Content uploaded by Kostas A Fanti
Author content
All content in this area was uploaded by Kostas A Fanti on Mar 05, 2021
Content may be subject to copyright.
Content uploaded by Kostas A Fanti
Author content
All content in this area was uploaded by Kostas A Fanti on Apr 21, 2019
Content may be subject to copyright.
AGGRESSIVE BEHAVIOR
Volume 35, pages 179–187 (2009)
Desensitization to Media Violence Over a Short Period
of Time
Kostas A. Fanti
1
, Eric Vanman
2
, Christopher C. Henrich
3
, and Marios N. Avraamides
1
1
Department of Psychology, University of Cyprus, Nicosia, Cyprus
2
Department of Psychology, University of Queensland, Brisbane, Australia
3
Department of Psychology, Georgia State University, Atlanta, Georgia
:::::::::::::::::::::::::::::::::::::::::
This study investigated the desensitization to violence over a short period of time. Participants watched nine violent movie scenes
and nine comedy scenes, and reported whether they enjoyed the violent or comedy scenes and whether they felt sympathetic toward
the victim of violence. Using latent growth modeling, analyses were carried out to investigate how participants responded to the
different scenes across time. The findings of this study suggested that repeated exposure to media violence reduces the
psychological impact of media violence in the short term, therefore desensitizing viewers to media violence. As a result, viewers
tended to feel less sympathetic toward the victims of violence and actually enjoy more the violence portrayed in the media.
Additionally, desensitization to media violence was better represented by a curvilinear pattern, whereas desensitization to comedy
scenes was better represented by a linear pattern. Finally, trait aggression was not related to the pattern of change over time,
although significant effects were found for initial reports of enjoyment and sympathy. Aggr. Behav. 35:179– 187, 2009.
r
2009
Wiley-Liss, Inc.
:::::::::::::::::::::::::::::::::::::::::
Keywords: media violence; desensitization; aggression; victim of violence; latent growth modeling
INTRODUCTION
More than 30 years ago, the possibility that
television (TV) and film violence may have adverse
effects on viewers was publicly presented [Steinfeld,
1972]. Exposure to TV, video game and film violence
may lead to several problematic outcomes, including
increased aggressive and violent behavior, increased
aggression-related thoughts or the accessibility of
violent constructs in memory and desensitization to
real-life aggression and to the suffering of victims
[e.g., Anderson et al., 2003; Berkowitz, 1984; Bush-
man, 1998; Gerbner et al., 1994; Gunter, 1994;
Huesmann and Kirwil, 2007; Huesmann et al., 1997,
2003; Linz et al., 1988, 1989; Malamuth and Check,
1981; Paik and Comstock, 1994].
This study was designed specifically to investigate
the process of desensitization. Desensitization has
been defined as the diminished emotional respon-
siveness to a negative or an aversive stimulus after
repeated exposure to it [Wolpe, 1982]. Initial
exposure to media violence typically produces
aversive responses such as fear, increased heart rate,
perspiration, discomfort and disgust, which is
consistent with the viewpoint that human beings
have innate negative responses to observing violence
[e.g., Cantor, 1998]. However, after prolonged and
repeated exposure across a person’s lifetime, the
psychological impact of screen-based media vio-
lence, including TV, movies and video games,
reduces or habituates; the observer becomes emo-
tionally and cognitively desensitized to media
violence across time [Cline et al., 1973; Flowers
et al., 2000; Funk et al., 2004; Huesmann and
Kirwil, 2007; Pennell and Browne, 1999; Smith and
Donnerstein, 1998; Thomas et al., 1977; Wei, 2007].
For example, Cline et al. [1973] and Thomas et al.
[1977] found that individuals who experienced high
levels of TV violence in the past were more likely to
be desensitized to violent film clips compared
with viewers who watched only small amounts of
TV violence. Furthermore, the previous research
Published online 26 January 2009 in Wiley InterScience (www.
interscience.wiley.com). DOI: 10.1002/ab.20295
Received 9 April 2008; Revised 14 October 2008; Accepted 15
October 2008
Correspondence to: Kostas A. Fanti, Department of Psychology,
University of Cyprus, P.O. Box 20537, CY 1678 Nicosia, Cyprus.
E-mail: kfanti@ucy.ac.cy
r
2009 Wiley-Liss, Inc.
provides evidence that children’s video game
exposure predicts emotional desensitization, indi-
cated by lower empathy or sympathy, and video
game and movie exposure predicted cognitive
desensitization, indicated by higher pro-violence
attitudes [Flowers et al., 2000; Funk et al., 2004;
Wei, 2007].
Chronic media violence exposure seems to have
long-term effects on viewers’ attitudes toward
violence and victims of violence through the process
of desensitization [Huesmann and Kirwil, 2007;
Smith and Donnerstein, 1998], but the exact
mechanisms of this desensitization are not yet fully
understood. One possibility is that repeated expo-
sure to violent media eliminates inhibitions to
violence [e.g., Carnagey et al., 2007]. Bartholow
et al. [2006] extended this hypothesis by showing
that exposure to violent images leads to blunted
evaluative categorization of violent stimuli. Their
study involved recording event-related potentials
(ERPs) from the scalps of violent and nonviolent
gamers during the presentation of neutral, violent
and negative nonviolent images within a context of
neutral images. Particular emphasis was given on
measuring the amplitude of the P300 component of
the ERP, which was previously reported to reflect
the extent of evaluative categorization of emotion-
ally relevant stimuli [i.e., greater P300 amplitudes
are observed when processing stimuli that are
inconsistent with the current context; Cacioppo
et al., 1994]. The study by Bartholow et al. [2006]
documented that, compared with nonviolent ga-
mers, violent gamers exhibited reduced P300 ampli-
tudes during exposure to violent images but not
during exposure to neutral or negative nonviolent
images. Furthermore, the amplitude of the P300
correlated negatively with the presence of aggressive
behavior during a subsequent task. The findings of
Bartholow et al. [2006] suggest that desensitization
to violent media takes the form of lowered inhibi-
tions toward violence and weakened evaluative
categorizations of violent stimuli.
Although many studies have provided evidence
for long-term desensitization toward violence after
chronic exposure to violent media, a number of
studies have also shown that traces of desensitiza-
tion are evident immediately after or even during a
single exposure to film violence [e.g., Lazarus et al.,
1962; Linz et al., 1989; Mullin and Linz, 1995]. Such
findings raise concern by suggesting that even brief
exposure to media violence may alter people’s
physiological and affective reactions to violence
and thus cause increased aggression, reduced
sympathy for victims of violent acts, decreased
probability of helping victims, and so on. Even
though the negative consequences of short-term
desensitization to violence are alarming, it is not yet
clear how such desensitization comes about and how
it develops in the short term after exposure to
violent media. For example, it is not clear whether
desensitization gradually increases during repeated
exposures to violence or whether it follows a step
function instead (i.e., it increases abruptly at a
specific point in time).
Furthermore, it is not yet known how aggressive
personality characteristics are related to short-term
desensitization. One possibility is that individuals
with high trait aggression have permanently reduced
inhibition to aggression and blunted evaluative
categorization of violent stimuli [Bartholow et al.,
2006; Huesmann and Kirwil, 2007]. If this is the
case, then only modest (if any) short-term desensi-
tization should be expected to take place for these
individuals. Additionally, previous studies have
linked aggression to higher pro-violence attitudes
and lower empathic responding [Anderson et al.,
2006; Mehrabian, 1997; Miller and Eisenberg, 1988],
which are both outcomes of desensitization. Because
of their pro-violence attitudes and lower empathy or
sympathy toward victims of violence, individuals
with high trait aggression might score higher on
cognitive desensitization and lower on emotional
desensitization.
The goal of this study was to investigate how
desensitization develops over a short period of time
that includes repeated exposure to media violence.
Participants were asked to view short video seg-
ments that included acts of violence, and their
reactivity was assessed through a questionnaire
administered after exposure to each segment. We
hypothesized that with repeated exposure (1) parti-
cipants would report higher levels of enjoyment
(indicating pro-violence attitudes), and (2) they
would report lower levels of sympathy for the
victims of violence (indicating lower empathic
responding). According to the previous research,
empathic responding—the capacity to perceive and
experience the state of another—reflects emotional
desensitization, whereas pro-violence attitudes sig-
nify cognitive desensitization [Eron, 2001; Funk
et al., 2004].
Additionally, trait aggression was used as an
indicator to examine any potential variability in
the development of desensitization between partici-
pants. On the basis of findings linking aggression to
lower empathy and higher pro-violence attitudes
[Anderson et al., 2006; Mehrabian, 1997; Miller and
Eisenberg, 1988], we expected individuals scoring
180 Fanti et al.
Aggr. Behav.
higher on aggression to enjoy the violent scenes
more and to be less concerned for the suffering of
victims of violence. Of particular importance for the
aims of this study is whether aggression will be
related to change over time, which will suggest
differences in the pattern of desensitization.
Finally, we examined whether short-term desensi-
tization to violence is unique to violent stimuli or
whether it reflects a more general process of
desensitization. To test this possibility we include a
condition in which participants view a series of video
segments that contained comedic elements instead of
violence. Inclusion of the comedy scenes provided a
test of whether desensitization applies to any
emotion or whether it is limited to negative stimuli
such as media violence. In the case that participants
get desensitized to both violent and comedy scenes,
we can infer that desensitization is not unique to
negative stimuli.
METHOD
Participants
The sample consisted of 96 college students (50%
female; age 18–26). The sample was diverse in terms
of ethnicity, including 13% Asian Americans, 27%
African Americans, 47% Caucasian Americans, 6%
Hispanic Americans and 7% non-US citizens. The
range for hours watching TV during a typical week
for this sample was also diverse, ranging from 0 to
56 viewing hours per week. Hours watching TV was
not correlated with viewers’ aggressiveness
(r5.08, P5.43), and was not significantly corre-
lated with the overall variables measuring whether
participants enjoyed the violent (r5.01, P5.93) or
comedy (r5.06, P5.53) scenes or sympathized
with the victim (r5.06, P5.56). Therefore, we did
not control for hours watching TV in the analyses
reported below.
Procedure
This study was introduced to students in intro-
ductory psychology classes as research in which they
could participate in exchange for course credit. The
participants were informed that they will be required
to watch different violent movie scenes, and that we
were interested to understand their reactions to
violent movie content. The cover story stated: ‘‘You
are invited to participate in a research study of
media violence and reactions to viewing violence.
The purpose of this study is to understand adults’
reactions to media violence. If you participate, you
will (1) view different segments of violent and
comedy clips, (2) complete a questionnaire before
the beginning of viewing the video and (3) complete
a questionnaire at the end of each video clip.’’
Participants who agreed to participate and signed
the informed consent completed the survey with the
demographic information and reported the number
of ‘‘extreme and inappropriate’’ aggressive acts that
they had committed during the last 6 months, along
with a description of these acts. After the initial
survey was completed, participants watched nine
violent movie scenes and nine comedy video clips.
After each clip participants were given 3 min to fill
out a questionnaire designed to measure whether
they enjoyed the violent or comedy scenes and
whether they were sympathetic toward the victim in
the violent scenes.
Violent and comedy scenes. Videotapes with
nine 2-min violent movie scenes and nine 2-min
comedy scenes with a 3-min break in between each
scene were prepared. Viewers watched one violent
scene followed by a comedy scene across time. The
scenes were in random order, and videotapes were
prepared that differed in the order of the scenes.
Assignment to a particular videotape was also
random. All the comedy scenes were from the series
‘‘Friends’’ and each scene depicted a different
episode. The violent scenes included short segments
from different movies. The scene from the movie
‘‘Elephant’’ depicts an adolescent shooting his fellow
students. The scene from the movie ‘‘Rembetico’’
depicts a domestic violence scene with the father
using corporal punishment toward his daughter and
also hitting the mother. The scene from the movie
‘‘Clockwork Orange’’ depicts a group of men
breaking into an apartment and acting violently
toward an elderly couple. Two scenes from the
movie ‘‘Fightclub’’ depict physical fighting between
two people with one of the characters being the
victim. The scene from the movie ‘‘The Vigilante’’
displays a group of drug dealers who mercilessly kill
in a gross and excessive way those who resist or
frustrate them. The scene from the movie ‘‘48
Hours’’ show several gun battles with vivid details
of the physical effects of the bullets. The scene from
the movie ‘‘The French Connection’’ shows a police-
man pursuing a murderer. During the pursuit, the
murderer kills another person. The scene from the
movie ‘‘Soylent Green’’ shows an attempt of a man
to evade several armed men chasing and threatening
him with clenched fists, guns and knives. Realism,
amount of physical damage and the emotionality
and suffering of victims were displayed in all the
selected violent scenes.
181Desensitization To Media Violence
Aggr. Behav.
After each scene participants were asked whether
they had previously watched the particular movie,
and according to their reports, 29.2% (n528) of the
participants had not watched any of the violent
movies presented, 12.5% (n512) had watched one
of the movies, 26% (n525) had watched two of the
movies, 21.9% (n521) had watched three of
the movies and 10.4% (n510) had watched four
of the movies. According to these findings the
selected violent clips were mostly unfamiliar to the
study’s participants. After each comedy scene,
participants were also asked whether they watched
the particular episode, and according to their reports
34.4% (n533) of the participants had not watched
any of the comedy scenes presented, 23% (n522)
had watched one to two of the comedy scenes,
14.6% (n514) had watched three to five of the
comedy scenes and 28% (n531) had watched six to
nine of the comedy scenes.
Measures
Real-life aggression. In this study we asked
participants whether they engaged in real-world
aggression. More specifically, participants were
asked to report the number and describe the
‘‘extreme and inappropriate’’ aggressive acts that
they had committed in the last 6 months. The
definition of an aggressive act that we used was
‘‘hitting and/or verbally insulting another person or
breaking objects because you are angry or fru-
strated.’’ This question was derived from the
Aggressive Acts Questionnaire [Barratt, 1991].
Participants described their aggressive acts, and
provided information in terms of the day it
happened, the duration (how long did the mood
last?) of the aggressive act and the time of the day
during which it happened. Moreover, in the case
that the description did not indicate an extreme or
inappropriate act, then it was not considered as one.
An example for an inappropriate/extreme act would
be, ‘‘I engaged in a physical and verbal fight with a
stranger for more than an hour because I was angry
with him,’’ and for a nonaggressive act, ‘‘I was
frustrated and slightly raised my voice toward my
mother, but it only lasted for a split of a second and
then I apologized.’’ Even though it might be
inappropriate to raise your voice toward your
mother, doing so for less than a second does not
make the act extreme. Therefore, to be regarded as
an aggressive act the description needed to indicate
an extreme and inappropriate act and the duration
needed to last for at least a couple of minutes. Two
researchers, the principal investigator and a research
assistant who was trained to identify extreme and
inappropriate acts, rated the participants’ responses
of aggression as extreme and inappropriate. All the
identified extreme aggressive acts were either physi-
cal, verbal or both. The intercoder agreement was
100%. Additionally, the participants seemed to have
a good grasp of the question, and the coders only
identified five responses that were not extreme or
inappropriate, including the example mentioned
above. Therefore, we created a trait aggression
variable based on the count of aggressive acts
participants reported. Participants’ responses ranged
from zero to seven aggressive acts.
Movie ratings. The reactivity of viewers to
violent media segments was assessed based on a self-
report questionnaire. The participants completed
four items on how ‘‘enjoyable’’ or ‘‘entertaining’’
they found the different violent and comedy scenes
(‘‘I enjoyed watching the scene,’’ ‘‘The scene was
amusing,’’ ‘‘The video segment was entertaining,’’ ‘‘I
did not like or enjoy the scene’’ (reverse coded)). The
items were adapted from the Film Evaluation
Instrument [Linz et al., 1984]. Additionally, partici-
pants responded to four questions designed to
measure their empathic responding and how much
they sympathized with the victim of violence (‘‘At
the end of the clip I felt sorry for the victim of
violence,’’ ‘‘I felt sympathy for the victim of violence
because he/she appeared to be in physical pain,’’ ‘‘I
felt sympathy for the victim of violence because he/
she appeared to suffer emotionally,’’ ‘‘During the
violent episode the victim showed fear and anxi-
ety’’). These items were adapted from the Victim
Evaluation Instrument [Linz et al., 1984]. Partici-
pants rated their responses using a scale ranging
from 0 (strongly disagree)to4(strongly agree). The
questions were added together to create an overall
score for each of the variables. Cronbach’s afor
each of these scales ranged from 0.82 through 0.96
for the different movie and comedy scenes.
Data Analysis
We conducted latent growth curve modeling to
investigate how participants responded to the
different violent and comedy scenes across time
within a structural equation modeling framework
using the Mplus 5.1 software package [Muthe
´n and
Muthe
´n, 2006]. Our first step was to fit a growth
model with no predictors. This model enabled us to
test whether change over time was linear or whether
it followed a more complex, curvilinear or cubic,
pattern over time. A quadratic or a cubic pattern of
change was accepted in the case that the growth
182 Fanti et al.
Aggr. Behav.
term was significant and its addition to the equation
improved the model fit. In addition, the use of
growth curve modeling enabled the identification of
the average intercept and growth terms for the nine
scenes. The residual growth factors were also used to
suggest whether there is variability in terms of the
initial levels of the variables and in terms of change
over time. In the case that variability was identified
in the intercept or growth terms, trait aggression was
used as a predictor. For a better depiction of the
analysis, Figure 1 illustrates a hypothetical curvi-
linear model in which variability has been identified
for the intercept and the linear and quadratic terms,
and because of that aggression was used to predict
initial levels and linear and quadratic change over
time.
Latent growth curve modeling uses a polynomial
function to model the relationship between an
attribute, in this case attitudes toward media
violence or comedy scenes and time [Muthe
´n and
Muthe
´n, 2006]. For example, in the case that we
identify a curvilinear pattern of change, the function
takes the form [Singer and Willett, 2003]:
yit ¼b0þb1Timeit þb2Time2
it þe
where yit is a latent variable that characterizes the
level of attitudes toward media violence or comedy
scenes for the participant iat time t. Timeit refers to
the specific scene that the participant iwatched at
time t, Time2
it is the square of the scene that the
participant iwatched at time tand eis a disturbance
assumed to be normally distributed with zero mean
and constant variance. The model’s coefficients, b
0
,
b
1
and b
2
, determine the average shape of the
trajectory. For this type of analysis the intercept,
linear and quadratic terms are assumed to take on a
normal distribution [Hedeker and Gibbons, 1994].
Three standard fit indexes were used in addition to
the w
2
statistic to evaluate model fit: the root mean-
square error of approximation (RMSEA), the
standardized root mean residual (SRMR) and the
Comparative Fit Index (CFI). Values less than .06
for the RMSEA and less than .09 for the SRMR are
considered to be a close fit, and a value higher than
.90 for CFI is considered to be acceptable [Hu and
Bentler, 1998; Kline, 1998]. Furthermore, according
to Hu and Bentler [1998], obtaining these values for
the fit indexes minimizes Type I and Type II error
rates.
RESULTS
Analysis Investigating Participants’ Responses
in Terms of Scene Enjoyment
To investigate the average trajectories of partici-
pants’ media violence enjoyment across time, a
linear growth curve model was initially estimated.
The linear growth model did not fit the data well,
w2
ð40;N¼96Þ¼365:80, RMSEA 5.116 (RMSEA CI:
.87|.145), SRMR 5.126, CFI 5.870. To test the
applicability of a curvilinear, rather than straight-
line, trajectory, we added a quadratic term to the
model, which improved the model fit substantially,
Dw2
ð4;N¼96Þ¼169:29, Po.001, and the model fit
the data well, w2
ð36;N¼96Þ¼196:51, RMSEA 5.056
(RMSEA CI: .050|.062), SRMR 5.061, CFI 5.955.
The cubic term was not significant nor did its
inclusion improve the model fit. On the basis of
these results, we concluded that the growth was
curvilinear.
The unstandardized intercept (i59.713,
SE 5.307, Po.001), the linear slope (s51.471,
SE 5.144, Po.001) and the quadratic acceleration
(q5.233, SE 5.017, Po.001) were all significant.
As shown in Figure 2, participants exhibited a
Fig. 1. The hypothetical structural equation model.
Fig. 2. The curvilinear growth model of whether the participants
enjoyed the violent scenes.
183Desensitization To Media Violence
Aggr. Behav.
quadratic growth over time, and they seemed to
report lower levels of enjoyment during the fourth
scene compared with the beginning and the end. In
addition, toward the end of the study participants
reported that they enjoyed the violent scenes more
compared with the beginning of the study. In
addition, the unstandardized residual intercept
(i53.825, SE 51.454, Po.01) was significant,
although the residual linear slope (s5.306,
SE 5.288, P5.29) and the quadratic acceleration
term (q5.005, SE 5.004, P5.23) were not sig-
nificant. These findings indicate that there was
significant variability between subjects in terms of
initial levels of movie violence enjoyment, but that
the change over time followed the same pattern for
all participants. Additional analysis investigated
whether aggression might influence the initial levels
of movie violence enjoyment. According to the
findings, aggression was positively related to initial
levels, b5.48, SE 5.13, Po.01. This finding sug-
gests that highly aggressive individuals enjoyed the
violent media more. This model also fit the data
well, w2
ð44;N¼96Þ¼203:16, RMSEA 5.064 (RMSEA
CI: .057|.071), SRMR 5.054, CFI 5.949.
Analysis Investigating Participants’ Responses
in Terms of Sympathy Toward the Victim
The linear growth model fit the data well,
w2
ð40;N¼96Þ¼475:59, RMSEA 5.057 (RMSEA CI:
.051|.063), SRMR 5.064, CFI 5.943. As with the
previous analysis, to test the applicability of a
curvilinear, rather than straight-line, trajectory of
change, we added a quadratic term to the model,
which improved the model fit, Dw2
ð4;N¼96Þ¼56:57,
Po.001, and the model fit the data well,
w2
ð36;N¼96Þ¼419:02, RMSEA 5.053 (RMSEA CI:
.048|.058), SRMR 5.046, CFI 5.962. The cubic
term was not significant nor did its inclusion
improve the model fit. On the basis of these results,
we concluded that the growth was curvilinear. The
unstandardized intercept (i58.395, SE 5.339,
Po.001), the linear slope (s51.159, SE 5.174,
Po.001) and the quadratic acceleration term
(q5.134, SE 5.019, Po.001) were all significant.
As shown in Figure 3, participants exhibited a
quadratic growth over time. As participants watched
more violent scenes, they started being more
sympathetic toward the victim, and midway through
the study (scene 5) they showed higher levels of
sympathy toward the victim. However, after the fifth
scene participants started reporting lower levels of
sympathy, and toward the end of the study they
reported lower levels of sympathy compared with
the beginning. In addition, the unstandardized
residual variability of intercept (i52.161,
SE 5.917, Po.01), the linear slope (s5.998,
SE 5.241, Po.001) and the quadratic acceleration
term (q5.015, SE 5.003, Po.001) were signifi-
cant. These findings indicate that there was sig-
nificant variability between subjects at the initial
levels of exposure to media violence and in terms of
change over time.
Additional analysis investigated whether aggres-
sion might influence the initial level of sympathy and
change over time. According to the findings,
aggression was significantly related to the intercept,
b5.18, SE 5.08, Po.05, but not to the linear
slope, b5.04, SE 5.05, P5.45, or the quadratic
term, b5.01, SE 5.01, P5.42. This finding
suggests that highly aggressive individuals expressed
less sympathy for the victim. The model fit the data
marginally well, w2
ð44;N¼96Þ¼424:07, RMSEA 5.088
(RMSEA CI: .076|.101), SRMR 5.062, CFI 5.922.
Analysis Investigating Participants’ Responses
to the Comedy Scenes
The linear growth model investigating desensitiza-
tion to comedy scenes fit the data well,
w2
ð40;N¼96Þ¼240:44, RMSEA 5.068 (RMSEA CI:
.061|.075), SRMR 5.058, CFI 5.913. The quadra-
tic term was significant (q5.077, SE 5.014,
Po.001), although its inclusion did not improve
the model fit, Dw2
ð40;N¼96Þ¼7:84, P5.09. On the
basis of these results, we concluded that the growth
was linear. The unstandardized intercept (i514.771,
SE 5.249, Po.001) and the linear slope (s5.194,
SE 5.040, Po.001) were significant. According to
Figure 4, the participants seemed to report decreas-
ing levels of enjoyment across time. In addition, the
unstandardized residual intercept (i54.172,
SE 5.899, Po.001) was significant, although the
residual linear slope was not (s5.029, SE 5.020,
Fig. 3. The curvilinear growth model of whether the participants
sympathized with the victim.
184 Fanti et al.
Aggr. Behav.
P5.14). These findings indicate that there was
significant variability at the initial levels of exposure
to comedy scenes. Additional analysis investigated
whether aggression might influence the initial level
of comedy enjoyment. According to the findings,
aggression was not significantly related to the
intercept, b5.02, SE 5.11, P5.82. This model
also fit the data well, w2
ð44;N¼96Þ¼210:99,
RMSEA 5.079 (RMSEA CI: .072|.086),
SRMR 5.078, CFI 5.947.
DISCUSSION
The findings of this study indicate that desensiti-
zation to media violence can occur after continuous
exposure to media violence in a short period of time,
and that desensitization to violent stimuli is better
represented by a curvilinear pattern of change.
Initial exposure to media violence produced some
aversive responses. During the first scenes partici-
pants reported that they enjoyed the violent scenes
less and reported more concern for the suffering of
victims. With repeated exposure, however, the
psychological impact of media violence was reduced
and participants indicated feeling less sympathy for
violence victims and started enjoying more the
violence portrayed in the media scenes. Therefore,
following the previous research, our findings suggest
that individuals tend to have negative responses
after initially observing violence, although after
repeated exposure they became desensitized to
media violence and reported that they actually
enjoyed the media violence more [Cantor, 1998;
Linz et al., 1989; Wolpe, 1982]. The implications of
these findings are important as it is possible that
reduced inhibitions toward violence might result in
increased aggressive behavior [Smith and Donner-
stein, 1998], which can have a number of negative
consequences for both the individual and the
society. Furthermore, the findings suggest that
desensitization might also result in decreased like-
lihood of being concerned for the victims of
violence, and therefore in decreased likelihood of
helping a victim of violence [Mullin and Linz, 1995].
This study also investigated whether there are
individual differences in short-term desensitization
effects. Aggressive behavior seemed to be related to
the initial level of enjoyment, indicating that
aggressive individuals enjoyed the violent scenes
more in comparison with the rest of the sample. In
terms of sympathy, the more aggressive participants
reported lower initial sympathy toward the victims
of violence. These findings are in agreement with
previous studies that linked aggressive behavior to
lower sympathy or empathy [Mehrabian, 1997;
Miller and Eisenberg, 1988]. Although we identified
differences in the degree of enjoyment and sympathy
toward the victim of violence, all individuals
regardless of aggressive habits were desensitized to
media violence in the same manner. This finding
seems to rule out the hypothesis that aggressive
individuals will exhibit a different desensitization to
violence pattern because individuals with high trait
aggression have permanently reduced inhibition to
aggression and blunted evaluative categorization of
violent stimuli.
A direction for future studies could be the
investigation of individual differences in desensitiza-
tion to media violence in other types of aggression as
well. For example, proactive aggressors view aggres-
sion as a positive behavior, and because of that they
have no negative emotions when acting aggressively
[e.g., Barratt et al., 1999]. However, reactive
aggressors are characterized in part by feelings of
remorse and by thought confusion following the
aggressive acts. The differences between the two
types of aggression may suggest that proactive
aggressors might enjoy media violence more and
reactive aggressors might sympathize with the
victims of violence more. Furthermore, future
research might be able to identify individual
differences when comparing individuals exhibiting
emotional problems such as depression and anxiety.
Importantly, the results of this study revealed a
different desensitization to comedy scenes, providing
evidence that desensitization may be different for
negative or aversive stimuli such as media violence.
Our results clearly show that desensitization to
positive emotions associated with watching the
comedic scenes was better represented by a linear
model, as opposed to a curvilinear model for violent
scenes. Therefore, with repeated exposure partici-
pants habituated to comedy scenes and did not
Fig. 4. The linear growth model of whether the participants enjoyed the
comedy scenes.
185Desensitization To Media Violence
Aggr. Behav.
exhibit the initial aversive response found for the
violent scenes, as comedy scenes are generally
related to positive emotions. Future studies should
investigate differences in how individuals process
negative aversive stimuli, such as violence, and how
they process more positive stimuli, such as comedy.
Moreover, although desensitization can have
maladaptive outcomes by resulting in lower respon-
siveness to real-world violence and by making the
viewing of violence more pleasurable [Carnagey
et al., 2007], the process of desensitization can also
be adaptive. As demonstrated by the previous
research, desensitization can be used as a form of
treatment for phobias and posttraumatic stress
disorder, and as a form of training for nurses,
medical doctors and even soldiers [Grossman and
DeGaetano, 1999; Pantalon and Motta, 1998;
Wolpe, 1982]. The current findings that desensitiza-
tion to aversive stimuli can occur in a very short
period may have important implications for these
types of treatment. However, future studies should
first examine whether short-term desensitization to
violence may generalize to other negative emotions.
The diverse sample of college students, the use of
both violent and comedy scenes and the use of video
clips depicting different types of violence, such as
physical, domestic, gang and weapon, are all
strengths of this investigation. In addition, the nine
time points assessed allowed for the investigation of
nonlinear trajectories of change over time [Singer
and Willett, 2003]. However, data were based solely
on self-report, and the investigation of physiological
desensitization would have enhanced the study’s
findings. Nevertheless, previous studies provided
sufficient evidence that exposure to violent media
physiologically desensitizes individuals to violence
[e.g., Bartholow et al., 2006; Carnagey et al., 2007].
Further, a more fine-grained assessment of different
forms of aggression (e.g., reactive vs. proactive and
relational vs. overt) and information on previous
levels of media and real-life violence exposure might
further elucidate the meaning of our findings.
In summary, this study investigated desensitiza-
tion to repeated exposures of violent media over a
short period of time. The findings indicated that
exposure to media violence even for a short period
of time may influence individuals’ attitudes toward
the victims of violence and the enjoyment of violent
content. Exposure to violence and the process of
desensitization may result in pro-violence attitudes,
which based on the previous research have the
power to make real-world violence more acceptable
or normative [Bushman and Huesmann, 2001;
Huesmann and Kirwil, 2007; Linz et al., 1984;
Mullin and Linz, 1995]. Moreover, a person
desensitized to violence may not be concerned,
empathetic or sympathetic toward the victims of
violence, and will be less likely to help the victims of
violence [Mullin and Linz, 1995]. Therefore, the
problem for society is the behavior resulting from
the process of desensitization because individuals
may become so desensitized to violence that they
come to accept that violence is normative. As this
study suggests, the process of desensitization might
occur in a short period of time, which raises the
possibility that maladaptive behaviors such as
violence are more likely and that adaptive behaviors
such as helping a victim of violence are less likely to
occur following the viewing of violent media.
REFERENCES
Anderson CA, Berkowitz L, Donnerstein E, Huesmann LR, Johnson J,
Linz D, Malamuth N, Wartella E. 2003. The influence of media
violence on youth. Psychol Sci Public Interest 4:81–110.
Anderson CA, Benjamin AJ, Wood PK, Bonacci AM. 2006.
Development and testing of the Velicer attitudes toward
violence scale: Evidence for a four-factor model. Aggr Behav
32:122–136.
Barratt ES. 1991. Measuring and predicting aggression within the
context of a personality theory. J Neuropsychiatry 3:35–39.
Barratt ES, Stanford MS, Dowdy L, Liebman MJ, Kent TA. 1999.
Impulsive and premeditated aggression: A factor analysis of self
reported acts. Psychiatry Res 86:163–173.
Bartholow BD, Bushman BJ, Sestir MA. 2006. Chronic violent video
game exposure and desensitization to violence: Behavioral and
event-related brain potential data. J Exp Soc Psychol 42:532–539.
Berkowitz L. 1984. Some effects of thoughts on anti- and prosocial
influences of media events: A cognitive-neoassociation analysis.
Psychol Bull 95:410–427.
Bushman BJ. 1998. Priming effects of media violence on the
accessibility of aggressive constructs in memory. Pers Soc Psychol
Bull 24:537–545.
Bushman BJ, Huesmann LR. 2001. Effects of televised violence
on aggression. In: Singer DG, Singer JL (eds.). Handbook
of Children and the Media. Thousand Oaks, CA: Sage,
pp 223–254.
Cacioppo JT, Crites Jr. SL, Gardner WL, Berntson GG. 1994.
Bioelectrical echoes from evaluative categorizations: I. A late
positive brain potential that varies as a function of trait negativity
and extremity. J Pers Soc Psychol 67:115–125.
Cantor J. 1998. Mommy, I’m Scared: How TV and Movies Frighten
Children and What We Can Do to Protect Them. San Diego, CA:
Harvest/Harcourt.
Carnagey NL, Anderson CA, Bushman BJ. 2007. The effect of video
game violence on physiological desensitization to real-life
violence. J Exp Soc Psychol 43:489–496.
Cline VB, Croft RG, Courrier S. 1973. Desensitization of children to
television violence. J Pers Soc Psychol 27:360–365.
Eron LD. 2001. Seeing is believing: How viewing violence alters
attitudes and aggressive behavior. In: Bohart AC, Stipek DJ
(eds.). Constructive and Destructive Behavior: Implications for
Family, School, and Society. Washington, DC: American
Psychological Association, pp 49–60.
186 Fanti et al.
Aggr. Behav.
Flowers AL, Hastings TL, Kelley ML. 2000. Development of a
screening instrument for exposure to violence in children: The
KID-SAVE. J Psychopathol Behav Assess 22:91–104.
Funk JB, Bechtoldt-Baldacci H, Pasold T, Baumgartner J. 2004.
Violence exposure in real-life, video games, television, movies,
and the internet: Is there desensitization? J Adolesc 27:23–39.
Gerbner G, Gross L, Morgan M, Signorielli N. 1994. Growing up with
television: The cultivation perspective. In: Bryant J, Zillman D (eds.).
Media Effects: Advances in Theory and Research. Hillsdale, NJ:
Lawrence Erlbaum Associates, pp 17–41.
Grossman D, Degaetano G. 1999. Stop teaching our kids to kill: A
call to action against TV, movie and video game violence. New
York: Crown Publishers.
Gunter B. 1994. The question of media violence. In: Bryant J, Zillman
D (eds.). Media Effects: Advances in Theory and Research.
Hillsdale, NJ: Lawrence Erlbaum Associates, pp 163–211.
Hedeker D, Gibbons RD. 1994. A random-effects ordinal regression
model for multilevel analysis. Biometrics 50:933–944.
Huesmann LR, Kirwil L. 2007. Why observing violence increases the
risk of violent behavior in the observer. In: Flannery DJ,
Vazsonyi AT, Waldman ID (eds.). The Cambridge Handbook
of Violent Behavior and Aggression. Cambridge, UK: Cambridge
University Press, pp 545–570.
Huesmann LR, Moise J, Podolski CP. 1997. The effects of media
violence on the development of antisocial behavior. In: Stoff D,
Breiling J, Masser J (eds.). Handbook of Antisocial Behavior.
New York: Wiley, pp 181–193.
Huesmann LR, Moise-Titus J, Podolski CP, Eron LD. 2003.
Longitudinal relations between childhood exposure to media
violence and adult aggression and violence: 1977–1992. Dev
Psychol 39:201–221.
Hu L, Bentler PM. 1998. Fit indices in covariance structure
modeling: Sensitivity to underparametrized model misspecifica-
tion. Psychol Methods 3:424–453.
Kline RB. 1998. Principles and practice of structural equation
modeling. New York: The Guilford Press.
Lazarus RS, Speisman M, Mordkov AM, Davison LA. 1962.
A laboratory study of psychological stress produced by a motion
picture film. Psychol Monogr: Gen Appl 34:553.
Linz D, Donnerstein E, Penrod S. 1984. The effects of long-term
exposure to filmed violence against women. J Commun
34:130–147.
Linz D, Donnerstein E, Penrod S. 1988. Effects of long-term
exposure to violent and sexually degrading depictions of women.
J Pers Soc Psychol 55:758–768.
Linz D, Donnerstein E, Adams SM. 1989. Physiological desensitiza-
tion and judgments about female victims of violence. Hum
Commun Res 15:509–522.
Malamuth NM, Check JVP. 1981. The effects of mass media
exposure on acceptance of violence against women: A field
experiment. J Res Pers 14:436–446.
Mehrabian A. 1997. Relations among personality scales of aggres-
sion, violence, and empathy: Validational evidence bearing on the
risk of eruptive violence scale. Aggr Behav 23:433–445.
Miller P, Eisenberg N. 1988. The relation of empathy to aggressive
and externalizing/antisocial behavior. Psychol Bull 103:
324–344.
Mullin CR, Linz D. 1995. Desensitization and resensitization to
violence against women: Effects of exposure to sexually violent
films on judgments of domestic violence victims. J Pers Soc
Psychol 69:449–459.
Muthe
´n LK, Muthe
´n B. 2006. Mplus User’s Guide. Los Angeles,
CA: Muthe
´n & Muthe
´n.
Paik H, Comstock G. 1994. The effects of television violence on anti-
social behavior: A meta-analysis. Commun Res 21:516–546.
Pantalon MV, Motta RW. 1998. Effectiveness of anxiety manage-
ment training in the treatment of posttraumatic stress disorder: A
preliminary report. J Behav Ther Exp Psy 29:21–29.
Pennell AE, Browne KD. 1999. Film violence and young offenders.
Aggr Violent Behav 4:13–28.
Singer JD, Willett JB. 2003. Applied longitudinal data analysis:
Modeling change and event occurrence. Oxford, NY: Oxford
University Press.
Smith SL, Donnerstein E. 1998. Harmful effects of exposure to
media violence: Learning of aggression, emotional desensitiza-
tion, and fear. In: Geen RG, Donnerstein E (eds.). Human
Aggression: Theories, Research, and Implications for Social
Policy. San Diego, CA: Academic Press, pp 167–202.
Steinfeld J. 1972. Statement in Hearings Before Subcommittee on
Communications of Committee on Commerce (U.S. Senate,
Serial No. 92-52). Washington, DC: U.S. Government Printing
Office, pp 25–27.
Thomas MH, Horton RW, Lippincott EC, Drabman RS. 1977.
Desensitization to portrayals of real life aggression as a function
of television violence. J Pers Soc Psychol 35:450–458.
Wei R. 2007. Effects of playing violent videogames on Chinese
adolescents’ pro-violence attitudes, attitudes toward others, and
aggressive behavior. Cyberpsychol Behav 10:371–380.
Wolpe J. 1982. The Practice of Behavior Therapy, 3rd edition. New
York: Pergamon Press.
187Desensitization To Media Violence
Aggr. Behav.
