Post’ refer to before and after participants completed the aggression measure,
respectively.
Measure
M
SD
Observed
Minimum
Observed
Maximum
α
Missing Data N
Gruesome Images
3.21
2.25
0.00
9.00
n/a
0
NTS - Negative Affect (Pre)
2.19
1.02
1.00
5.00
.86
1
NTS - Positive Affect (Pre)
2.93
0.93
1.00
5.00
.87
1
NTS - Negative Affect (Post)
3.27
0.91
1.33
5.00
.86
5
NTS - Positive Affect (Post)
2.01
0.93
1.00
4.67
.89
5
Short Sadistic Impulse Scale
1.86
0.89
1.00
5.60
.85
1
SADISM AND AGGRESSION 65
Supplemental Table 6. Zero-order correlations between key variables from Study 3. NTS =
Need Threat Scale. ‘Pre’ and ‘Post’ refer to before and after participants completed the
aggression measure, respectively. Gender is coded: male = 1, female = -1. *p < .05, **p <
.01, ***p < .001
1
2
3
4
5
6
1. Gender
2. Gruesome Images
.06
3. NTS - Negative Affect (Pre)
.09
-.00
4. NTS - Positive Affect (Pre)
.09
.08
-.26**
5. NTS - Negative Affect (Post)
.06
.08
.53***
-.08
6. NTS - Positive Affect (Post)
.03
-.03
-.15
.46***
-.38***
7. Short Sadistic Impulse Scale
.17
.18*
.08
-.14
.24**
-.16
SADISM AND AGGRESSION 66
Supplemental Table 7. Item text for the Positive Affect During Aggression Scale, taken from the Need Threat
Affect Subscales (NTAS), Positive Affect Negative Affect Schedule (PANAS), Profile of Mood States (POMS),
Discrete Emotions Questionnaire (DEQ), and generated by the authors (ADHOC). *Retained in final 9-item
Aggressive Pleasure subscale.
Original Scale
Indicate the extent to which this statement described how you felt
when your opponent received the noise blasts that you picked in
the competitive reaction-time task...
PANAS
Active
PANAS
Alert
DEQ
Anticipation
ADHOC
At peace
PANAS
Attentive
ADHOC
Blessed
DEQ
Calm
ADHOC
Carefree
ADHOC
Delighted*
SADISM AND AGGRESSION 67
PANAS
Determined
ADHOC
Ecstatic
POMS
Energetic
PANAS
Enthusiastic
PANAS
Excited
NTAS
Friendly
ADHOC
Glad*
NTAS
Good*
ADHOC
Gratified*
NTAS/DEQ
Happy*
ADHOC
Hopeful
ADHOC
In Love
PANAS
Inspired
PANAS
Interested
ADHOC
Joy
ADHOC
Laid-back
SADISM AND AGGRESSION 68
POMS
Lively
NTAS
Pleasant*
ADHOC
Pleased*
PANAS
Proud*
ADHOC
Relaxed
ADHOC
Rewarded
ADHOC
Satisfied*
ADHOC
Serene
PANAS
Strong
DEQ
Thankful
ADHOC
Tranquil
SADISM AND AGGRESSION 69
Supplemental Table 8. Original items of the Positive Affect During Aggression Scale and their
associated component loadings from the EFA from Study 7. *Retained in final 9-item Aggressive
Pleasure subscale.
Original
Scale Items
Component
1
Component
2
Component
3
Component
4
Component
5
Active
.00
.05
.04
.09
.35
Alert
.10
-.06
-.05
.04
.64
Anticipation
-.02
-.13
.22
.06
.26
At peace
.36
.47
-.02
.09
.15
Attentive
-.03
.19
.08
-.07
.66
Blessed
.02
.05
.37
.43
.12
Calm
.07
.77
-.09
.03
.07
Carefree
-.04
.73
-.03
-.01
-.01
Delighted*
.62
.05
.21
-.03
-.01
Determined
.01
.05
.18
.12
.40
Ecstatic
.18
-.10
.63
.13
-.12
SADISM AND AGGRESSION 70
Energetic
-.10
.05
.60
.17
.17
Enthusiastic
.38
.03
.26
.01
.02
Excited
.38
.08
.22
-.09
.00
Friendly
.30
.31
-.09
.27
.13
Glad*
.75
-.08
.03
.04
.05
Good*
.67
.26
-.04
.08
.07
Gratified*
.54
-.05
.07
.07
.08
Happy*
.76
.19
-.04
.00
.07
Hopeful
.14
.00
.11
.46
.14
In Love
-.04
.06
.25
.29
.03
Inspired
.31
.04
.21
.00
.18
Interested
.04
.20
-.09
.02
.04
Joy
.30
-.01
.21
.48
-.15
Laid-back
.01
.78
.07
-.04
-.07
Lively
.01
.07
.59
.11
.09
Pleasant*
.63
.18
.05
.13
.00
SADISM AND AGGRESSION 71
Pleased*
.70
.03
.07
.06
.05
Proud*
.54
.01
-.04
.22
.09
Relaxed
.05
.76
-.02
.12
.05
Rewarded
.24
.12
.17
.12
-.06
Satisfied*
.66
.10
-.01
.02
.10
Serene
.29
.31
.27
.04
.27
Strong
.20
.01
.01
.15
.44
Thankful
-.06
.02
-.02
.92
-.01
Tranquil
.18
.47
.31
.00
.13
SADISM AND AGGRESSION 72
Supplemental Table 9. Descriptive statistics and internal scale consistencies for key variables from Study 4.
HPFS = History of Physical Fights, HSAT = Hot Sauce Aggression Task, PADAS = Positive Affect During
Aggression Scale, SSIS = Short Sadistic Impulse Scale, VDAT = Voodoo Doll Aggression Task.
Measure
M
SD
Observed
Minimum
Observed
Maximum
α
Missing Data N
HPFS - Past Five Years
0.79
1.50
0.00
10.00
n/a
11
HPFS - Past Year
0.21
0.66
0.00
5.00
n/a
6
HSAT (Log Transformed)
0.56
0.33
0.00
1.43
n/a
8
PADAS - Pleasure
2.62
1.67
1.00
7.00
.96
13
SSIS (10 items)
1.48
0.53
1.00
4.10
.62
29
SSIS (9 items)
1.29
0.50
1.00
4.11
.77
29
Taylor Aggression Paradigm
5.10
2.01
0.00
10.00
.98
0
VDAT (Log Transformed)
0.54
0.55
0.00
1.72
n/a
2
SADISM AND AGGRESSION 73
Supplemental Table 10. Zero-order correlations between key variables from Study 4. HPFS = History of
Physical Fights, HSAT = Hot Sauce Aggression Task, PADAS = Positive Affect During Aggression Scale,
VDAT = Voodoo Doll Aggression Task. Gender is coded: male = 1, female = -1. *p < .05, **p < .01, ***p < .001
1
2
3
4
5
6
7
1. Gender
2. HPFS - Past Five Years
.38***
3. HPFS - Past Year
.29***
.78***
4. HSAT (Log Transformed)
.24**
.11
.02
5. PADAS - Pleasure
.21**
.12
.17*
-.01
6. Short Sadistic Impulse Scale
.23**
.40***
.45***
.06
.19*
7. Taylor Aggression Paradigm
-.11
-.04
.20
.09
-.16**
.19*
8. VDAT (Log Transformed)
-.16*
-.06
.01
-.00
.17*
.14
.38***
SADISM AND AGGRESSION 74
Supplemental Table 11. Descriptive statistics and internal scale consistencies for key variables from Study 5.
BAQ = Brief Aggression Questionnaire, NTS = Need Threat Scale, VDAT = Voodoo Doll Aggression Task. ‘Pre’,
‘During’, and ‘Post’ refer to before, during, and after participants completed the aggression measure,
respectively.
Measure
M
SD
Observed
Minimum
Observed
Maximum
α
Missing Data N
BAQ - Anger
3.05
1.45
1.00
6.33
.73
1
BAQ - Hostility
3.63
1.50
1.00
7.00
.76
1
BAQ - Physical Aggression
3.48
1.82
1.00
7.00
.83
1
BAQ - Verbal Aggression
4.41
1.32
1.00
7.00
.68
1
NTS - Negative Affect (Pre)
2.88
1.67
1.00
6.33
.87
1
NTS - Positive Affect (Pre)
4.70
1.71
1.00
7.00
.92
1
NTS - Negative Affect (During)
3.65
1.67
1.00
7.00
.81
1
NTS - Positive Affect (During)
3.63
1.98
1.00
7.00
.95
1
NTS - Negative Affect (Post)
2.59
1.59
1.00
7.00
.86
1
NTS - Positive Affect (Post)
4.94
1.61
1.00
7.00
.93
1
Short Sadistic Impulse Scale
2.00
1.18
1.00
5.80
.90
1
SADISM AND AGGRESSION 75
VDAT (Log Transformed)
0.54
0.65
0.00
1.72
n/a
1
SADISM AND AGGRESSION 76
Supplemental Table 12. Zero-order correlations between key variables from Study 5. BAQ = Brief Aggression
Questionnaire, NTS = Need Threat Scale, SSIS = Short Sadistic Impulse Scale, VDAT = Voodoo Doll Aggression Task.
Gender is coded: male = 1, female = -1. ‘Pre’, ‘During’, and ‘Post’ refer to before, during, and after participants
completed the aggression measure, respectively. *p < .05, **p < .01, ***p < .001
1
2
5
6
10
11
1. BAQ - Anger
2. BAQ - Hostility
.54***
3. BAQ - Physical Aggression
.44***
.44***
4. BAQ - Verbal Aggression
.39***
.41***
5. Gender
.19*
.08
6. NTS - Negative Affect (Pre)
.31***
.29***
-.01
7. NTS - Positive Affect (Pre)
-.10
-.16
.11
-.61***
8. NTS - Negative Affect
(During)
.12
.05
.05
.19*
9. NTS - Positive Affect
(During)
.15
.20*
.12
.23**
10. NTS - Negative Affect
(Post)
.42***
.33***
.07
.66***
SADISM AND AGGRESSION 77
11. NTS - Positive Affect (Post)
-.11
-.07
.12
-.35***
-.33***
12. SSIS
.55***
.52***
.10
.35***
.41***
-.02
13. VDAT (Log Transformed)
.27**
.38***
.02
.43***
.41***
-.00
SADISM AND AGGRESSION 78
Supplemental Table 13. Descriptive statistics and internal scale consistencies for key variables from Study 6.
PANAS = Positive Affect Negative Affect Schedule, SSIS = Short Sadistic Impulse Scale, VDAT = Voodoo Doll
Aggression Task. ‘Pre’, ‘During’, and ‘Post’ refer to before, during, and after participants completed the aggression
measure, respectively.
Measure
Observed Minimum
Observed Maximum
α
Missing Data N
PANAS - Negative Affect (Pre)
1.00
5.00
.83
1
PANAS - Positive Affect (Pre)
1.00
7.00
.92
1
PANAS - Negative Affect (During)
1.00
4.30
.87
0
PANAS - Positive Affect (During)
1.00
5.00
.93
0
PANAS - Negative Affect (Post)
1.00
5.00
.83
0
PANAS - Positive Affect (Post)
1.00
7.00
.93
0
SSIS (Ten Items)
1.00
3.40
.64
0
SSIS (Nine Items)
1.00
3.67
.72
0
VDAT (Log Transformed)
0.00
1.72
n/a
1
SADISM AND AGGRESSION 79
Supplemental Table 14. Zero-order correlations between key variables from Study 6. PANAS = Positive Affect Negative Affect
Schedule, VDAT = Voodoo Doll Aggression Task. Gender is coded: male = 1, female = -1. ‘Pre’, ‘During’, and ‘Post’ refer to
before, during, and after participants completed the aggression measure, respectively. *p < .05, **p < .01, ***p < .001
1
2
3
4
5
6
7
8
1. Gender
2. PANAS - Negative Affect (Pre)
.01
3. PANAS - Positive Affect (Pre)
.25***
-.10
4. PANAS - Negative Affect (During)
-.06
.35***
.10
5. PANAS - Positive Affect (During)
.14*
.14*
.35***
.09
6. PANAS - Negative Affect (Post)
.06
80***
-.07
.48***
.13*
7. PANAS - Positive Affect (Post)
.21**
-.13*
.82***
.08
.38***
-.10
8. Short Sadistic Impulse Scale
.31***
.27***
.03
.09
.18**
.31***
.01
9. VDAT (Log Transformed)
.04
.12
.02
.19**
.25***
.10
-.02
.21**
SADISM AND AGGRESSION 80
Supplemental Table 15. Original items from Victim Suffering Scale and their associated component loadings from
the PCA in Study 7. *Used in Study 8’s manipulation check.
The pins that I placed in the doll that represented
my partner...
Factor 1
Factor 2
Included in final
VSS?
really hurt them.*
.82
-.05
Yes
caused them to feel real pain. [Study 8: caused
them real harm]*
.89
-.04
Yes
failed to harm them.
.05
.75
No
were painful to them.*
.83
-.07
Yes
didn’t hurt them.
-.10
.86
No
inflicted actual damage to them.
.88
.07
Yes
made them feel pain.
.89
-.05
Yes
pained them.
.91
-.05
Yes
caused them to suffer.
.92
.02
Yes
inflicted a real wound in them.
.90
.11
Yes
SADISM AND AGGRESSION 81
Supplemental Table 16. Descriptive statistics and internal scale consistencies for key variables from Study 7.
PADAS = Positive Affect During Aggression Scale, VDAT = Voodoo Doll Aggression Task.
Measure
Observed Minimum
Observed Maximum
α
Missing Data N
Sadism Index
-0.23
3.15
.95
0
PADAS - Arousal
1.00
7.00
.80
0
PADAS - Calmness
1.00
7.00
.87
0
PADAS - Pleasure
1.00
7.00
.95
0
VDAT (Log Transformed)
0.00
1.72
n/a
0
Victim Suffering Scale
1.00
7.00
.97
0
SADISM AND AGGRESSION 82
Supplemental Table 17. Zero-order correlations between key variables from Study 7. DD = Dirty Dozen Questionnaire, PADAS
= Positive Affect During Aggression Scale, VDAT = Voodoo Doll Aggression Task. Gender is coded: male = 1, female = -1. *p <
.05, **p < .01, ***p < .001
Measure
1
2
3
4
5
6
7
8
9
1. Gender
2. Sadism Index
.18**
3. DD - Machiavellianism
.08
.51***
4. DD - Narcissism
.12*
.58***
.61***
5. DD - Psychopathy
.11*
.34***
.56***
.40***
6. PADAS - Arousal
.17**
.22***
.06
.07
.11*
7. PADAS - Calmness
.16**
.15**
.08
.06
.07
.46***
8. PADAS - Pleasure
.18**
.19***
.07
.04
.11*
.64***
.63***
9. VDAT (Log Transformed)
.12*
.14**
.03
.03
-.06
.06
.04
.03
10. Victim Suffering Scale
.11*
.26***
.08
.11*
.04
.23***
.02
.08
.34***
SADISM AND AGGRESSION 83
Supplemental Table 18. Descriptive statistics and internal scale consistencies for key variables from Study 8.
PADAS = Positive Affect During Aggression Scale.
Measure
Observed Minimum
Observed Maximum
α
Missing Data N
Sadism Index
-0.44
2.77
.85
0
PADAS - Pleasure
1.00
7.00
.96
2
Taylor Aggression Paradigm
0.00
9.74
.98
3
Victim Suffering Scale
1.00
7.00
.88
2
SADISM AND AGGRESSION 84
Supplemental Table 19. Zero-order correlations between key variables from Study 8. PADAS = Positive Affect
During Aggression Scale, SD3 = Short Dark Triad Scale, TAP = Taylor Aggression Paradigm, VSS = Victim
Suffering Scale (short version). Gender is coded: male = 1, female = -1. *p < .05, **p < .01, ***p < .001
Measure
1
2
3
4
5
6
7
1. Gender
2. Sadism Index
.27***
3. PADAS - Pleasure
.08
-.01
4. SD3 - Machiavellianism
.13
.38***
-.03
5. SD3 - Narcissism
.10
.09
.14*
.24**
6. SD3 - Psychopathy
.18**
.59***
-.03
.52***
.22**
7. Taylor Aggression Paradigm
.07
.04
.04
.14*
.26***
.14*
8. Victim Suffering Scale
-.10
.08
-.15*
.20**
-.01
.16*
.37***
SADISM AND AGGRESSION 85
Supplemental Table 20. Effect sizes and sample sizes entered into internal meta-analysis.
Study
Effect Size (r)
Sample Size (n)
Aggression Measure
1
.25
147
Noise Blasts
2
.16
149
Hot Sauce
3
.18
125
Gruesome Images
4
.19
182
Noise Blasts
5
.46
152
Voodoo Doll Pins
6
.21
237
Voodoo Doll Pins
7
.14
356
Voodoo Doll Pins
8
.04
204
Noise Blasts
Running head: SADISM AND AGGRESSION 86
Methodology Attachment
Sadism and Aggressive Behavior: Inflicting Pain to Feel Pleasure
-----Study 1-----
Cover Story:
Caffeine and placebo conditions: “The purpose of this study is to investigate how
caffeine affects cognition. To do so, you will consume two pills containing either 200 mg
of caffeine or a corn starch placebo. You will not be told which one you have consumed
until the end of the study. After you take the pills, you will fill out a series of
questionnaires, watch a short video, perform a word identification task, perform a shape
identification task, complete a language task, and play a reaction time game.”
Control condition: The purpose of this study is to investigate how caffeine affects
cognition. You are a member of our control group and will not receive caffeine. In this
study, you will fill out a series of questionnaires, watch a short video, perform a word
identification task, perform a shape identification task, complete a language task, and
play a reaction time game.”
Independent Variables:
-Experimental Provocation Manipulation experimenter script:
1. “Now we will move on to the anagram task. Anagrams are words with
scrambled letters. Your goal in this task is to unscramble the letters so that
they form an actual word in the English language, as fast as you can.
2. Hand them an anagrams form.
3. Say “Here is the sheet with the anagrams you will complete. There are 14 to
complete and they are all solvable. I will come back in several minutes to
collect your responses.”
4. Leave the room and time 3 minutes.
5. Come back into the room and WITHOUT LOOKING AT THE SCRIPT, say “You
are not done yet? ………….Okay, I will give you some more time.”
6. Leave the room and time 1 minute and 30 seconds.
SADISM AND AGGRESSION 87
7. Re-enter the room and WITHOUT LOOKING AT THE SCRIPT say, “Are you
still not done? ………….Well, because you seem to be struggling, I will give
you some more time.”
8. Leave the room, and time 1 minute.
9. Reenter the room and WITHOUT LOOKING AT THE SCRIPT, sigh, and say
Alright, this seems to be a waste of my time. Let’s just move on to the next
task.”
Anagram Task
Below are 14 10-letter anagrams. Each anagram, when solved, becomes a word. In
the next 5 minutes, please solve as many anagrams as possible.
1. a b l s o u t e l y: __ __ __ __ __ __ __ __ __ __
2. s z i a n t i i r g: __ __ __ __ __ __ __ __ __ __
3. h c i m o a n a r s: __ __ __ __ __ __ __ __ __ __
4. w n a g z i h b z s: __ __ __ __ __ __ __ __ __ __
5. u n a b a l c n e d: __ __ __ __ __ __ __ __ __ __
6. q i z n i u t a n g: __ __ __ __ __ __ __ __ __ __
7. p c z e u i l b i d: __ __ __ __ __ __ __ __ __ __
8. c z l i n i v i i g: __ __ __ __ __ __ __ __ __ __
SADISM AND AGGRESSION 88
9. m t c y h i z i e d: __ __ __ __ __ __ __ __ __ __
10. c n e t i m t e e r: __ __ __ __ __ __ __ __ __ __
11. v o i n c l a z i g: __ __ __ __ __ __ __ __ __ __
12. b m m k o g a c a n: __ __ __ __ __ __ __ __ __ __
13. l u m e r b a j c k: __ __ __ __ __ __ __ __ __ __
14. a x i m a o s t i e: __ __ __ __ __ __ __ __ __ __
-Short Sadistic Impulse Scale (available here:
http://psycnet.apa.org/doi/10.1037/a0022400)
Dependent Variables:
-Taylor Aggression Paradigm (stimuli available here: https://osf.io/a2wft/files/)
-Need Threat Scale (available here: https://doi.org/10.1016/S0065-2601(08)00406-1)
-----Study 2-----
Cover Story: The effect mental visualization abilities on taste preferences. The
Cyberball task was explained as an opportunity to practice mental visualization skills
and the hot sauce aggression task was explained as a means to assess taste
preferences.
Independent Variables:
-Rejection Manipulation:
SADISM AND AGGRESSION 89
Acceptance condition website:
http://laits.utexas.edu/cyberball/cyberball.htm?userid=agileProgger&amp;settings=3plin
&amp;p2name=Me&amp;pic4=images/stevejobs.png&amp;pics=false&amp;pic1=image
s/stevejobs.png&amp;pic3=images/mark.jpg&amp;chat=false&amp;p1name=Sam&amp
;p3name=Jordan" target="_blank"><em><em><span style="font-
family:&quot;Arial&quot;,sans-serif;mso-fareast-font-
family:&quot;TimesNewRoman&quot;;color:navy
Rejection condition website:
http://laits.utexas.edu/cyberball/cyberball.htm?userid=agileProgger&amp;settings=3plos
&amp;p2name=Me&amp;pic4=images/stevejobs.png&amp;pics=false&amp;pic1=image
s/stevejobs.png&amp;pic3=images/mark.jpg&amp;chat=false&amp;p1name=Sam&amp
;p3name=Jordan"target="_blank"><em><em><spanstyle="fontfamily:&quot;Arial&quot;,
sans-serif;mso-fareast-font-family:&quot;Times New Roman&quot;;color:navy
-Short Sadistic Impulse Scale (see Study 1)
SADISM AND AGGRESSION 90
Dependent Variables:
-Hot Sauce Aggression Task (instructions and stimuli of hot sauce bottles used
available here: https://osf.io/a2wft/files/)
-Need Threat Scale (see Study 1)
-----Study 3-----
Cover Story: “This study is an online survey about individuals’ ability to mentally
visualize events.The Cyberball task was explained as an opportunity to practice
mental visualization skills and the cover story for the aggression measure is depicted in
the ‘Dependent Variables’ section below.
Independent Variables:
-Rejection manipulation (see Study 2)
-Short Sadistic Impulse Scale (see Study 1)
Dependent Variables:
-Gruesome image aggression task:
-Need Threat Scale (see Study 1)
-----Study 4-----
Cover Story: This is detailed in the experimenter script available here:
https://osf.io/a2wft/files/
Independent Variables:
SADISM AND AGGRESSION 91
-Experimental provocation materials and scripts are available here:
https://osf.io/a2wft/files/
-Short Sadistic Impulse Scale (see Study 1)
-Positive Affect During Aggression Scale (see Supplemental Table 1)
Dependent Variables:
-Stimuli and experimenter scripts for all dependent variables are available here:
https://osf.io/a2wft/files/
-----Study 5-----
Cover Story: “This online study is about individuals’ ability to mentally visualize events
and write about them.” The essay manipulation cover story is detailed below in the
‘Independent Variables’ section. The voodoo doll aggression task was introduced as an
opportunity to “interact with a virtual avatar to release pent up energy, this will take
approximately 5 minutes.
Independent Variables:
-Short Sadistic Impulse Scale (see Study 1)
-Essay manipulation:
SADISM AND AGGRESSION 92
SADISM AND AGGRESSION 93
SADISM AND AGGRESSION 94
SADISM AND AGGRESSION 95
Dependent Variables:
-Need Threat Scale (see Study 1)
-Voodoo Doll Aggression Task: (sample stimuli available here: https://osf.io/a2wft/files/)
-----Study 6-----
Cover Story: “The purpose of this study is to investigate your cognitive and mental
abilities. To do so, we will repeatedly measure your cognitive abilities which often
change minute-by-minute.”
The self-regulation fatigue manipulation cover story is detailed in the ‘Independent
Variables’ section below and no cover story was given for the voodoo doll aggression
measure.
Independent Variables:
-Short Sadistic Impulse Scale (see Study 1)
SADISM AND AGGRESSION 96
-Self-control fatigue manipulation:
Experimenter script:
1. Say: “Now you will complete a short essay that will help us understand
your cognitive abilities. I will give you 5 minutes to simply describe the
last 3 years of your life in whatever manner you wish to do so. However,
there is a catch, you are forbidden from using two letters as you write the
essay. To see which letters you cannot use, please read the instructions
at the top of the essay form that is in front of you.”
2. Start the stopwatch and say “Please begin writing, I will ask you to stop in 5
minutes.”
3. Close the door and wait out in the hallway.
4. When the stopwatch reaches 5 minutes, instruct participants to stop.
5. Collect their essay,
Control condition:
Please write a description of the last 3 years of your life, recalling key
events. You will have 5 minutes to do so. While writing, you CANNOT use
the letters X or Z at any time.
Fatigue Condition:
Please write a description of the last 3 years of your life, recalling key
events. You will have 5 minutes to do so. While writing, you CANNOT use
the letters A or N at any time.
Dependent Variables:
-Positive Affect Negative Affect Schedule (available here:
https://www.ncbi.nlm.nih.gov/pubmed/3397865)
-Voodoo Doll Aggression Task (see Study 5)
-----Study 7-----
SADISM AND AGGRESSION 97
Cover Story: “This study explores how your ability to imagine experiences and settings
influences your responses on several computer tasks and questionnaires.” The
Cyberball task was explained as an opportunity to practice mental visualization skills
and the voodoo doll task was explained as “a short imagination task”.
Independent Variables:
-Rejection paradigm (see Study 2)
-Short Sadistic Impulse Scale (see Study 1)
-Assessment of Sadistic Personality (available here:
https://doi.org/10.1016/j.paid.2016.07.043)
-Comprehensive Assessment of Sadistic Tendencies (available here:
http://www2.psych.ubc.ca/~dpaulhus/research/DARK_TRIAD/MEASURES/CAST.3.0.do
cx)
Dependent Variables:
-Positive Affect During Aggression Scale (see Supplemental Table 1)
-Voodoo Doll Aggression Task (see Study 5)
-Victim Suffering Scale (see Supplemental Table 3)
-----Study 8-----
Cover Story: The study was introduced as a project meant to assess participants on a
reaction-time task. Other measures were explained as potential having effects on
reaction-time or are detailed in the sections below.
Independent Variables:
-Short Sadistic Impulse Scale (see Study 1)
SADISM AND AGGRESSION 98
-Varieties of Sadistic Tendencies Scale (available here:
http://www.psych.ubc.ca/~dpaulhus/Paulhus_measures/VAST.3.3.docx)
-Victim Suffering Manipulation text and protocol is explicitly detailed in the Procedures
subsection of the Study 8 description in the manuscript.
Dependent Variables:
-Victim Suffering Scale (see Supplemental Table 3)
-Taylor Aggression Paradigm (see Study 1)
-Positive Affect During Aggression Scale (see Supplemental Table 1)
... A data extraction sheet was developed, pilot-tested and refined accordingly (Appendix C). Additional information which was unavailable in one published article was requested and obtained from the author (Chester et al., 2019). During this process, one duplicate publication was noted; Buckels et al. (2013) was the published version of Buckels (2012). ...
... These were the: SSIS (n = 25); CAST (n = 15); VAST (n = 7); and the ASP (n = 2). Three studies combined items from these measures to create the Sadism Index (Chester et al., 2019) and the Direct Sadism Composite Score . ...
... High heterogeneity was still observed via the Q and I 2 statistic and sensitivity analysis showed that the overall moderate pooled effect size (0.39) remained (Appendix G). It was also noted that some studies (Buckels et al., 2013;Chester et al., 2019) could be viewed as measuring behavioural operationalisations of sadism, rather than aggressive behaviour. A sensitivity analysis showed that the overall moderate pooled effect size (0.36) remained when these studies were excluded. ...
Article
Previous research suggests that individuals with increased sadistic traits may seek out opportunities to exercise cruelty towards others. This study extends research which has examined the emotional processing of individuals with increased sadistic traits. Results showed that all dark traits were significantly correlated with experiencing positive emotion in response to violent stimuli. This study also examined the relationship between sadism and the temperamental traits of the behavioural inhibition and approach systems. Results suggest that sadism could be understood as a high-approach low-avoidance temperamental trait. A structural equation model predicting emotional processing was constructed; a direct pathway was found between sadism and positive emotional responding to violent stimuli and lack of negative emotion to violent stimuli, over and above latent dark tetrad antagonism. Dark tetrad traits were predictive of a deficit in BIS-Anxiety. Implications of these findings are discussed.
... A data extraction sheet was developed, pilot-tested and refined accordingly (Appendix C). Additional information which was unavailable in one published article was requested and obtained from the author (Chester et al., 2019). During this process, one duplicate publication was noted; Buckels et al. (2013) was the published version of Buckels (2012). ...
... These were the: SSIS (n = 25); CAST (n = 15); VAST (n = 7); and the ASP (n = 2). Three studies combined items from these measures to create the Sadism Index (Chester et al., 2019) and the Direct Sadism Composite Score . ...
... High heterogeneity was still observed via the Q and I 2 statistic and sensitivity analysis showed that the overall moderate pooled effect size (0.39) remained (Appendix G). It was also noted that some studies (Buckels et al., 2013;Chester et al., 2019) could be viewed as measuring behavioural operationalisations of sadism, rather than aggressive behaviour. A sensitivity analysis showed that the overall moderate pooled effect size (0.36) remained when these studies were excluded. ...
... was performed analogically as the provocation manipulation check (negative affect: M = 2.00, SD = 1.27; positive affect: M = 3.01, SD = 1.60). Participants responded using a scale ranging from 1 (not at all) to 5 (extremely).Act of revenge was assessed as the number of gruesome images (0-9) assigned by participants to their partner for detailed description (i.e., Gruesome Image Aggression Task;Chester et al., 2019;Gollwitzer & Bushman, 2012); M = 2.88, SD = 3.08. ...
Article
Full-text available
The extrinsic reward should impede revenge-seeking if revenge is solely driven by the desire to feel gratified. Study 1 (N = 114) showed that satisfaction from receiving monetary compensation decreased thinking about getting back at the provocateur. However, Study 2 (N = 213) found that insulted participants aggressed against their partners despite fulfillment from receiving the unexpected monetary reward. This evidence indicates that gratification is insufficient to impede revenge following provocation, suggesting that avengers want to feel pleasure when retaliating and want to balance the scales by sending offenders a message.
... Indeed, the Dark Tetrad traits have consistently positive associations with aggressive behaviors. It is worth mentioning that psychopathy and sadism stand out for their closer links with various types of aggression, including physical, verbal, between partners, prejudice, bullying, extremism, radicalization, contempt, workplace mistreatment, child corporal punishment, mild and moderate self-harm, and cheating (Chester et al., 2018;Chung et al., 2022;Fernández-del-Río et al., 2021;, 2020Min et al., 2019;Moor et al., 2019;Morgades-Bamba et al., 2018;Paulhus, Gupta et al., 2021;Pfattheicher et al., 2017;Pfattheicher, Schindler, & Nockur, 2018;Plouffe et al., 2020a;Ritchie et al., 2019;Schriber et al., 2017;Tetreault et al., 2018;Tetreault & Hoff, 2019). Narcissism was unrelated to violent behavior in most studies but was still connected to lighter forms of aggression, as the use of brainteaser questions in personal selection (Highhouse et al., 2018) and with preference for painting with violent motives (Tucaković & Marković, 2021). ...
Article
With this meta-analytic review, we aimed to estimate the relationship that sadistic personality has with the Dark Triad traits and, secondarily, describe the research on the Dark Tetrad traits. We searched for articles in the following databases, PsycARTICLES, PsycINFO, PubMed, ScienceDirect, SCOPUS, and Web of Science, where we found 128 articles to qualitatively evaluate and 103 articles to quantitatively analyze. Sadism correlated with narcissism (r = 0.26), Machiavellianism (r = 0.43), and psychopathy (r = 0.58). The most common themes across the studies were: (a) structural aspects of the Dark Tetrad; (b) online behavior; (c) aggressiveness; (d) moral beliefs and behavior; (e) video games; (f) sexual behavior, and (g) emotional functioning. The Dark Tetrad traits correlated with several dysfunctional behaviors and socially maladaptive outcomes. Finally, sadism is potentially more similar to psychopathy and Machiavellianism, than narcissism.
... Participant Characteristics from Study 3Measures Image-Based TAP. To measure aggression we implemented a modified, online variant of the TAP that employed aversive images as the harm-inducting stimuli instead of noise blasts (an Image-Based TAP; hereafter IB-TAP; similar toChester, DeWall, & Enjaian, 2019; Gollwitzer & ...
Preprint
Full-text available
Do people harm others with the broader altruistic goal of helping them? Across four studies, we empirically examined whether people believe in such prosocial aggression and whether they would enact it when given the opportunity. In Study 1 (N = 493), participants reported considerable belief in the existence of prosocial aggression, which was simultaneously associated with greater antisocial traits and greater altruism. We further observed a self-serving bias, in which participants indicated that their aggression was relatively more prosocially-motivated. In Studies 2 and 3 (combined N = 426), participants were often prosocially-aggressive, inflicting more harm when their aggression could help (versus only hurt) the target. In Study 4 (N = 285), such prosocial aggression was directed preferentially towards agreeable (versus antagonistic) targets. Our findings highlight the need to better understand how harm can arise from a desire to help and to revise theory to accommodate this behavioral phenotype.
Article
Purpose The purpose of this research is to investigate the concept of revenge against banks. Design/methodology/approach Structural equation modeling was used to test hypotheses based on collected survey data. A total of 625 questionnaires were collected from Lebanese customers. Findings Research findings identify multiple antecedents to consumers' desire for revenge against banks by encapsulating the cognitive, emotional, social, and personal patterns that influence the desire for revenge. The author found that the desire for revenge fuels direct behavioral reactions towards bank. Practical implications Consumer revenge remains underdeveloped in marketing research. This study provides managerial recommendations to assist bank response strategies in managing consumer revenge behaviors. Originality/value This study is one of very few that explores the concept of revenge against financial institutions, specifically by connecting the literature to the discovery of cognitive, affective, and social factors. This paper contributes to the existing body of knowledge by highlighting the role of personality traits in consumer revenge. This study’s research implications are built on unique findings in a developing country, while most extensive studies that boost negative public attitudes toward the banking industry are established in developed countries.
Article
Background Everyday sadism is unique among the dark personality traits that form the Dark Tetrad. Whilst it shares the Dark Core of Personality with psychopathy, narcissism, and Machiavellianism, it is theoretically distinct from the other members in that involves an appetite for cruelty. Objective Given mounting evidence of everyday sadism being associated with socially aversive behaviour, and its potential utility in predicting harmful behaviour within the general population; the current review examined the relationship between sadism and aggressive behaviour. Data sources Seven well-established electronic bibliographic databases and two other sources were systematically searched. Supplementary searches were employed. Study selection The initial search identified 627 full-text articles. Following application of inclusion and exclusion criteria, 39 full-text articles inclusive of 52 studies were subject to quality assessment. Data was extracted from 52 studies, with 48 studies eligible for quantitative synthesis. Data synthesis Correlational analyses were the most common method of examining sadism and aggression. A multi-method approach was employed to standardize effect sizes into a common metric (correlation coefficient). A meta-analysis was conducted using 57 effect sizes from 48 studies with 22,179 participants, revealing a pooled correlation coefficient of 0.35 (95% CI = 0.29, 0.40). Significant heterogeneity was observed which warranted sensitivity and subgroup analyses. Conclusions Overall, a moderate relationship was found between subclinical sadism and aggressive behaviour, as defined by acts ranging from verbal to physical, and sexual aggression and violence. A stronger, moderate relationship was found between sadism and perpetration of aggressive behaviour online, indicating that anonymity may unleash the darker side of personality within the general population. The implications of this finding are considered.
Article
Full-text available
Subclinical sadism, characterized by infliction of cruelty, aggression, or humiliation on another for subjugation or pleasure, provides important information in the prediction of aversive behaviors that have implications for individuals’ and society’s well-being worldwide. Given sadism’s universal relevance, it is imperative that researchers ensure valid and reliable trait measurement not only among English-speaking individuals, but also cross-nationally among countries in which sadism remains relatively understudied. The objective of the current research was to validate the revised version of the Assessment of Sadistic Personality (ASP-8) (Plouffe et al., 2017) across samples of Russian (n = 1087, Mage = 37.36, SD = 10.36), Greek (n = 1195, Mage = 35.64, SDage = 13.08), Serbian (n = 443, Mage = 28.10, SDage = 6.60), and British (n = 511, Mage = 28.50, SDage = 11.62) adults. Overall, results supported the reliability, dimensionality, and scalar/partial scalar measurement invariance of the ASP-8 across cross-national samples. Convergent and discriminant validity were mostly supported through correlations with general personality traits, the Dark Triad, emotional intelligence, mental toughness, depression, anxiety, stress, satisfaction with life, aggression, and attitudes toward social groups. Based on our findings, we recommend the use of the ASP-8 in future investigations of aversive traits.
Chapter
The present chapter presents traditional and contemporary work on the relevance of emotion for clinical forensic psychology, as well as the recent but rapidly growing literature on emotion regulation and offending. First, theoretical and empirical work is discussed that emphasizes the role of negative emotions for aggression and antisocial behavior, as well as the importance of focusing on the experience (or lack thereof) of specific emotions, including but not limited to anger. Then, a case is made that, in addition to and in interaction with the experience of emotions, the individual ability to regulate emotions is consequential for offending outcomes. Finally, these two streams of literature are combined to advance practical implications for clinical work (i.e., prevention and treatment) and policymaking (i.e., staffing and training).
Article
Full-text available
Aggressive behavior hurts us all and is studied across psychology's sub-disciplines. Classical theories discuss the causes of aggression in the context of negative affect (e.g., frustration, pain). However, more recent research implicates positive affect as an important correlate and cause of aggression. Such aggressive pleasure likely evolved from ancient predatory tendencies that later yielded reproductive benefits, holds across reactive and proactive forms of aggression, and is used strategically as an item in many people's emotion-regulation toolkit. Findings from psychological and neural sciences have converged to detail aggression's hedonically pleasant qualities and the motivational and biological mechanisms through which they occur. This new approach
Article
Full-text available
The Taylor Aggression Paradigm (TAP) is a frequently used laboratory measure of aggression. However, the flexibility inherent in its implementation and analysis can undermine its validity. To test whether the TAP is a valid aggression measure irrespective of this flexibility, we conducted two preregistered studies (Study 1 n = 177, Study 2 n = 167) of a standardized version of the TAP. Across both studies, TAP scores showed agreement with other laboratory aggression measures, were magnified by an experimental provocation manipulation, and were associated with traits typically linked to aggressive behavior. Mixed evidence was found for the external and discriminant validity of the task. Individual responses largely loaded onto a single component, suggesting that the aggregate scoring approach accurately represents the underlying data structure. These results provide preliminary support for the internal validity of this TAP approach and highlight the utility of preregistration in psychometric research.
Article
Full-text available
Social rejection is a painful event that often increases aggression. However, the neural mechanisms of this rejection-aggression link remain unclear. A potential clue may be that rejected people often recruit the ventrolateral prefrontal cortex's (VLPFC) self-regulatory processes to manage the pain of rejection. Using functional MRI, we replicated previous links between rejection and activity in the brain's mentalizing network, social pain network, and VLPFC. VLPFC recruitment during rejection was associated with greater activity in the brain's reward network (i.e., the ventral striatum) when individuals were given an opportunity to retaliate. This retaliation-related striatal response was associated with greater levels of retaliatory aggression. Dispositionally aggressive individuals exhibited less functional connectivity between the ventral striatum and the right VLPFC during aggression. This connectivity exerted a suppressing effect on dispositionally aggressive individuals' greater aggressive responses to rejection. These results help explain how the pain of rejection and reward of revenge motivate rejected people to behave aggressively.
Article
Full-text available
People differ in how much they seek retribution for interpersonal insults, slights, rejections, and other antagonistic actions. Identifying individuals who are most prone towards such revenge-seeking is a theoretically-informative and potentially violence-reducing endeavor. However, we have yet to understand the extent to which revenge-seeking individuals exhibit specific features of aggressiveness, impulsivity, and what motivates their hunt for retribution. Toward this end, we conducted three studies (total N = 673), in which revenge-seeking was measured alongside these other constructs. Analyses repeatedly demonstrated that revenge-seeking was associated with greater physical (but not verbal) aggressiveness, anger, and hostility. Revenge-seeking's link to physical aggression was partially accounted for by impulses toward enjoying aggression and the tendency to use aggression to improve mood. Dominance analyses revealed that sadism explained the most variance in revenge-seeking. Revenge-seeking was associated with greater impulsive responses to negative and positive affect, as well as greater premeditation of behavior. These findings paint a picture of revenge-seekers as physically aggressive curators of anger, whose retributive acts are performed with planned malice and motivated by the act's entertaining and therapeutic qualities.
Article
Full-text available
Aggressive behavior hurts us all and is studied across psychology’s subdisciplines. Classical theories discuss the causes of aggression in the context of negative affect (e.g., frustration, pain). However, more recent research implicates positive affect as an important correlate and cause of aggression. Such aggressive pleasure likely evolved from ancient predatory tendencies that later yielded reproductive benefits, holds across reactive and proactive forms of aggression, and is used strategically as an item in many people’s emotion-regulation toolkit. Findings from psychological and neural sciences have converged to detail aggression’s hedonically pleasant qualities and the motivational and biological mechanisms through which they occur. This new approach generates novel hypotheses and might lead to effective interventions that mollify mankind’s aggressive tendencies.
Article
Laboratory experiments investigating aggressive behavior have operationalized and assessed aggression in a variety of ways; however, these measures are often problematic because they do not create a situation in which participants perceive potential for real harm to come to the target, there is a risk of actual harm to the target, or they are too familiar to participants. To overcome these limitations, we developed a new method for measuring aggression, specifically, the amount of hot sauce administered to a target known to dislike spicy foods. We summarize a series of experiments assessing theory‐based hypotheses regarding aggression in which this measure is employed. We then briefly consider the strengths and limitations of this new measure. Aggr. Behav. 25:331–348, 1999. © 1999 Wiley‐Liss, Inc.
Article
On the basis of an evolutionary theory of self-esteem (SE), it was hypothesized that the SE-aggression relationship differs across functionally distinct domains of SE and across contexts. In 2 experiments, participants had the opportunity to aggress against the evaluator of an essay they had written. In Study 1, self-perceived superiority was positively related to aggression, whereas social inclusion was inversely related to aggression. In Study 2, in which the context was altered to simulate a mating competition, only a measure of self-perceived mate value emerged as a (positive) predictor of aggression. Global SE failed to contribute to the prediction of aggression in either experiment. Statistically controlling for narcissism did not eliminate either set of findings. Implications for the conceptualization and measurement of SE and narcissism are discussed.
Article
Aggression is often construed as a unitary trait fully captured by the Buss-Perry Aggression Questionnaire (BPAQ). Our review of the literature questions that assumption in several respects. Instead of a top-down approach, we argue for a bottom-up conception based on the Dark Tetrad of personality, that is, narcissism, Machiavellianism, psychopathy, and sadism. We highlight research showing that each member of the tetrad responds to different provocations. We conclude that the unitary trait conception of aggression has yielded more confusion than understanding. The term aggression should be reserved for outcomes, with many possible trait x situation predictors. Future research should continue the investigation of moderators as well as cognitive mediators to clarify the triggering of aggression in the individual tetrad members.
Article
No longer conceptualised as only for the “desperate”, online dating offers many benefits over face-to-face dating. Accompanying the benefits of online dating is the potential for new, distinct forms of antisocial behaviour online, such as trolling. The current study (N = 357) sought to explore the antisocial behaviour of trolling on Location-Based Real-Time Dating applications (i.e., LBRTD apps) in an online sample of Australians sourced from the community. Specifically, we examined the role of participant's sex and of the personality traits of narcissism, Machiavellianism, psychopathy, sadism, and impulsivity in predicting perpetration of trolling behaviours on LBRTD apps. Although there were no sex differences, the traits of psychopathy, sadism, and dysfunctional impulsivity were significantly associated with trolling behaviours. Subsequent moderation analysis revealed that dysfunctional impulsivity predicts perpetration of trolling, but only if the individual has medium or high levels of trait psychopathy. Results of the current study aid in further conceptualising the personality of the Internet “troll”. Future research should further explore antisocial online behaviours, such as other hostile behaviour that occurs on LBRTD apps.