Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=hmep20
Download by: [Tulane University] Date: 05 April 2016, At: 00:33
Media Psychology
ISSN: 1521-3269 (Print) 1532-785X (Online) Journal homepage: http://www.tandfonline.com/loi/hmep20
Repeated Play Reduces Video Games’ Ability
to Elicit Guilt: Evidence from a Longitudinal
Experiment
Matthew Grizzard, Ron Tamborini, John L. Sherry & René Weber
To cite this article: Matthew Grizzard, Ron Tamborini, John L. Sherry & René Weber (2016):
Repeated Play Reduces Video Games’ Ability to Elicit Guilt: Evidence from a Longitudinal
Experiment, Media Psychology, DOI: 10.1080/15213269.2016.1142382
To link to this article: http://dx.doi.org/10.1080/15213269.2016.1142382
Published online: 30 Mar 2016.
Submit your article to this journal
Article views: 8
View related articles
View Crossmark data
Repeated Play Reduces Video Games’ Ability
to Elicit Guilt: Evidence from a Longitudinal
Experiment
MATTHEW GRIZZARD
Department of Communication, University at Buffalo, State University of New York,
Buffalo, New York, USA
RON TAMBORINI and JOHN L. SHERRY
Department of Communication, Michigan State University, East Lansing, Michigan, USA
RENÉ WEBER
Department of Communication, University of California Santa Barbara,
Santa Barbara, California, USA
Statements supported mostly by correlational and cross-sectional
studies suggest that playing violent video games can cause emotional
desensitization. A longitudinal experiment examined a)
whether repeated violent game play leads to emotional desensitization
and b) whether desensitization generalizes to other play and
real-life experiences. Participants played alternative versions of the
same violent game for the first four days; on these days, the character
role was varied between-subjects to be moral (United Nations
soldier) or immoral (terrorist soldier). On Day 5, all participants
played a novel game as a terrorist. Results indicate two things. First,
habituation occurs over repeated game play: Repeated exposure
decreased the ability of the original game to elicit guilt. Second,
the decreased ability to elicit guilt can generalize to other gameplay
experiences: Guilt elicited by the novel game on Day 5 was
Address correspondence to Matthew Grizzard, Department of Communication, 359 Baldy
Hall, Buffalo, NY 14260-1060, USA. E-mail: mngrizza@buffalo.edu
The lead author would like to acknowledge Alison Grizzard for aiding in data collection
and Tim Pleskac (Senior Research, Center for Adaptive Rationality at the Max Planck Institute
for Human Development) for reviewing early drafts of this manuscript. The authors would also
like to acknowledge three anonymous reviewers for their critical and detailed reviews during
the review process.
Media Psychology, 00:1–24, 2016
Copyright © Taylor & Francis
ISSN: 1521-3269 print/1532-785X online
DOI: 10.1080/15213269.2016.1142382
1
Downloadedby[TulaneUniversity]at00:3305April2016
reduced for the immoral character condition compared to the
moral character condition. The current study provides causal,
longitudinal evidence regarding the potential for video game play
to lead to emotional desensitization with regard to future video
game-play experiences.
Determining the extent to which video games are capable of eliciting moral
emotions has attracted considerable scholarly attention (Grizzard, Tamborini,
Lewis, Wang, & Prabhu, 2014; Hartmann, Toz, & Brandon, 2010; Lin, 2011;
Schmierbach & Limperos, 2013). Moral emotions serve as “an emotional moral
barometer, providing immediate and salient feedback” regarding the social
and moral acceptability of an action (Tangney, Stuewig, & Mashek, 2007, p.
347). Research on the ability of video games to elicit moral emotions presents
a consistent, replicable finding: Players engaging in unjustified violence in
video games feel guilt for their virtual transgressions (e.g., Hartmann et al.,
2010). Notably, the findings of this research contrast strongly with gamers’
claims that killing virtual characters on a screen is no different than taking
pawns in a game of chess, with neither action being particularly relevant to
real-world morality (see Hartmann, 2012; Hartmann et al., 2010). The preponderance
of evidence clearly shows that engaging in unjustified violence in
a video game elicits significantly higher levels of guilt than engaging in
justified violence (cf., Grizzard et al., 2014; Hartmann et al., 2010). If virtual
violence was indeed morally irrelevant, then both unjustified and justified
violence should elicit similar levels of guilt. Thus, current evidence suggests
that video game actions are morally relevant.
In a somewhat interesting twist, video game scholars and avid gamers
both argue that playing the role of an immoral character in a violent video
game is unlikely to lead to antisocial consequences, albeit for different reasons.
Gamers tend to claim that virtual behaviors are morally inconsequential,
whereas researchers argue that because immoral game play elicits guilt, guilt
felt by the player may mitigate antisocial consequences of violent video game
play or even lead to prosocial responses (Grizzard et al., 2014; Hartmann
et al., 2010; Weaver & Lewis, 2012). This argument is supported by research
that shows guilt motivates prosocial behaviors (Boster et al., 1999; de Hooge,
Breugelmans, & Zeelenberg, 2008). However, the possibility remains that the
ability of games to elicit guilt may be reduced through repeated play. Prior
work on moral emotions and video game play examined the ability of games
to elicit guilt in single-exposure studies. Thus, it is plausible that the ability of
video games to elicit guilt might habituate over time: Initial exposures to
virtual moral transgressions elicit guilt, but repeated play leads to habituation
of this response. Consequently, game play could lose its capacity to induce a
moral response. If such effects become completely habituated, gamers may
experience their in-game actions as morally inconsequential.
2 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
The current study seeks to examine whether the elicitation of guilt
resulting from video game play is reduced with repeated play. Reduction of
response intensity associated with repeated stimulus presentation is labeled as
habituation in the academic literature (Groves & Thompson, 1970; Rankin
et al., 2009; Stein, 1966). This study also seeks to establish whether habituation
of the guilt response will extend or generalize to future play experiences.
Habituation coupled with generalization is often described as “desensitization”
in the video game aggression literature (cf. Carnagey, Anderson, & Bushman,
2007). This article begins with a discussion of the desensitization hypothesis
and its relationship to research on phobia reduction. We then present the
current study with a focus on its design and the specific hypotheses related to
habituation and generalization. Finally, we present and discuss the findings of
an experiment designed to test these hypotheses.
THE DESENSITIZATION HYPOTHESIS AND ITS RELATIONSHIP TO
VIDEO GAME THEORY AND RESEARCH
Habituation and generalization are established psychological processes with
decades of research characterizing patterns of response (Rankin et al., 2009).
Rankin et al. (2009) define habituation as “a behavioral response decrement
that results from repeated stimulation and that does not involve sensory
adaptation/sensory fatigue or motor fatigue” (p. 136). In the case of media
stimuli, viewers would be expected to show diminished psychophysiological
response after multiple exposures. Generalization is defined as the extension
of the habituation response to new stimuli that are similar in nature to the
original stimulus. For example, Faulkender, Wright, and Waldron (1974)
habituated toddlers to images of animals and then exposed the toddlers to a
mixed set of images of another category (fruit) and images of the same
category (other animals). Toddlers spent significantly less time looking at
images from the familiar category than from the new category, suggesting
that habituation generalized from the initial set of animal images to the new
set. Research on habituation and generalization has led video game researchers
and theorists to propose that aggressive video game play can lead to
increased aggression through desensitization (i.e., the desensitization hypothesis;
Bushman & Anderson, 2009; DeWall, Anderson, & Bushman, 2011). The
desensitization hypothesis is a combination of habituation and generalization
mechanisms with regard to media. Repeated exposure to media aggression is
thought to decrease both physiological responses (e.g., heart rate) and emotional
responses (e.g., fear, guilt) to similar future experiences, both mediated
and non-mediated.
One area where desensitization due to exposure to virtual environments
has been studied is for the treatment of phobias (Parsons & Rizzo, 2008).
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 3
Downloadedby[TulaneUniversity]at00:3305April2016
Treatment for phobias involves repeatedly exposing patients to fear-inducing
stimuli in a safe environment. Investigations have shown that this repeated
exposure results in both habituation and generalization (Foa & Kozak, 1986).
In many cases, repeated exposure can take place in virtual environments. For
example, Rizzo et al. (2009) used virtual reality to desensitize returning
soldiers with posttraumatic stress disorder. Research on phobia reduction
through repeated stimulus exposure is particularly important to the current
investigation in two ways. First, it provides a theoretical mechanism for
desensitization of guilt by suggesting that desensitization can affect both
physiological responses, such as heart rate, as well as emotional responses,
such as fear (Lang, Melamed, & Hart, 1970). Second, it shows that virtual
environments are capable of eliciting real-world desensitization (Tarr & Warren,
2002). At first glance, research showing that virtual environments are
capable of desensitizing emotional responses, such as fear, suggests that this
mechanism should work with other emotions, such as guilt. At the same time,
it is important to recognize that basic emotions, such as fear, differ from more
complex emotional responses, such as moral emotions. Unlike fear and other
basic emotions, moral emotions, such as guilt, rely on higher levels of cognitive
deliberation (Tangney et al., 2007). To experience guilt, individuals must
recognize that their actions have violated a moral or social tenet. Thus, it is
unclear whether desensitization will occur with emotions that require more
complex cognitive evaluations, such as guilt, or whether the cognitive evaluative
structures associated with the elicitation of guilt are so ingrained that
desensitization of these complex emotions is impossible.
Although desensitization logic is present in much of the theory on video
game violence (see Anderson & Carnagey, 2004) and is applied equally to
physiological and emotional responses, it is simply assumed to occur in most
investigations. Moreover, when it has been examined, causal claims have
been hindered by the cross-sectional nature of the research (see discussion
by Grizzard et al., 2015). Most studies examining desensitization are not
longitudinal and instead take place within a single laboratory sessions (cf.
Anderson et al., 2010; Bartholow, Bushman, & Sestir, 2006).
The current study is distinct from previous research on desensitization to
video game violence in key ways. First, unlike previous single-session studies,
the present investigation examines desensitization across multiple sessions.
Research comparing single-session versus multiple-session desensitization
training (Rowe & Craske, 1998) concludes that single-session desensitization
is a) less likely to generalize to novel stimuli and b) less likely to be
permanent.1
Second, unlike previous research that focused only on cognitive
or physiological desensitization, the current study examines emotional
responses. This fact is particularly important as current research in moral
psychology focuses on the importance of emotions in guiding cognitions
and behaviors. For example, dual-process theories of morality argue that
4 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
emotional responses underlie moral judgments and behaviors (for an overview,
see Eden, Grizzard, & Lewis, 2013, and Haidt, 2001).
THE CURRENT STUDY
The study reported here was a longitudinal experiment separated into two
phases to test the desensitization hypothesis. The habituation phase was
designed to test the study’s basic hypothesis that the ability of a video game
to induce moral emotions will dissipate with repeated play. During the habituation
phase, participants played the same game on four sequential occasions.
The generalization phase was included to establish that habituation is
not limited to the original game, but extends to new game stimuli. During the
generalization phase, participants played a new game.
The Habituation Phase
Literature regarding the amount of time necessary for habituation to occur due
to media exposure varies greatly. Several studies have demonstrated habituation
within a single session (Bartholow et al., 2006; Linz, Donnerstein, &
Adams 1989; Popova, 2009; Wilson, 1989). However, longitudinal experiments
over multiple days have also demonstrated habituation with regard to
violence and pornography (Linz, Donnerstein, & Penrod, 1984; Zillmann &
Bryant, 1982). The habituation schedule for the current research consisted of
four exposures to the same game across four separate days. This mirrored the
habituation schedule of Rowe and Craske’s (1998) phobia reduction study.
Game play on each day consisted of approximately 10–15 minutes, for a total
of 40–60 minutes across all exposures. This schedule should be sufficient for
establishing desensitization for several reasons. First, and as previously stated,
media research has demonstrated desensitization in single sessions with significantly
less exposure time (Wilson, 1989). Second, multiple exposures over
a series of days is more likely to lead to desensitization than a single session
(Rowe & Craske, 1998).
The current project adopted the stimuli used in the Hartmann et al.
(2010) study, a modified version of Operation Flashpoint. In their study,
participants engaged either in normatively immoral behaviors (i.e., playing
as a paramilitary terrorist who kills U.N. soldiers to defend a torture camp;
henceforth, referred to as the immoral condition) or in normatively moral
behaviors (i.e., playing as a U.N. solider who has to kill paramilitary
terrorists to liberate a torture camp; henceforth, referred to as the moral
condition). Thus, the current study replicates Hartmann et al.’s first
hypothesis.
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 5
Downloadedby[TulaneUniversity]at00:3305April2016
H1: Participants in the immoral condition will feel more guilt postgame play
than participants in the moral condition on Day 1 of the study.
Next, we hypothesize that the level of guilt elicited by the game should
decrease with repeated play in a manner that interacts with condition. Both
conditions should experience decreases in guilt over time (previous studies
show that even moral violence is capable of eliciting low levels of guilt;
Hartmann et al., 2010; Grizzard et al., 2014; Lowinger & Solomon, 2004), but
the reduction in guilt should be stronger for the immoral condition, as this
group is expected to elicit higher levels of guilt on Day 1 of the study (see
Figure 1 for a graphical depiction of the desensitization hypothesis).
H2: There will be an interaction between game condition and time, such that
the postgame play guilt difference between the moral and immoral conditions
will be smaller on Day 4 of the study than on Day 1 of the study.
The Generalization Phase
The generalization phase tests whether the effect of the immoral game on
reducing guilt will generalize to a similar three-dimensional shooter game. The
similar game is the “No Russian” level from the game, Call of Duty: Modern
Warfare 2 in which players help a terrorist organization carry out an attack at
a crowded airport by walking through the airport and killing any civilians they
encounter. The game is graphic and realistic in its depiction of the violence
and in the reactions of the civilian characters.
The game was selected for its similarity to the immoral condition of the
desensitization phase game. Both games include terrorist groups committing
unjust violence, both have identical controls, and both use the first-person
perspective. Previous research paradigms on emotional desensitization to
media content generally select novel content that is similar, but not identical,
FIGURE 1. Visual depiction of habituation and generalization hypotheses. Days 1–4 represent
the habituation phase of the study while Day 5 represents the generalization phase.
6 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
to the desensitizing content (Wilson, 1989; Zillmann & Bryant, 1982), as
generalization is a function of the similarity between stimuli (Shepard, 1987).
Consistent with previous research on habituation and generalization (Bradley,
Lang, & Cuthbert, 1993; Rowe & Craske, 1998; Zillmann & Bryant, 1982), it is
expected that participants in both conditions will experience an increase in
guilt when they are exposed to the novel stimuli as compared to the original
stimuli. However, if habituation of a general-nature occurred and it is lasting,
participants in the immoral game condition should experience a smaller
increase in guilt to the novel game as compared to the moral game condition,
as the immoral game has desensitized the individual to immoral violence
specifically. This leads to Hypotheses 3a and 3b, which are displayed graphically
in Figure 1. Note that the novel game is expected to be more guilt
inducing than the original game. This specificity in the prediction is based
on the fact that the novel game had more advanced graphics than the original
game, and is supported by past investigations on the influence of advanced
graphics on player responses in general (see Ivory & Kalyanaraman, 2007)
and on the arousal potential of the two specific games used in this study (see
Grizzard et al., 2015).
H3a: There will be an increase in postgame-play guilt for both conditions
between Day 4 of the study, when participants play the original game for the
last time, and Day 5 of the study, when participants play the novel game for
the first time.
H3b: There will be an interaction between game condition and time, such that
the increase in postgame-play guilt between Day 4 and Day 5 will be greater
for participants in the moral condition as compared with participants in the
immoral game condition.
METHOD2
Procedure Overview
The study employed a mixed design with a game condition factor and a
repeated measures factor. The between-subjects factor was the game condition
(moral and immoral). The repeated measures factor was time (Days 1–5).
Days 1–4 represented the habituation phase, during which participants played
the Operation Flashpoint game used by Hartmann et al. (2010). Day 5
represented the generalization phase examining generalization of the habituation
response to future game-play experiences, during which participants
played the aforementioned, novel Call of Duty game.
To account for potential testing effects associated with repeatedly measuring
guilt, a separate between-subjects factor was added to the study. This
additional, between-subjects factor represented two separate groups of participants
(Group A and Group B). During the desensitization phase, both
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 7
Downloadedby[TulaneUniversity]at00:3305April2016
groups played Flashpoint over the four days. The assessment of guilt, however,
differed depending on the group to allow for a modified Solomon fourgroup
design. Group A completed the guilt measure on all six days of the
experiment; Group B completed the guilt measure only on Days 4 and 5 (see
Table 1). The group between-subjects factor allows for the testing of hypotheses
related to desensitization through Group A and hypotheses related to
generalization through the combination of Group A and Group B. Participants
were randomly assigned to conditions.
Participants
A power analysis based on the effect size observed in Hartmann et al. (2010),
Cohen’s d = .87, assuming power level of β = .80, and α = .05 (two-tailed)
indicated that 22 participants per condition was sufficient to detect a significant
effect. A total of 145 university undergraduates were recruited to take part
in the experiment. Undergraduates are an appropriate sample for this study
because it seems improbable that video game-induced desensitization would
affect college students differently than other adult populations. Moreover, it
has been noted elsewhere that college students represent a population particularly
well suited to studying video game-related psychological processes
(Anderson & Dill, 2000). Ninety-five participants (65.54%) identified themselves
as female, 49 (33.8%) identified themselves as male, and 1 (0.7%) did
not disclose their sex. The majority of participants identified themselves as
White or Caucasian (n = 105, 72.4%). Minority races accounted for the
following: Asian (n = 12, 8.3%), Black or African American (n = 5, 3.4%),
Hispanic (n = 2, 1.4%), Native American (n = 1, 0.7%), and other (n = 4, 2.8%);
16 participants (11.0%) did not disclose their race. The average age of participants
was 19.53 years (SD = 1.64, min = 18, max = 27). Seventy participants
were randomly assigned to Group A (35 to the immoral condition, 35 to the
moral condition), and 75 participants were randomly assigned to Group B (35
TABLE 1 Procedure Illustrating Guilt Assessment and Induction by Day
Desensitization phase Generalization phase
Condition Day 1 Day 2 Day 3 Day 4 Day 5
Group A
Immoral O O O O O
Moral O O O O O
Group B
Immoral O O
Moral O O
Induction Game Game Game Game Novel game
Note. O = represents observation of guilt.
8 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
to the immoral condition, 40 to the moral condition), who had guilt assessed
only after Day 4.
Measures
Guilt. The current project employed similar measures as Hartmann et al.
(2010) to replicate their findings. Their scale consisted of three items: “When
playing the game, how often did you feel…” (1) “regret,” (2) “sorry about
something you did,” and (3) “like you did something wrong.” The current
study added three items: guilty, blameworthy, and ashamed, taken from the
guilt subscale of the PANAS-X (Watson & Clark, 1994). The prompt was also
changed to “What extent do you feel…,” to ensure that guilt observed was not
simply a demand characteristic (i.e., participants thinking about how they
should have felt when playing the game rather than how they felt in the
moment). Participants responded on a 0 (not at all) to 10 (extremely) scale. A
confirmatory factor analysis using structural equation modeling conducted on
the data from Day 1 indicated a one-factor solution fit the data well after
dropping the blameworthy and ashamed items,3
χ2
(2) = 1.86, p = .39, CFI =
1.00, RMSEA = .00, SRMR = .01. The four items were averaged to form a
composite for each day, and internal consistency was high across all days
(minimum observed α = .95).
In addition to guilt, the current study also recorded several covariates
thought to be important. These included the participants’ sex, video gameplay
habits, state affect and state arousal, and familiarity with the games used
in the study.
Video Game-Play Habits. If the hypothesized relationships regarding
desensitization (both habituation and generalization) exist in the population,
then participants who are gamers are likely to have experienced desensitization
already. Thus, two factors related to video game play were also measured
using items created for this study: A five-item attitudes toward video games
scale (sample items include “I consider myself to be a gamer” and “I enjoy
playing video games”; α = .86) and a question regarding the amount of time
spent playing video games per week (“On average, how many hours do you
spend playing video games each week? Remember to include weekdays and
weekends;” Response scale: 1 = 0 hours (I never play); 2 = just a few hours a
year; 3 = 1 or 2 hours a month; 4 = 1 or two hours a week; 5 = 1 or 2 hours a
few days each week; 6 = 1 or 2 hours every day; 7 = more than 2 hours a day).
State Affect and State Arousal. Guilt is an emotion that has both cognitive
(i.e., recognizing that one’s action contradicts one’s values) and affective
components (i.e., feeling negative emotions resulting from recognition). As
such, the negative affect associated with guilt is relatively specific in nature
(i.e., associated with the recognition, not negative affect in general). To
account for the relationship between guilt and generalized negative affect
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 9
Downloadedby[TulaneUniversity]at00:3305April2016
and to control for such a relationship in hypothesis tests, affect and arousal
were measured after game play on each day of the study using the affect grid
(Russell, Weiss, & Mendelsohn, 1989). The affect grid is a 9-×-9 table that plots
arousal (low to high) from 1 to 9 on the y-axis and affect (unpleasant to
pleasant) from 1 to 9 on the x-axis. The midpoint of both factors (5 on each
axis) represents neutral affect and arousal.
Game Familiarity. Finally, familiarity with the games used in the study
was measured on the final day of the study. Participants were asked how
familiar they were with the games in general and more specifically whether
they had played or heard of the “No Russian” mission from Call of Duty:
Modern Warfare 2.
Procedure
The procedure was relatively similar across all days of the study. Upon arrival
at the lab, informed consent was gathered. Next, participants’ blood pressure
was measured through an automatic blood pressure machine as part of a data
collection unrelated to the current study (see Grizzard et al. [2015] for a
reporting of these data). Next, participants completed the affect grid so that
they would become familiar with its use. Following this, participants were
informed of the game that they would be playing through the use of recorded
instructions, and the controls of the game were described. Next, participants
played the game; the Operation Flashpoint game play lasted for approximately
7–10 minutes and the novel game, Call of Duty: Modern Warfare 2
was played for approximately the same amount of time. Immediately following
game play, participants completed the guilt measure, followed by another
reading of the blood pressure machine, the affect grid, and several questions
about perceptions of the game (see Grizzard et al. [2015] for a discussion of
these data). The only deviation from these procedures that occurred was on
the final day of the study when participants’ demographic information and
their familiarity of the games were measured at the end of the survey.
RESULTS
Inclusion of Covariates
Prior to hypothesis testing, a correlation matrix of the independent variable
(Condition: coded 0 = moral, 1 = immoral), covariates (sex, attitudes toward
video games, amount of time playing video games, state affect, state arousal,
and game familiarity), and the dependent variable guilt (Day 1) was examined
to determine which covariates to include in hypothesis tests. Covariates
were included in analyses, if a) they were significantly correlated with the
dependent variable across all days of the study and b) they were not
10 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
correlated with the independent variable. In addition, if covariates correlated
with each other greater than r ≥ .50, the covariate that correlated most
strongly with the dependent variable was included and the others were
excluded. This decision was made to ensure that power would not be
reduced by the inclusion of multiple covariates with high intercorrelations.
Based on these criteria, two covariates were determined to be important for
inclusion in analyses: sex and affect. Neither sex (dummy-coded; 0 = female,
1 = male) nor affect correlated significantly with condition (r = –.05 and r =
–.09, respectively), but both correlated significantly with guilt (rsex = –.36, p <
.001, and raffect = –.60, p < .001). Three potential covariates were not
included, despite their positive correlation with guilt, due to their high
correlation with sex. These covariates were attitudes toward video games
(rsex = .53, rguilt = –.30), game hours played (rsex = .75, rguilt = –.36), and
familiarity with Call of Duty: Modern Warfare 2 (rsex = .58, rguilt = –.33).
Notably, sex and game hours played correlated similarly with guilt (both r =
–.36). The inclusion of both covariates in analyses presented in the paper did
not alter the pattern of findings for other predictors, and only reduced the
sex effect to zero (F < 1), as would be expected for covariates with a high
correlation.
Participant Mortality
A manipulation check question was included on each day that asked participants
whether the character they controlled during the game was “a terrorist,”
“a U.N. solider,” or “I don’t remember,” with participants answering incorrectly
being removed from analysis. Rolling elimination was used as a method of
excluding participants who answered incorrectly. For example, if a participant
answered incorrectly on Day 1, he was excluded from all analyses. However,
if a participant answered correctly on Day 1 and answered incorrectly on Day
5, she was included in analyses that dealt with Days 1–4, but was excluded
from all analyses that dealt with Day 5. Participants were also excluded from
analyses if they: a) missed experimental sessions, b) did not complete the guilt
measures on all days, or c) did not complete the covariate measures.
Hypothesis 1 examined the difference between conditions on Day 1 for
participants in Group A. Sixty eight of the 70 participants recruited into Group
A were included in analysis of Hypothesis 1 based on the above criteria
(97.1% inclusion rate). Hypothesis 2 examined an interaction between time
(Day 1–4) and condition, and Hypothesis 3 examined an interaction between
time (Days 4 and 5) and condition. The combined tests of Hypotheses 2 and 3
included 56 participants (80.0% inclusion rate) from Group A. The tests of
Hypothesis 3 where Groups A and B were combined included 114 participants
in total (Group A, n = 56 and Group B, n = 58, for 80.0% and 77.3%
inclusion rates, respectively).
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 11
Downloadedby[TulaneUniversity]at00:3305April2016
Hypothesis Testing
Hypothesis 1—Main Effect on Day 1. Hypothesis 1 predicted that guilt
would be higher for participants in the immoral game condition than for
participants in the moral game condition. An analysis of covariance
(ANCOVA) was conducted on participants from Group A with condition
(immoral vs. moral) as the between-subjects factor, Guilt as the dependent
variable, and affect and sex as the covariates. Results from the test indicated
significant effects of both covariates, sex F(1, 64) = 5.01, p = .03, η2
= .05, and
affect F(1, 64) = 23.96.36, p < .001. η2
= .23. The direction of the parameter
estimates indicates that males felt less guilt than females and positive affect
was associated with less guilt. More notably, there was a main effect of
condition, F(1, 64) = 5.59, p = .02, η2
= .05. Examination of the estimated
marginal means indicate that participants in the immoral condition experienced
significantly more guilt than participants in the moral condition (Mimmoral
= 3.33, SD = 2.43, Mmoral = 1.93, SD = 2.43), consistent with the first
hypothesis and replicating Hartmann et al. (2010) and Grizzard et al. (2014).
Hypotheses 2 and 3—Interactions Between Condition and Time. Combined
together, Hypotheses 2 and 3 predicted a specific pattern across all
five days of the study (see Figure 1). To reduce the number of tests and
inflation of Type I errors as well as to provide a test of the specified pattern,
a focused contrast analysis (see Rosenthal, Rosnow, & Rubin, 2000) was
conducted to test these hypotheses in a single statistical test for participants
in Group A. A contrast analysis allow for a powerful statistical test of
hypothesized patterns in data by examining the extent to which the
observed data adhere to a predicted pattern. This approach is particularly
preferred in ANOVA designs that incorporate more than two levels of an
independent variable (e.g., time in the current study). Contrast analyses
require the predicted pattern to be described in contrast coefficients that
correspond to the predicted pattern (e.g., a hypothesized positive linear
increase in a one-way design with three levels of the independent factor
would be –1 0 +1) and sum to zero. Note that this approach is far superior
than a simple F test as the F test only tells the researcher that a significant
difference exists somewhere between the three levels of the independent
factor, and it tells the researcher nothing about the extent to which their data
fit the predicted pattern (see Rosenthal et al., 2000).
The contrast coefficients used to test the combined hypotheses were
created to represent the pattern found in Figure 1, but recentered so that
they would sum to zero (as required for contrasts). These contrast coefficients
can be found in Table 2 and are visually depicted in Figure 2. To test the
combined hypotheses using a contrast analysis, contrast scores (L scores)
were calculated for each participant, and the significance of the contrast was
tested using the standard t test approach (see Rosenthal et al., 2000).
12 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
Results of the t test on the contrast score were significant, t(54) = 5.91, p <
.001, Cohen’s d = 1.61, indicating that the data in the study were consistent
with both the habituation and generalization hypotheses. In addition, the
pattern of observed means correlates with the pattern of predicted means r
(8) = .98, p < .001 (see Figure 3). Although this analysis is purely for heuristic
purposes, the high degree of correlation indicates that the observed data fit
the predicted pattern well. We also tested this contrast with the presence of
the covariates. This was accomplished through a regression approach
whereby guilt on each day was regressed onto condition, sex, and affect for
the specific day. The predicted values from these regressions were then used
as the outcome variable for the contrast test. The results of this contrast were
similarly significant, but with a larger effect size, as would be expected when
controlling for additional variance in guilt elicited by the covariates, t(54) =
11.45, p < .001, Cohen’s d = 3.12.
FIGURE 2. Visual depiction of the contrast coefficients used to test Hypotheses 2 and 3.
FIGURE 3. Observed and predicted means standardized to be on the same scale.
TABLE 2 Contrast Coefficients used in Contrast Analysis for Combined Testing of Habituation
and Generalization
Condition Day 1 Day 2 Day 3 Day 4 Day 5 Sum
Immoral 1 0 –1 –2 3 1
Moral –1 –1.33 –1.67 –2 5 –1
Sum 0 –1.33 –2.67 –4 8 0
Note. Row effects and column effects sum to zero as required for contrast analyses on interactions.
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 13
Downloadedby[TulaneUniversity]at00:3305April2016
Further Analysis of Hypothesis 3. In addition to the contrast analysis, we
further examined Hypothesis 3a and 3b. These analyses included participants
from Group B, who only provided data on Days 4 and 5 of the study. A mixed
ANOVA was conducted with group (Group A vs. Group B) and condition
(immoral vs. moral) as between-subjects factors and time (Day 4 and Day 5)
as a within-subjects factor. The multivariate tests indicated a significant effect
of time, Wilks’ λ = .56, F(1, 110) = 85.38, p < .001, η2
p = .44, indicating that guilt
was higher on Day 5 (M = 5.50, SD = 3.60) than Day 4 (M = 2.72, SD = 2.89),
and a significant interaction of time × condition, Wilks’ λ = .95, F(1, 110) =
5.27, p =. 02, η2
p = .05, indicating that while the immoral condition (M = 3.03,
SD = 3.10) experienced more guilt on Day 4 than the moral condition (M =
2.44, SD = 2.68), the immoral condition (M = 5.10, SD = 3.41) experienced less
guilt on Day 5 than the moral condition (M = 5.86, SD = 3.75). Both of these
effects are consistent with the hypotheses.
Importantly, the Time × Group interaction and the Time × Group ×
Condition interactions were both nonsignificant indicating that the two-way
interaction of Time × Condition did not differ based on group, Wilks’ λ = 1.00,
F(1, 110) = .05, p = .83, η2
p = .00, and Wilks’ λ = .99, F(1, 110) = 1.31, p = .26,
η2
p = .01, respectively. The between-subjects effect of Condition was nonsignificant,
F(1, 110) = .00, p = .97, η2
p = .00, and the interaction of condition
with group was also nonsignificant, F(1, 110) = .002, p = .97, η2
p = .00. There
was a significant main effect of group, F(1, 110) = 5.21, p = .02, η2
p = .05, with
Group B (M = 4.70, SD = 2.83) scoring significantly higher than Group A (M =
3.50, SD = 2.81). However, because group did not interact with Time, Condition,
or Time × Condition, this main effect simply indicates that the values of
the significant effects of time and Time × Condition were higher for Group B
as compared to Group A. Overall, the results are consistent with Hypotheses
3a and 3b as indicated by the significant effects of a) Time, b) Time ×
Condition and the nonsignificant effects of c) Condition, d) Condition ×
Group, e) Group × Time, and f) Group × Time × Condition. See Figure 4 for
a depiction of the interaction with Groups A and B combined.
FIGURE 4. Hypothesis 3a and 3b pattern of means for conditions from Day 4 to Day 5.
14 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
DISCUSSION
Previous research indicates that engaging in unjust or immoral behaviors in a video
game is capable of eliciting the moral emotion of guilt (Grizzard et al., 2014;
Hartmann et al., 2010; Lin, 2011; Weaver & Lewis, 2012). Video game scholars
have argued that the ability of immoral game play to elicit guilt suggests two
important implications. First, video game play is morally consequential (Hartmann
et al., 2010; Lin, 2011; Weaver & Lewis, 2012). In other words, the actions undertaken
by players in a video game do not occur in a moral vacuum. Instead, virtual
behaviors interact with an individual’s sense of real-world morality and, thus, are
capable of eliciting guilt. Second, because immoral game play induces guilt,
assuming the role of an immoral character in a video game should reduce the
potential for antisocial consequences associated with violent video games (Grizzard
et al., 2014; Hartmann et al., 2010; Weaver & Lewis. 2012). Hartmann and
colleagues (2010) argue that guilt should cause players to “enter a more critical and
distanced reception mode, which may also serve to diminish the aggressive effects
of displayed virtual violence” (p. 357). Weaver and Lewis (2012) echo this theoretical
argument: “If a player has moral agency and feels guilty for a certain behavior
or choice, a certain level of conscious consideration of the repercussion of one’s
behavior is implied” (p. 613). Grizzard and colleagues (2010) provided empirical
evidence consistent with this logic showing that guilt elicited by video game play
led to increased moral sensitivity related to the moral transgression committed
during game play. This research indicated potential domain-specific mechanisms
related to virtual immoral behavior leading to increased moral sensitivity of moral
issues related to the virtual behavior. The findings of the current study support the
proposition that video game play is morally consequential but challenge the
proposition that routinely assuming an immoral character would mitigate the
aggressive effects of violent video game play.
Desensitization to Moral Violations through Repeated Game-Play
Although past research argues that guilt-inducing video games might lead to
prosocial responses including increased moral sensitivity (see Grizzard et al.,
2014; Hartmann et al., 2010; Weaver & Lewis, 2012), the current findings modify
this hypothesis by demonstrating that the ability of game play to elicit guilt
decreases over time and that this decreased sensitivity generalizes to future
game-play experiences. These findings indicate that if atrocious virtual behaviors
increase moral sensitivity (see Grizzard et al., 2014), the increases may be short
lived, as the increase in moral sensitivity seems to be dependent on the mediating
effect of guilt. Importantly, these findings indicate that desensitization, which has
been found for basic emotions, may occur for complex emotions that require
cognitive deliberation. At the same time, the question of why this type of
desensitization occurs for guilt requires future research.
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 15
Downloadedby[TulaneUniversity]at00:3305April2016
As stated in the introduction, guilt is a rather complex emotion. It requires
an individual to recognize that he or she has done something to violate
internal standards of behavior, and it is this recognition that brings about
accompanying negative affect. The findings of the current study do not
indicate whether it is the recognition that changes with repeated play or
whether the resultant affect is less intense. For example, it could be the case
that the desensitization observed in the current study was purely the result of
players altering the cognitive recognition process of guilt; through repeatedly
engaging in virtual behaviors, players came to view their actions as less
morally consequential. In this manner, the desensitization observed here
may simply reflect the formation of a cognitive schema related to the processing
of video game violence. This hypothesis reflects a suggestion made by
Raney (2004) when he suggested that media may have its own moral codes
that are qualitatively unique from real-world morality (see also Young &
Whitty, 2010). This interpretation is similar to the tunnel vision hypothesis
(Breuer, Scharkow, & Quandt, 2014).
A second potential explanation regarding our observation of guilt desensitization
is that repeated play did not alter the cognitive recognition process,
but reduced the amount of negative affect associated with the response. For
example, it may be the case that players on Day 4 recognized their actions to
be just as immoral as on Day 1, but they did not feel as much negative affect
leading to the observed pattern. This explanation would suggest that it was
the emotional processes of guilt that were altered rather than the cognitive
processes. The data in the current study cannot rule out either of these
explanations. Neither would biophysiological data related to arousal, as both
explanations would predict reduced arousal. In fact, the only way to rule out
one of these explanations may be through functional magnetic resonance
imaging (fMRI) data.
If the first explanation is correct and desensitization relates to changes in
recognition, fMRI data should indicate time-dependent decreased connectivity
between executive prefrontal and subcortical limbic areas. If, on the other hand,
desensitization is related purely to changes in strength of emotional responsiveness
with no change in recognition, fMRI data should indicate time-dependent
decreased activations in limbic areas that are largely independent from activity
in executive networks (see Koban & Pourtois, 2014, and Wagner, N’Diaye,
Ethofer, & Vuilleumier, 2011). Moreover, this type of research could also
determine the relevance of a third explanation related to moral disengagement
(see Breuer et al., 2014). It may be the case that the desensitization observed in
the current study was the result of post hoc emotion regulation. According to
this explanation, players recognize that they did something wrong; guilt and
negative affect arise; and the player then uses cognitive strategies to reduce
resulting negative affect (see Bandura, 2001, for a discussion of moral disengagement;
see also Hartmann, 2012, for a description of how moral disengagement
might relate to video game behaviors). If this explanation were
16 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
responsible for the underlying pattern, fMRI research should indicate a specific
temporal dynamic between subcortical and executive networks such that the
upregulation of executive networks moderates or suppresses subcortical activation
(see McRae et al., 2012). Finally, a reverse mechanism—that is, some sort of
anticipated emotion regulation—may also be possible: Players who engage in
violent video game play already know (or have learned through repeated game
play) that their behaviors within the game violate social norms and actively
suppress the emotional response (guilt, empathy, etc.) in favor of executive
control, which is crucial to master the game. This mechanism would be consistent
with previous findings in neuroimaging studies on playing violent video
games (see Weber, Ritterfeld, & Mathiak, 2006; see also, Mathiak & Weber,
2006) and studies on the elicitation of disgust (see Gallo, McCulloch, & Gollwitzer,
2012). Future research may be able to disentangle these alternative
explanations, each associated with specific hypotheses related to connectivity
patterns and their dynamics within the brain. At the same time, particularly
clever designs may be able to provide evidence of these competing explanations
using self-report measures, especially if one could establish expected
patterns that are dissociated with one of the proposed mechanisms.
The Moral Consequence of Actions Preformed in Virtual Environments
The findings of the current study add to building evidence that virtual actions
are capable of eliciting moral emotions, such as guilt. Thus far, most research
of this type has focused on moral emotions elicited by moral transgressions,
such as unjust violence. Questions remain as to whether moral upholding in
virtual environments (e.g., being kind and compassionate) are equally as
capable of eliciting positive moral emotions, such as moral pride and elevation
(see Oliver, Hartmann, & Woolley, 2012). Past research (Baumeister, Bratslavsky,
Finkenauer, & Vohs, 2001) showing greater sensitivity to moral
violations than moral upholding indicates that guilt may be an easier moral
emotion to elicit than moral pride. Future research should attempt to determine
whether this is the case or not. The stimuli used in this study could
provide a test of this hypothesis by examining whether the binomial relationship
between condition and guilt is similar in magnitude but in the opposite
direction for measures of moral pride.
Generalization to Moral Emotions Elicited by Non-mediated Stimuli
Research on habituation and generalization, the mechanisms that underlie
desensitization, indicate that two outcomes are possible when habituation
occurs: 1) stimulus generalization, the extension of habituation effects to
novel stimuli; and 2) stimulus specificity, the failure of generalization to occur
(Rankin et al., 2009). Although the findings here indicate some stimulus
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 17
Downloadedby[TulaneUniversity]at00:3305April2016
generalization (i.e., to future play experiences), generalization to non-mediated
stimuli may not occur, as the probability of generalization is negatively related
to the sensory intensity of the eliciting stimulus (Rankin et al., 2009). For
particularly strong stimuli, such as non-mediated stimuli, generalization may
not occur at all. For example, participants may habituate to video game violence,
including horrendous acts of virtual torture, rape, and murder, but that
habituation may not extend to their viewing the same actions in the real world.
Accordingly, we are cautious in describing the ability of video game-induced
desensitization to generalize outside of the game until such a finding is demonstrated
empirically (see Young & Whitty, 2010, for further discussion).
Limitations
There are several limitations of the current study worthy of mention. First, it
is important to note that no other specific emotion was measured in the
current study except for guilt. Thus, it is possible that any guilt elicited from
game play is simply a demand characteristic. At the same time, there are
several reasons that a demand characteristic explanation is unlikely to
account for the entirety of the data. Foremost among these is the fact that
the study predicted an interaction between condition and time resulting in a
very specific pattern of means. It seems unlikely that demand characteristics
would lead to such an interaction. In fact, a simple demand-characteristic
explanation would suggest a reduction in variance associated with the
between-subject factor. Given that most studies on video game violence
utilize a violent versus nonviolent comparison, participants in the moral
condition would have likely assumed that they were in the violent condition
of the study, artificially inflating their guilt responses. Still, future research
should include measures of other specific emotions (e.g., sadness, happiness,
anger, etc.), which could provide additional discriminant validity
regarding the potential for demand characteristics.
Second, the data revealed an unexpected pattern regarding the level of
guilt experienced on Day 4 by Group B. Although not statistically significant,
Group B participants did report slightly more guilt on Day 4 of the study than
Group A participants reported on Day 1 (see Figure 5). Although this finding
seems at odds with the desensitization phenomenon, it is important to remember
that desensitization is inherently a within-subjects effect. That two
between-subjects groups would not represent a pattern expected for desensitization
is not problematic, especially when one considers that the interaction
contrasts conducted on the within-subjects data were consistent with hypotheses,
and the ANOVA conducted demonstrated that group did not significantly
interact with condition, time, or the two way interaction of condition and time.
Still, this pattern is worthy of mention and demonstrates the notable problems
of utilizing between-subjects designs to examine within-subjects effects.
18 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
This idiosyncratic finding also suggests future research questions regarding
guilt. It is possible that asking participants about their levels of guilt on
each day exacerbated habituation, or that waiting until Day 4 to ask Group B
exacerbated the guilt they reported. This post hoc explanation relates to the
amount of virtual harm done before the guilt question was asked for both
groups. For participants in Group A, the guilt questions were asked after the
participants had played the game only once. For participants in Group B, the
guilt questions were asked after the participant had played the game four
times. This difference resulted in the participants of Group B committing more
virtual transgressions than participants in Group A prior to the initial assessment
of guilt. Thus, asking the guilt questions on a day-to-day basis could
have caused participants in Group A to consider their behaviors with respect
to the current day alone while waiting to ask the question until Day 4 could
have caused participants in Group B to consider their in-game behaviors with
respect to all four previous days. Although this explanation seems plausible, it
should be considered speculative until future research can provide a critical
test of this hypothesis. Notably, this is another area where the inclusion of
measures of other specific emotions might be useful, as they could serve as a
masking agent for the effects on guilt.
The third limitation relates to the mortality rate of the participants. All
longitudinal studies suffer from participant mortality, and some might consider
the rates of mortality here to be moderate, particularly with respect to
the test of Hypothesis 4. However, participant mortality is only a concern
regarding the internal validity of an experimental study if there is an interaction
between mortality and condition such that participants who dropped
out of the experiment are systematically different from participants who
remained in. A chi-square test crossing participant mortality (remained in
vs. dropped out) with the four conditions (moral vs. immoral for Groups A
and B) was conducted. The results indicated that mortality was unrelated to
experimental condition, χ2
(3) = .52, p = .92. Furthermore, correlations
between participant mortality and the dependent variables and covariates
FIGURE 5. Comparison of guilt induced by play on Day 1 for Group A versus Day 4 for
Group B.
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 19
Downloadedby[TulaneUniversity]at00:3305April2016
were also nonsignificant indicating that participants who dropped out were
not significantly different from participants who remained in the study on
any of the covariates or dependent variables. In the current case, participant
mortality seemed to be random and unassociated with the covariates, independent
variables, or dependent variables. Because participant mortality
was random, it does not limit or restrict the validity of the current findings;
the only effect should be a loss of power.
CONCLUSION
The current article provides evidence regarding potential effects of
repeated video game play on the ability of video games to elicit moral
emotions. The current research indicates that any mitigating effect of such
play (cf. Grizzard et al., 2014) is likely to be short lived, as repeated play
desensitizes players to feeling video game-induced guilt. At the same time,
it is important to note that there is no evidence of video game-induced
desensitization generalizing to real-world behaviors. This proposition
requires additional research, as it would be faulty to assume that such
generalization is guaranteed to occur. Still, this study is the first to provide
experimental evidence of desensitization of moral emotional responses
within a video game. This type of longitudinal experimental framework
may, thus, be useful for further exploring the complex relationship
between video game play and moral emotions.
NOTES
1. In this research (Rowe & Craske, 1998), spider-fearful individuals underwent desensitization
training. Some of these participants experienced all of the desensitization training (the duration of which
was 1 hour and 45 minutes) in a single session on Day 1 of the experiment; the desensitization training for
the other participants was diffused over four days. Researchers held the total amount of desensitization
training time constant between the two groups. Notably, the single-session group experienced a reoccurrence
of fear to spiders (i.e., evidence that the desensitization they experienced was not permanent),
whereas the multi-session group did not.
2. Data were also collected to examine the possibility of generalization of habituation to real-world
judgments. On the day following the lab sessions, a novel guilt induction was administered in an online
survey. The novel guilt induction asked participants to imagine themselves in a situation in which they had
done something to wrong another person (adapted from de Hooge et al., 2008). Subsequent discussion of
this induction illuminated several confounds that made us distrust the results, so we decided to not include
this induction in this article. We do not believe this has an effect on the other tested hypotheses because it
occurred one day subsequent to all other data collection. Please contact the first author if interested in these
results.
3. That the blameworthy and ashamed items had a negative effect on the fit of a one factor solution is
not surprising. Guilt and shame are similar but distinct moral emotions with the former being associated
with actions (i.e., “I did something bad”) and the latter being associated with the self (i.e., “I am bad”;
Niedenthal, Tangney, & Gavanski, 1994).
20 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
REFERENCES
Anderson, C. A., & Carnagey, N. L. (2004). Violent evil and the general aggression
model. In A. Miller (Ed.), The social psychology of good and evil (pp. 168–192).
New York, NY: Guilford.
Anderson, C. A., & Dill, K. E. (2000). Video games and aggressive thoughts, feelings,
and behaviors in the laboratory and in life. Journal of Personality and Social
Psychology, 78, 772–790. doi:10.1037/0022-3514.78.4.772
Anderson, C. A., Shibuya, A., Ihori, N., Swing, E. L., Bushman, B. J., Sakamoto, A., …
& Saleem, M. (2010). Violent video game effects on aggression, empathy, and
prosocial behavior in eastern and western countries: A meta-analytic review.
Psychological bulletin, 136(2), 151–173. doi:10.1037/a0018251
Bandura, A. (2001). Social cognitive theory of mass communication. Media Psychology,
3, 265–299. doi:10.1207/S1532785XMEP0303_03
Bartholow, B. D., Bushman, B. J., & Sestir, M. A. (2006). Chronic violent video game
exposure and desensitization to violence: Behavioral and event-related brain
potential data. Journal of Experimental Social Psychology, 42, 532–539.
doi:10.1016/j.jesp.2005.08.006
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is
stronger than good. Review of General Psychology, 5, 323–370. doi:10.1037/
1089-2680.5.4.323
Boster, F. J., Mitchell, M. M., Lapinski, M. K., Cooper, H., Orrego, V.O., & Reinke, R.
(1999). The impact of guilt and type of compliance-gaining message on compliance.
Communication Monographs, 66, 168–177. doi:10.1080/03637759909376470
Bradley, M. M., Lang, P. J., & Cuthbert, B. N. (1993). Emotion, novelty, and the startle
reflex: Habituation in humans. Behavioral Neuroscience, 107, 970–980.
doi:10.1037/0735-7044.107.6.970
Breuer, J., Scharkow, M., & Quandt, T. (2014). Tunnel vision or desensitization?
Journal of Media Psychology, 26, 176–188. doi:10.1027/1864-1105/a000122
Bushman, B. J., & Anderson, C. A. (2009). Comfortably numb: Desensitizing effects of
violent media on helping others. Psychological Science, 20, 273–277.
doi:10.1111/j.1467-9280.2009.02287.x
Carnagey, N. L., Anderson, C. A., & Bushman, B. J. (2007). The effect of video game
violence on physiological desensitization to real-life violence. Journal of Experimental
Social Psychology, 43, 489–496. doi:10.1016/j.jesp.2006.05.003
de Hooge, I. E., Breugelmans, S. M., & Zeelenberg, M. (2008). Not so ugly after all:
When shame acts as a commitment device. Journal of Personality and Social
Psychology, 95, 933–943. doi:10.1037/a0011991
DeWall, C. N., Anderson, C. A., & Bushman, B. J. (2011). The general aggression
model: Theoretical extensions to violence. Psychology of Violence, 1(3), 245–258.
doi:10.1037/a0023842
Eden, A., Grizzard, M., & Lewis, R. J. (2013). Moral psychology and media theory:
Historical and emerging viewpoints. In R. Tamborini (Ed.), Media and the moral
mind (pp. 1–25). New York, NY: Routledge.
Faulkender, P. J., Wright, J. C., & Waldron, A. (1974). Generalized habituation of
concept stimuli in toddlers. Child Development, 45, 1002–1010. doi:10.2307/
1128088
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 21
Downloadedby[TulaneUniversity]at00:3305April2016
Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: Exposure to corrective
information. Psychological Bulletin, 99, 20–35. doi:10.1037/0033-2909.99.1.20
Gallo, I. S., McCulloch, K. C., & Gollwitzer, P. M. (2012). Differential effects of various
types of implementation intentions on the regulation of disgust. Social Cognition,
30, 1–17. doi:10.1521/soco.2012.30.1.1
Grizzard, M., Tamborini, R., Lewis, R. J., Wang, L., & Prabhu, S. (2014). Being bad in a
video game can make us morally sensitive. Cyberpsychology, Behavior, & Social
Networking, 17, 499–504. doi:10.1089/cyber.2013.0658
Grizzard, M., Tamborini, R., Sherry, J. L., Weber, R., Prabhu, S., Hahn, L., & Idzik,
P. (2015). The thrill is gone, but you might not know: Habituation and
generalization of biophysiological and self-reported arousal responses to
video games. Communication Monographs, 82, 64–87. doi:10.1080/03637751.
2014.971418
Groves, P. M., & Thompson, R. F. (1970). Habituation: a dual-process theory. Psychological
Review, 77, 419–450. doi:10.1037/h0029810
Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach
to moral judgment. Psychological Review, 108, 814–834. doi:10.1037/0033-
295X.108.4.814
Hartmann, T. (2012). Moral disengagement during exposure to media violence:
Would it feel right to shoot an innocent civilian in a video game? In R.
Tamborini (Ed.), Media and the moral mind (pp. 109–131). New York, NY:
Routledge.
Hartmann, T., Toz, E., & Brandon, M. (2010). Just a game? Unjustified virtual violence
produces guilt in empathetic players. Media Psychology, 13, 339–363.
doi:10.1080/15213269.2010.524912
Ivory, J. D., & Kalyanaraman, S. (2007). The effects of technological advancement and
violent content in video games on players’ feelings of presence, involvement,
physiological arousal, and aggression. Journal of Communication, 57, 532–555.
doi:10.1111/j.1460-2466.2007.00356.x
Koban, L., & Pourtois, G. (2014). Brain systems underlying the affective and social
monitoring of actions: an integrative review. Neuroscience & Biobehavioral
Reviews, 46, 71–84. doi:10.1016/j.neubiorev.2014.02.014
Lang, P. J., Melamed, B. G., & Hart, J. (1970). A psychophysiological analysis of fear
modification using and automated desensitization procedure. Journal of Abnormal
Psychology, 76, 220–234. doi:10.1037/h0029875
Lin, S. (2011). Effect of opponent type on moral emotions and responses to video
game play. Cyberpsychology, Behavior, & Social Networking, 14, 695–698.
doi:10.1089/cyber.2010.0523.
Linz, D., Donnerstein, E., & Adams, S. M. (1989). Physiological desensitization and
judgments about female victims of violence. Human Communication Research,
15, 509–522. doi:10.1111/j.1468-2958.1989.tb00197.x
Linz, D., Donnerstein, E., & Penrod, E. (1984). The effects of multiple exposures to
filmed violence against women. Journal of Communication, 34, 130–147.
doi:10.1111/j.1460-2466.1984.tb02180.x
Lowinger, T., & Solomon, Z. (2004). PTSD, guilt, and shame among reckless drivers.
Journal of Loss and Trauma, 9, 327–344. doi:10.1080/15325020490477704
22 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016
Mathiak, K., & Weber, R. (2006). Toward brain correlates of natural behavior: fMRI
during violent video games. Human Brain Mapping, 27, 948–956.
McRae, K., Gross, J. J., Weber, J., Robertson, E. R., Sokol-Hessner, P., Ray, R. D.,…&
Ochsner, K. N. (2012). The development of emotion regulation: an fMRI study of
cognitive reappraisal in children, adolescents and young adults. Social Cognitive
and Affective Neuroscience, 7, 11–22. doi:10.1093/scan/nsr093
Nelissen, R. M. A., Dijker, A. J., & De Vries, N. H. (2007). How to turn a hawk into a
dove and vice versa: Interactions between emotions and goals in a give-some
dilemma game. Journal of Experimental Social Psychology, 43, 280–286.
doi:10.1016/j.jesp.2006.01.009
Niedenthal, P. M., Tangney, J. P., & Gavanski, I. (1994). “If only I weren’t” versus “if
only I hadn’t”: Distinguishing shame and guilt in counterfactual thinking. Journal
of Personality and Social Psychology, 67, 585–595. doi:10.1037/0022-3514.
67.4.585
Oliver, M. B., Hartmann, T., & Woolley, J. K. (2012). Elevation in response to
entertainment portrayals of moral virtue. Human Communication Research,
38, 360–378. doi:10.1111/j.1468-2958.2012.01427.x
Parsons, T. D., & Rizzo, A. A. (2008). Affective outcomes of virtual reality exposure
therapy for anxiety and specific phobias: A meta-analysis. Journal of Behavior
Therapy and Experimental Psychiatry, 39, 250–261. doi:10.1016/j.jbtep.
2007.07.007
Popova, L. (2009, May). Compassion desensitization: Comparing reactions to disaster
and fictional television violence. Paper presented at the annual meeting of the
International Communication Association, Chicago, IL.
Raney, A. A. (2004). Expanding disposition theory: Reconsidering character liking,
moral evaluations, and enjoyment. Communication Theory, 14, 348–369.
doi:10.1111/j.1468-2885.2004.tb00319.x
Rankin, C. H., Abrams, T., Barry, R. J., Bhatnagar, S., Clayton, D. F., Colombo, J.,…&
Thompson, R. F. (2009). Habituation revisited: an updated and revised description
of the behavioral characteristics of habituation. Neurobiology of Learning
and Memory, 92, 135–138. doi:10.1016/j.nlm.2008.09.012
Rizzo, A. A., Difede, J., Rothbaum, B. O., Johnston, S., McLay, R. N., Reger, G., … &
Pair, J. (2009). VR PTSD exposure therapy results with active duty OIF/OEF
combatants. Studies in Health Technology and Informatics, 142, 277–282.
doi:10.3233/978-1-58603-964-6-277
Rosenthal, R., Rosnow, R. L., & Rubin, D. B. (2000). Contrasts and effect sizes in
behavioral research: A correlational approach. New York, NY: Cambridge University
Press.
Rowe, M. K., & Craske, M. G. (1998). Effects of an expanding-spaced versus massed
exposure schedule on fear reduction and the return of fear. Behavior Research
and Therapy, 36, 701–718. doi:10.1016/S0005-7967(97)10016-X
Russell, J. A., Weiss, A., & Mendelsohn, G. A. (1989). Affect grid: A single-item scale of
pleasure and arousal. Journal of Personality and Social Psychology, 57, 493–502.
doi:10.1037/0022-3514.57.3.493
Schmierbach, M., & Limperos, A. M. (2013). Virtual justice: Testing disposition theory
in the context of a story-driven video game. Journal of Broadcasting & Electronic
Media, 57(4), 526–542. doi: 10.1080/08838151.2013.845828
Repeated Play Reduces Video Games’ Ability to Elicit Guilt 23
Downloadedby[TulaneUniversity]at00:3305April2016
Shepard, R. N. (1987). Toward a universal law of generalization for psychological
science. Science, 237, 1317–1323. doi:10.1126/science.3629243
Stein, L. (1966). Habituation and stimulus novelty: A model based on classical conditioning.
Psychological Review, 73, 352–356. doi:10.1037/h0023449
Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). Moral emotions and moral
behavior. Annual Review of Psychology, 58, 345–372. doi:10.1146/annurev.
psych.56.091103.070145
Tarr, M. J., & Warren, W. H. (2002). Virtual reality in behavioral neuroscience and
beyond. Nature Neuroscience, 5, 1089–1092. doi:10.1038/nn948
Wagner, U., N’Diaye, K., Ethofer, T., & Vuilleumier, P. (2011). Guilt-specific processing
in the prefrontal cortex. Cerebral Cortex, 21, 2461–2470. doi:10.1093/cercor/
bhr016
Watson, D., & Clark, L. A. (1994). The PANAS-X: Manual for the positive and negative
affect schedule—Expanded form. Retrieved from http://www.psychology.uiowa.
edu/faculty/clark/panas-x.pdf
Weaver, A. J., & Lewis, N. (2012). Mirrored morality: An exploration of moral choice in
video games. Cyberpsychology, Behavior, & Social Networking, 15, 610–614.
Retrieved from http://online.liebertpub.com/doi/full/10.1089/cyber.2012.0235#
utm_source=PR&utm_medium=email&utm_campaign=CYBER
Weber, R., Ritterfeld, U., & Mathiak, K. (2006). Does playing violent video games
induce aggression? Empirical evidence of a functional magnetic resonance imaging
study. Media Psychology, 8(1), 39–60.
Wilson, B. J. (1989) Desensitizing children’s emotional reactions to the mass media.
Communication Research, 16, 723–745. doi:10.1177/009365089016006001
Young, G., & Whitty, M. T. (2010). Games without frontiers: On the moral and
psychological implications of violating taboos within multi-player virtual
spaces. Computers in Human Behavior, 26, 1228–1236. doi:10.1016/j.chb.
2010.03.023
Zillmann, D., & Bryant, J. (1982). Pornography, sexual callousness, and the trivialization
of rape. Journal of Communication, 32, 10–21. doi:10.1111/j.1460-
2466.1982.tb02514.x
24 M. Grizzard et al.
Downloadedby[TulaneUniversity]at00:3305April2016