Content-Type: text/html

Hypermasculinity, Aggression, and Television Violence: An Experiment

By
Erica Scharrer
Assistant Professor
Department of Communication
309 Machmer Hall
University of Massachusetts Amherst
Amherst, MA 01003
Phone: (413) 545-4765
Fax: (413) 545-6399
E-mail: <a href="/cgi-bin/wa?LOGON=A3%3Dind0309D%26L%3DAEJMC%26E%3D8bit%26P%3D4507450%26B%3D--%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>


Submitted to:
Mass Communication & Society Division
AEJMC
April 1, 2003


This study was supported by a Faculty Research Grant from the University of
Massachusetts Amherst. The author would like to thank Nina Huntemann for
her assistance with the project.






Hypermasculinity, Aggression, and Television Violence: An Experiment


Abstract


This experiment tests the role of hypermasculinity (HM) and trait
aggression in predicting aggressive responses to violent television. 91
male college students were exposed to a violent and HM television program,
a violence only program, or a control program. Results find that some
dimensions of HM and pre-existing aggression interact with exposure to the
treatment stimulus to predict aggressive responses, and that HM can also be
treated as a dependent variable that is affected by television exposure.









Hypermasculinity, Aggression, and Television Violence
In the United States, young males are both the most frequent perpetrators
and the most frequent victims of crimes (Maguire & Pastore, 1995; U.S.
Department of Health and Human Services, 2001). Public debate surrounding
the influence of media violence on antisocial behavior has reached a
fevered pitch in recent years in the wake of a spate of school shootings
involving young male perpetrators (cite). In social science research, the
association between exposure to television violence and aggression has been
consistently supported (see Comstock & Scharrer, 1999; Potter, 1999).
However, calls for additional research have been issued to examine the
conditions, including the individual differences that audience members
bring to the viewing experience (Oliver, 2002), that are most likely to
produce aggressive responses to violent media (Jo & Berkowitz, 1994;
Potter, 1999).
This research project examines a gender-related variable, hypermasculinity,
and measures its role in young males' aggressive responses to violent
television stimuli. It positions hypermasculinity not only as a
pre-existing condition that qualifies as an individual difference that
potentially shapes responses to television exposure, but it also
investigates whether exposure to television programming that contains
hypermasculine depictions can also contribute to self-reports of one's own
hypermasculinity. It also examines the role of pre-existing aggressive
tendencies in predicting aggressive responses to violent television.
Finally, it examines the impact of two different types of violent
portrayals—a somewhat realistic drama in which characters who use violence
also meet the definition of hypermasculinity, and a fantasy-based drama in
which violence occurs without accompanying hypermasculine gender
portrayals. The objective,

Hypermasculinity, Aggression, and Television Violence
therefore, is to advance understanding in an area of the utmost
significance: Is hypermasculinity a factor in one's psychological make-up
that predicts an aggressive response to viewing a particular type of
television violence? Can hypermasculinity itself vary in response with
television messages that make it seem normative? The importance of the
answers to those questions for a society with high frequencies of
male-perpetrated violence is undeniable.
Literature Review
Hypermasculinity
Hypermasculinity (HM) is a personality construct that occurs primarily in
males, in which stereotypically "macho" traits are held up as an ideal.
Hypermasculine males exhibit strong and exaggerated forms of masculinity,
virility, and physicality, as well as a tendency toward disrespect for
women and the trivializing of romantic and sexual relationships (Zaitchik &
Mosher, 1993). HM is also characterized by high sensation seeking and
pursuit of thrills associated with dangerous situations as well as disdain
for emotions viewed as weak or feminine, such as compassion or the
expression of sadness (Mosher & Sirkin, 1984). Mosher and Sirkin (1984)
measure this concept using three components: (1) Callused Sex Attitudes,
(2) Danger as Exciting, and (3) Violence as Manly. Later, Hall (1992) added
an element known as Toughness as Self-Control. Thus, HM is seen in
tendencies to view physical aggression as a typical and, indeed,
inevitable, response for men and to exercise a sense of emotional
callousness or stoicism.
Mosher and Tompkins (1988) have suggested that young males and those from
lower socioeconomic classes are most likely to adopt the macho personality
constellation,
Hypermasculinity, Aggression, and Television Violence
and they point to cultural norms and socialization as contributing factors.
Mosher and Sirkin (1984) suggest that HM begins in childhood, during which
parents and other important individuals in some boys' lives negatively
reinforce feelings of fear and distress using contempt and humiliation. As
the child grows, other aspects of society reinforce the male ideal,
including norms at school or, as is tested here, in the media. Therefore,
though the concept is referred to as a "personality construct" which
presupposes a certain degree of stability, there is some reason to believe
that HM can be viewed as subject to change as well, in light of one's
perception of norms and in association with other social factors.
Although it has been suggested that socialization is primarily responsible
for the development of HM (Broude, 1990; Collison, 1996; Mosher & Sirkin,
1984), no empirical study exists that has examined whether television—one
of the primary socializing agents of contemporary times (Singer & Singer,
2001)—is a causal contributor to HM. This study adds to the body of
literature by testing whether exposure to television programming can
contribute to HM self-reports. Because HM males are rather prevalent in
television programming, especially in police and detective dramas (XXXX,
2001b), and films (Sparks, 1996), television exposure may well contribute
to the development of HM. Furthermore, Zaitchik and Mosher (1993) have
determined that an individual is more likely to exhibit HM behaviors and
attitudes when in the company of other HM men. It is possible that through
identification with HM characters, male audience members may feel as if
they are in the company of other HM men and may disinhibit HM attitudes and
behavioral self-reports. Therefore, the study at hand
Hypermasculinity, Aggression, and Television Violence
examines not only the possibility that HM is a pre-existing condition that
can shape responses to television exposure (XXXX, 2001a), but also that
television exposure itself can contribute to HM attitudes and identity.
Hypermasculinity and Media Exposure
Previous research has found associations between HM and general aggression
(Gold, Fultz, Burke, Prisco, & Willett, 1992) as well as sexual aggression,
the latter including such variables as history of sexual aggression (Weir &
Branscombe, 1990), self-reported likelihood of rape (Smeaton & Byrne,
1987), and acceptance of rape myths (Sullivan & Mosher, 1990). A smaller
number of past studies have included a media element in the research
design, not framing the research as pertaining to media effects, per se,
but using a videotaped stimulus, nonetheless. For example, Beaver, Gold,
and Prisco (1992) performed an experiment in which over 175 undergraduate
males, divided according to low, moderate, and high HM, were exposed to a
videotaped date rape scenario. High HM males rated the male perpetrator as
less responsible, were less sympathetic to the victim, and perceived a
lesser degree of force compared to those with lower levels of HM. Russell
(1992) examined the responses of 110 college-aged males to a videotaped
sports stimulus, either a violent boxing match (treatment) or a non-violent
skiing event (control). Results revealed higher post-test aggression among
those with high pre-existing HM who watched the boxing match compared to
both those who watched the skiing event and those lower in HM.
However, only one previous study (XXXX, 2001a) uses a media effects
framework to examine the connection between HM and aggressive responses to
violent
Hypermasculinity, Aggression, and Television Violence
television. XXXX performed an experiment with 60 college-aged males,
exposing the treatment group to an episode of Miami Vice that contained
violence and HM and exposing the control group to an episode of Home
Improvement that contained HM but no violence. Subjects' HM was measured
using 20 items from the HM index (Mosher & Sirkin, 1984) and aggression was
measured using 28 items from the Buss-Durkee Hostility Inventory (Buss &
Durkee, 1957). Results showed that those treatment-group subjects high in
pre-existing HM had a larger increase in aggression between the pre- and
post-test compared to those treatment-group subjects low in pre-existing HM.
However, the study had a number of important flaws that will be rectified
in the current research. First, the use of a sitcom as the control group
condition may have decreased control subjects' aggression, so that the
difference in post-test aggression between them and the treatment group was
exaggerated. Second, in the previous study, HM was measured solely as an
independent variable. The current study will examine the possibility that
not only can one's HM (as an independent variable) influence how one
responds to media violence, but exposure to media violence could also
influence one's HM (as a dependent variable). Thus, the current study will
measure HM at the pre- and post-test and will also seek to replicate the
result that HM is associated with more aggressive responses to violent and
HM television.
Theoretical Foundations
Berkowitz' (1984, 1990, 1993) neo-association theory suggests that exposure
to a mediated portrayal of violence can serve to disinhibit viewers'
aggression, through a series of cognitive links between their past, stored
experiences and information presented
Hypermasculinity, Aggression, and Television Violence
in the portrayal. The central notion of the theory is that when an
individual encounters new information, such as via exposure to a television
portrayal, stored mental images that pertain to that new information are
called forth through semantic links. Exposure to violence on television,
then, conjures stored information about violence in one's mind, which can
then color short-term emotions and behaviors. Furthermore, past research
has determined that individuals with higher pre-existing aggression have
more extensive experience and stored thoughts associated with aggression
that are available to call forth after violent media exposure compared to
individuals lower in pre-existing aggression (Bushman, 1995, 1996).
Also according to the theory, not all television or media portrayals of
violence are equally likely to produce an aggressive response. Rather,
factors in the depiction of violence can result in either the inhibition or
the disinhibition of an aggressive response by the viewer. One such factor
is the degree of identification between the audience member and the
character that uses violence, which is, in turn, often associated with a
perceived similarity between viewer and character (Berkowitz & Rawlings,
1963; Turner & Berkowitz, 1972). Jo and Berkowitz (1994) have reasoned that
a viewer who is strongly identifying with aggressive characters while
viewing television will have more accessible aggressive thoughts in mind
immediately after viewing because they have called them to the forefront
while viewing. Another such factor is the degree of realism in the
television depiction (Berkowitz & Alioto, 1973; Atkin, 1983). Jo and
Berkowitz (1994) argue that identification and involvement are facilitated
by realism. Furthermore, the theory would suggests that cognitive
associations made between media information
Hypermasculinity, Aggression, and Television Violence
and one's own behavior in specific situations are stronger when the media
information pertains to one's own situation, such as when it is reality-
rather than fantasy-based.
This study is also informed by an extension of neo-associationism, the
theory that "mental models" play a significant role in individuals'
responses to media (Roskos-Ewoldsen, Roskos-Ewoldsen, & Dillman Carpentier,
2002). The mental models theory suggests that media exposure can make
certain attitudes accessible by bringing them to the fore, which, in turn,
contributes to the use of a "mental model" to determine how to define a
current situation. In other words, the media exposure is "the prime (that)
influences how later information is interpreted by influencing the type of
mental model that is constructed to understand the situation"
(Roskos-Ewoldsen, Roskos-Ewoldsen, & Dillman Carpentier, 2002, p.
112).  Factors that influence which type of mental model will be
constructed include the similarity between the media prime and the
situation, as well as the individual differences of the individual that
lead to the accessibility of particular mental models. Exposure to violent
media, therefore, would create a scenario in which, when presented with a
new situation, treatment group subjects—especially those with individual
differences that make them more prone to aggression—would have greater
access to an aggressive mental model of the new situation compared to
control group subjects.
The study at hand suggests that, in addition to viewers high in trait
aggressiveness (Bushman, 1995, 1996), viewers high in HM will also perceive
a link between themselves and the HM characters they see in one of the
treatment conditions who use violence in successful and rewarded ways.
Thus, according to an application of mental
Hypermasculinity, Aggression, and Television Violence
models theory, those viewers will be more likely than lower HM viewers to
respond aggressively to the television stimulus because they are more
likely to have ready access to aggressive mental models to apply to new
situations. It is argued that HM viewers are more likely to have a greater
number of stored experiences with violence that they view positively and
therefore will make stronger cognitive connections with a violent and HM
stimulus than less HM viewers. Central to the hypotheses tested in this
research, therefore, is the notion of identification and perceived
similarity between audience member and television character that has been
consistently linked with media effects in past research (Comstock &
Scharrer, 1999).
Hypotheses
H1: Those with higher pre-existing aggressive tendencies will be more
likely to experience an increase in aggression following exposure to media
violence compared to those with lower pre-existing aggressive tendencies.
        Aggression has been determined to be a rather stable personality trait,
manifested dynamically in response to changing factors in the social
environment, but also somewhat unchanging as a predisposition (Huesmann,
Eron, Lefkowitz, & Walder, 1984). Bushman (1996) has identified trait
aggressiveness as a key factor in predicting aggressive responses to
violent media stimuli. Therefore, the prior experiences with and tendencies
toward aggression that a viewer brings to the screen as a personality trait
are predicted to shape the amount of situational aggression in responses to
the two violent television program stimuli (regardless of whether the
stimulus contains HM or not). High aggressive individuals are more likely
to apply aggressive mental models to new
Hypermasculinity, Aggression, and Television Violence
situations, especially if they perceive a link between themselves and the
television content to which they are exposed (Berkowitz, 1984;
Roskos-Ewoldsen, Roskos-Ewoldsen, & Dillman Carpentier, 2002).
H2: Those with higher pre-existing hypermasculinity will be more likely to
experience an increase in aggression following exposure to the stimulus
that combines violent and hypermasculine portrayals compared to those with
lower pre-existing hypermasculinity.
        In the author's previous study (XXXX, 2001a), it was determined that males
with high per-existing HM had greater increases in self-reported aggression
after watching a television stimulus with high levels of violence and HM
compared to males with low pre-existing HM. However, by using a stimulus
program that contained both violence and HM, that study neglected to
determine whether HM subjects were responding to the violence in the
stimulus or to the violence and the HM portrayals. The current study
attempts to isolate the effects of violent portrayals from those of
portrayals that contain both violence and HM by using two different
treatment stimuli. An application of neo-associationism (Berkowitz, 1984)
and the mental models theory (Roskos-Ewoldsen, Roskos-Ewoldsen, & Dillman
Carpentier, 2002) suggests that HM may be an individual difference that
results in greater accessibility of aggressive frames to use to apply to
new situations. This link between HM and the readiness of aggressive mental
models is made here due to the centrality of a physically aggressive
orientation in the very definition of HM (Mosher & Sirkin, 1984).
Furthermore, HM and aggression have been consistently linked in past
studies (Beaver, Gold, & Prisco, 1992; Gold, et al., 1992; Russell, 1992;
Hypermasculinity, Aggression, and Television Violence
Smeaton & Byrne, 1987; Sullivan & Mosher, 1990; Weir & Branscombe, 1990).
High HM viewers will also presumably identify and perceive greater
similarity with characters that also display HM characteristics compared to
less HM viewers (Berkowitz & Rawlings, 1963), which can contribute to
stronger media effects.
H3: Those exposed to the more realistic television violence stimulus will
be more likely to experience as increase in aggression than those exposed
to the less realistic, fantasy-based television violence stimulus.
The third hypothesis predicts that, regardless of the subject's previously
existing HM level, those exposed to a more realistic television stimulus
will have a larger increase in aggression compared to those exposed to a
fantasy-based television stimulus. Neo-associationism suggests that the
degree of realism in the television depiction (Berkowitz & Alioto, 1973;
Atkin, 1983) impacts the extent to which viewers will be influenced by
television exposure, in part by encouraging identification and involvement
(Jo & Berkowitz, 1994). Realistic television exposure would also presumably
strengthen the cognitive associations made between media information and
one's own decisions about behavior in specific situations.
H4: Those exposed to either the violent and HM television stimulus or the
violent television stimulus will be more likely to experience an increase
in aggression compared to those in the control group.
This hypothesis suggests that exposure to either stimulus containing
violence will lead to an increase in aggression whereas exposure to the
control television program stimulus that does not contain violence will
not. It is based on the decades of research
Hypermasculinity, Aggression, and Television Violence
that support an association between violent television exposure and
aggression (see Comstock & Scharrer, 1999 for summary). Therefore, when
combined with H3, a progression from no increase (for the control group) to
a small increase (for the fantasy-based violence only group), to a larger
increase (for the realistic violence and HM group) from pre- to
post-aggression measures is predicted.
RQ: Will exposure to the stimulus containing violence and HM lead to
increased HM levels?
        A review of the literature suggests that HM has rarely (if ever) been
tested as a dependent measure. However, it has been argued that HM
attitudes and tendencies, like other gender roles, are formed in large part
by socializing factors, of which television is one (in addition to parents,
peers, etc.) (Broude, 1990; Collison, 1996; Mosher & Sirkin, 1984; Singer &
Singer, 2001). Indeed, studies have shown that HM characters have a
considerable presence in television programming (Sparks, 1996, XXXX,
2001b), thereby providing multiple models for potential socialization.
However, since this issue has not been adequately addressed in the previous
literature, a research question rather than a hypothesis is posed here. In
this study, then, HM is measured at both the pre- and post-test times in
order to determine whether exposure to a television program that shows HM
as normative will increase subjects' self-reports of HM tendencies.
Methods
Research Design and Procedure
This experiment tests the impact of exposure to two types of television
violence portrayals on subjects' aggression levels, and determines the role
of subjects' HM as an
Hypermasculinity, Aggression, and Television Violence
antecedent personality variable as well as a potential dependent variable.
The study employs a 3 (control vs. violent, fantasy-based, non-HM stimulus
vs. violent, reality-based HM stimulus) X 2 (low HM, high HM) X 2 (low
trait aggression, high trait aggression) repeated measures design.
The pre-test questionnaire, measuring HM and aggression as well as
demographics, was given to each group (control, and both treatment
conditions) at Time 1. One week later, each group was exposed to a
television program (either treatment or control) and then immediately
completed the post-test questionnaire, measuring their HM and aggression
again. A questionnaire numbering system was used to link the pre- and
post-tests of specific individuals without using their names.
Participants
93 subjects participated in the study and were randomly assigned to groups
using a table of random numbers. Subjects were males enrolled in
undergraduate or graduate study at a large, public university in New
England. Male college students were chosen as appropriate subjects because
hypermasculinity is typically viewed as a primarily male phenomenon and
because young males are statistically the most frequent perpetrators of
violence. The subjects were given $25 for their participation in the study
and were recruited using an ad in the campus newspaper and flyers posted
throughout the area.
Materials
The television program used for the violence and HM condition was
determined by asking a panel of colleagues to list the television drama(s)
that they felt contained the most violence and had the strongest depictions
of HM, selecting from current programs
Hypermasculinity, Aggression, and Television Violence
available on broadcast or cable. The panel consisted of graduate students
and faculty members studying communication, and the poll resulted in The
Sopranos receiving the highest number of votes. A similar process was used
to arrive upon the use of Seventh Heaven as a control condition. The panel
was asked what one-hour television drama currently available through
broadcast or cable contained the least amount of violence or the least HM
male characters. Seventh Heaven, a drama that airs on the WB network about
a family in which the father is a minister, received the highest number of
votes.
The use of an episode of Buffy the Vampire Slayer for the violence without
HM condition was based on the researcher's own knowledge of the program and
its depictions of violence and gender. It is a program in which the main
character is a female who frequently uses physical violence in a classic
good vs. evil theme. Not only is the character who uses violence female
rather than male (a relatively rare television phenomenon), but most of the
male characters in the program do not meet the criteria for HM. The main
male characters, for the most part, are not callous about sex, do not seek
danger for thrill, and do not see violence as a necessary extension of
their gender. Finally, the program was also chosen because it is
fantasy-based and therefore is an appropriate test of the fantasy versus
reality hypothesis.
Thus, each of the television programs used in the study was a one-hour
drama. The episode was shown to subjects in its entirety, including
commercial breaks, in order to approximate the manner in which viewers
would encounter the program in a natural viewing environment. Subjects
watched the programs in groups rather than individually, in an attempt to
make them feel less scrutinized, and also to approximate natural viewing.
Hypermasculinity, Aggression, and Television Violence
After subjects had completed the post-test questionnaire, a manipulation
check was utilized to determine whether the choice of television programs
to use in the study was appropriate in terms of participants' views of the
degree of violence and HM contained in each. Results of the manipulation
check show that participants did, indeed, find that the control television
program (Seventh Heaven) contained minimal amounts of HM and violence
(rated M = 1.09 on a scale of 1 to 5 for violence, and M = 2.18 on a scale
of 1 to 5 for "macho portrayals"). Participants also gave the Buffy the
Vampire Slayer episode fairly high marks for the presence of violence (M =
3.33) but low marks for HM (M = 2.50). Finally, participants viewed The
Sopranos episode that they had seen as violent (M = 3.64) and as containing
high levels of HM (M = 4.06). Therefore, the researcher's and the panel's
view of the programs' depictions of violence and HM—that led to the
decision to use the programs in the study— corresponded well to
participants' views of the programs as reported in the manipulation check.
Measures
Demographic data was gathered in the pre-test questionnaire, including the
age, year in school, and race of the subject, as well as the subject's
family income, and the education level of his father and mother. The
pre-test questionnaire also measured television exposure, asking subjects
to estimate the number of hours that they watch television on a typical
weekday and a typical weekend day, as well as the average number of days
per week that they watch. Subjects were also asked to list their three
favorite current television programs that they watch most frequently. The
ratings for each program listed by the subjects were consulted in TV Guide
to determine whether they
Hypermasculinity, Aggression, and Television Violence
typically contain violence. Three consecutive issues of TV Guide were
consulted. If the program was rated a V (for violence) more than once in
the three issues, it was given a score of 2 for amount of violence. If it
was rated a V just once, it was given a 1, and if it was never rated a V,
it was given a 0. If a program listed by a subject was not included in the
TV Guide, the official website of the program was consulted to determine if
violence appeared to be a major or minor theme. The overall television
exposure and amount of violence in favorite programs variables were then
used as controls in data analysis.
HM was measured using 15 of the 30 forced-choice self-report scale items
developed by Mosher and Sirkin (1984), an index that had been tested for
validity and reliability in past research. In the forced-choice items,
subjects are asked to choose one of two opposing statements that comes
closest to describing themselves. Equal numbers of items from the HM index
were chosen from each of the three components, callous sex attitudes, being
drawn to danger for thrill, and perceiving violence as "manly" (see
Appendix). Numerous false purpose forced-choice items were included in this
section of the questionnaire as well, interspersed with the HM items, in
order to reduce sensitivity to the research design. The HM items, as well
as the false purpose forced-choice items, were present in both the pre- and
post-test questionnaires so that possible changes in HM levels reported
after viewing the television stimulus could be determined.
Trait aggression was measured using selected items from the Buss and Perry
(1992) aggression / hostility (AH) scale, which also had been tested for
validity and reliability in the past. Equal numbers of items were selected
from the components of aggression and hostility identified by the authors,
physical aggression, verbal aggression,
Hypermasculinity, Aggression, and Television Violence
anger, and hostility (see Appendix). False purpose items that also asked
subjects to describe their attitudes and behavioral tendencies were
interspersed with the AH items in the questionnaire. For both the AH and
the false purpose items in this section, students were presented with a
statement about an attitude or a behavioral tendency and were asked to rate
it from 1 (that is not typical of me at all) to 5 (that is very typical of
me).
Situational or state aggression was measured by asking subjects to select
from a variety of possible behavioral responses after reading a series of
constructed, hypothetical scenario descriptions written by the researcher
and a research assistant. All of the scenarios raised frustrating
situations that could realistically happen in a young man's life. They
included being insulted by a professor, being stuck behind a slow-moving
car when running late for an appointment, being "cut" in line resulting in
a concert being sold out before obtaining tickets, coming across friends
who are vandalizing a car that bears a rival sports team's bumper sticker,
and encountering an overly aggressive opponent in a pick-up football game.
For each of the five scenarios, respondents were asked what they would be
most likely to do in the situation, and then were asked "if there were no
consequences and you knew you would get away with anything, what would you
like to do" in the same situation. For both the "what would you do" and
"what would you like to do" questions, question-specific response options
were given that varied in nature from strongly aggressive, to moderately
aggressive, to non-aggressive options. Examples of strongly aggressive
responses include writing on an evaluation that the professor should be
fired, starting a fight with the line cutters or the football aggressor,
joining in on the vandalizing of the car, and honking repeatedly and
swearing at the slow driver. The
Hypermasculinity, Aggression, and Television Violence
scenarios and related response items were posed to subjects at both the
pre-test and post-test, so that possible changes in situational
aggressiveness after viewing the television stimulus could be determined.
Asking subjects to state how they would respond to hypothetical situations
is a measure that has been used in past research (Cahoon & Edmonds, 1984,
1985). It is also an appropriate test of the neo-associationism-based
mental models theory (Roskos-Ewoldsen, Roskos-Ewoldsen, & Dillman
Carpentier, 2002) that suggests that when confronting a new situation, some
individuals may employ aggressive mental models to define that situation,
whereas others may not. The measures were designed to test the extent to
which subjects define the new situation presented in the scenario with an
aggressive mental model that they have called to mind while watching the
treatment television stimuli.
Finally, additional attempts were made to mask the purpose of the study. In
the post-test questionnaire, an entire section of false purpose items was
presented. These questions asked about subjects' opinions of the program
they had viewed, how positively or negatively the program presented men and
women, and their beliefs about the overall degree of realism in television
content. Although it was possible that the treatment group subjects figured
that the study was about television and aggression, they most likely would
not have been aware of the HM element of the study and would not have known
about the differences in the different viewing conditions.


Hypermasculinity, Aggression, and Television Violence
Results
Descriptive Statistics
        The mean age of the college males participating in the study was 20.47
(SD= 2.67), and age ranged from 18 to 39. One-third of the sample (33%) was
enrolled in the senior year of college, about one-fourth were freshmen
(24.2%), about one-fourth sophomores (26.4%), nearly 10% juniors, and 6.6%
graduate students. In an open-ended question asking about race and
ethnicity, just under three-fourths of the sample (72.5%) defined
themselves as white or Caucasian. Eleven percent described themselves as
Asian or Asian American, 4.4% as Italian American, 3.3% as Black or African
American, 2.2% as multiracial, and 2.2% did not answer. Other ethnic
descriptions were used by one subject each, including Dominican-Italian,
Cambodian, Indian, and Irish Cuban.
        The subjects reported a wide distribution of annual family incomes, with
18.7% falling in the category $75,001-90,000 and 16.5% each in
$50,001-75,000 and $90,001-105,000. Fifteen percent of the sample reported
a family income of $35,001-50,000, 14.3% more than $105,000, 21.1%
$20,001-35,000, and 4.4% less than $20,000. Education levels achieved by
the subjects' parents were also widely distributed, with the most common
responses for one's father bachelor's degree (38.5%) and master's degree
(23.1%), and the most common responses for one's mother the same
(bachelor's degree 29.7%, master's degree 24.2%). The average number of
hours on weekdays the sample viewed television was 4.77 (SD = 5.52), and on
weekends was 4.40 (SD = 2.97). The sample watched television an average of
4.4 days per week (SD = 2.18). Finally, although equal numbers of subjects
were originally assigned to each condition, some failed to
Hypermasculinity, Aggression, and Television Violence
appear for the second stage of the study. Ultimately, 38.5% (N = 35) of the
sample provided valid data (pre- and post-questionnaires completed) for the
control group, 35.2% for the reality-based violence and HM group (N = 32),
and 26.4% for the fantasy-based violence without HM group (N = 24).
Factor Structure and Data Reduction
        Because the HM and the AH items had previously identified dimensions,
confirmatory factor analysis was run on the items comprising each of the
dimensions of HM and AH to test whether in the present data they would,
indeed, form only one factor. A principal component analysis, requesting
eigenvalues over 1, was run for each of the dimensions of HM and AH, at
both the pre- and post-test. Results were largely supportive of the factor
structure as defined by previous authors (for HM, Mosher & Sirkin, 1984,
for AH, Buss & Perry, 1992).
        For HM, at both times, the four items chosen to represent the danger as
thrilling dimension (response to dangerous experience, pre-test = .49,
post-test = .44; rather gamble or play it safe, pre-test = .66, post-test =
.81; type of party preferred, pre-test = .73, post-test = .78; drive safely
or fast, pre-test = .66, post-test = .36) did, indeed, form one factor that
explained 40.94% of the variance at the pre-test and 39.53% at the
post-test. Only three of the four items chosen to represent the violence as
manly dimension formed one factor; therefore, the fourth item was
eliminated from subsequent analyses. The factor loadings for the dimension
were response to being called a name, pre-test = .71, post-test = .63;
response to remembering past fights, pre-test = .63, post-test = .72;
natural for men to fight, pre-test = .68, post-test = .72. The items
explained 45.39% of the
Hypermasculinity, Aggression, and Television Violence
variance at the pre-test and 47.52% at the post-test. Finally, four of the
five items chosen to represent the callous about sex dimension loaded
cleanly onto one component in the data; therefore, the fifth item was
eliminated from subsequent analyses. The factor loadings for the dimension
were alcohol/drugs tactics to have sex, pre-test = .58, post-test = .86;
men need to have sex, pre-test = .56, post-test = .36, consent necessary to
have sex, pre-test = .71, post-test = .83; and women good for one thing,
pre-test = .77, post-test = .61. The items explained 43.49% of the variance
at the pre-test and 48.2% at the post-test.
        All of the HM items were also examined together to determine whether an
additive HM scale comprised of all 15 items would be reliable. Cronbach's
alpha was a sufficiently reliable .72 for the HM items at the pre-test.
Therefore, an additive scale of responses to the HM items was formed. In
the subsequent analyses used to test the hypotheses, the overall HM scale
is used in order to simplify and reduce the data. However, to address the
research question, the individual dimensions of HM are used in order to
provide a more detailed description.
        Confirmatory factor analysis showed that the items selected to represent
each of the four dimensions of AH loaded cleanly onto one component in the
data from both the pre-test and post-test. The factor loadings for the
physical aggression items were resort to violence, pre-test = .72,
post-test = .68; won't hit back, pre-test = .81, post-test = .70; will not
hit back, pre-test = .79, post-test = .78; and come to blows, pre-test =
.73, post-test = .81. The items explained 57.95% of the variance in the
pre-test and 55.33% in the post-test. The factor loadings for the verbal
aggression items were when people annoy me, pre-test = .75, post-test =
.70; friends say I'm argumentative, pre-test = .77, post-test
Hypermasculinity, Aggression, and Television Violence
= .71; and I keep it to myself, pre-test = .67, post-test = .77. The items
explained 53.14% of the variance at the pre-test and 52.80% at the
post-test. The factor loadings for the anger items were trouble controlling
temper, pre-test = .72, post-test = .87; friends think I'm a hothead,
pre-test = .84, post-test = .76; I fly off the handle, pre-test = .73;
post-test = .75; and I'm an even-tempered person, pre-test = .81, post-test
= .75. The items explained 60.29% of the variance at the pre-test and
60.84% at the post-test. Finally, factor loadings for the items measuring
hostility were people laugh behind my back, pre-test = .65, post-test =
.76; wonder what nice people want, pre-test = .46; post-test = .65; others
always get the breaks, pre-test = .85, post-test = .82; and I've gotten a
raw deal, pre-test = .86, post-test = .82. The items explained 52.07% of
the variance in the pre-test data and 59.49% in the post-test data. Thus,
additive scales were formed for each of the dimensions of AH, and these
newly formed scales were used in subsequent analyses.
Furthermore, subjects' responses to the scenarios were also compiled into
additive scales. Reliability tests revealed that responses to the five
scenarios in which subjects were asked what would you do in that situation
were reliably related in the pre-test data (Cronbach's alpha = .75) as well
as in the post-test data (Cronbach's alpha = .80). Interestingly, responses
were even more consistent when the same scenarios were used to ask subjects
what you would like to do in the situation if they knew there would be no
consequences involved (Cronbach's alpha = .90 in pre-test, .89 in the
post-test). Thus, additive scales were formed for all of these
scenario-related variables ("what would you do," pre- and post-test and
"what would you like to do," pre- and post-test).

Hypermasculinity, Aggression, and Television Violence
Hypothesis Tests
        The first hypothesis predicted that pre-existing, trait aggression would
lead to more aggression in responses to the constructed scenarios following
exposure to violent media stimuli. First, subjects' scores on each of the
four dimensions of AH—anger, physical aggression, verbal aggression, and
hostility—were categorized into high and low, using a median split. After
selecting only those that were exposed to the two violent television
program stimuli (N = 56), repeated measures ANOVA was run. Results find
support for the role of some of the dimensions of AH in predicting reports
of responses to the scenarios (see Table 1). The hypotheses are tested in
terms of whether the dimensions of AH interacted with time of response to
predict the what would you do and the what would you like to do reports.
________________________________________________________________________Table
1 about here
________________________________________________________________________
Of the four AH dimensions, anger is the only one that interacted with time
of response to predict change in subjects' reports to what would you do, F
(1, 54) = 6.36, p < .05. None of the other dimensions of AH interacted with
time of response significantly to predict what would you do
reports—physical aggression (F [1, 54] = .18, ns), verbal aggression (F [1,
54] = .22, ns), or hostility (F [1, 54] = 1.32, ns). Two of the four AH
dimensions interacted with time of response to predict within-subjects
change for those exposed to the two violent stimuli in reports of what
would you like to do in the scenarios. Physical aggression interacted with
time of response to predict change in reports of what
Hypermasculinity, Aggression, and Television Violence
would you like to do, F (1, 54) = 3.83, p = .05, as did anger, F (1, 54) =
4.45, p < .05. No such interaction occurred for time of response and verbal
aggression (F [1, 54 = .68, ns) or time of response and hostility (F [1,
54] = .18, ns). Thus, H1 received partial support.
        The second hypothesis predicted that high levels of pre-existing HM would
lead to more aggression in responses to the constructed scenarios following
exposure to the violent and HM stimulus compared to low levels of
pre-existing HM. Selecting only that treatment group (N = 32), therefore,
repeated measures ANOVA was run with the change in responses to the
scenarios from the pre-test to the post-test as the factor (see Table 2).
Results find support for the role of HM interacting with time of response
to predict an increase in aggression in responses to what subjects would do
in the scenarios, but not in responses to what subjects would like to do if
there were no consequences.
________________________________________________________________________
Table 2 about here
________________________________________________________________________
HM, categorized by median split into high and low, had a main-effect on the
responses to what would you do as a between-subjects factor, F (1, 30) =
4.79, p < .05. The time at which subjects responded to the what would you
do scenarios had a significant within-subjects main effect, F (1, 30) =
20.20, p < .001. Most importantly for the hypothesis, time of response and
degree of HM also interacted significantly to predict within-subjects
responses to the what would you do scenarios, F (1, 30) = 5.22, p < .05.
Those low in pre-existing HM increased from a mean of 8.94 (SD = 6.51) to a
mean of

Hypermasculinity, Aggression, and Television Violence
10.67 (SD = 2.91). Those high in pre-existing HM experienced a larger
increase, from a mean of 10.00 (SD = 2.39) to 15.29 (SD = 3.31) on the what
would you do additive scale.
Dichotomized HM also had a main-effect on responses to what would you like
to do as a between-subjects factor, F (1, 30) = 4.96, p < .05. The time at
which subjects responded to the what would you like to do scenarios had a
main effect as a within-subjects factor, F (1, 30) = 87.03, p < .001.
However, in this case, time of response and HM did not interact
significantly to predict within-subjects responses to the what would you
like to do scenarios, F (1, 30) = 1.63, ns. Those low in pre-existing HM
increased from a mean of 11.00 (SD = 2.43) to a mean of 13.28 (SD = 2.97).
Those high in pre-existing HM increased from a mean of 12.64 (SD = 2.59) to
a mean of 15.64 (SD = 2.53) on the what would you like to do additive
scale. The difference in size of the increase according to degree of HM was
not large enough to achieve statistical significance. Thus, overall, there
is partial evidence for H2, supported in the case of the what would you do
reports but not in the case of the what would you like to do (if no
consequences) reports.
        The third hypothesis predicted that the increase in state aggression
experienced by subjects would be higher if they were exposed to the
realistic television stimulus compared to the fantasy-based television
stimulus. The fourth hypothesis predicted that both of these groups—those
exposed to the realistic stimulus and to the fantasy-based stimulus, both
of which contained violence—would be more likely to experience an increase
in state aggression compared to the control group. Both hypotheses were
tested with repeated measures ANCOVA, controlling for overall television
exposure and amount of violence in favorite programs (see Table 3). The
control variables are used
Hypermasculinity, Aggression, and Television Violence
here, but not used in the preceding tests, because the entire sample is
employed to test H3 and H4, whereas H1 and H2 were relevant only to a
portion of the sample. Concern regarding whether the sample size would
allow for sufficient power made this necessary.
________________________________________________________________________
Table 3 about here
________________________________________________________________________
For the what would you do additive scale, group/condition was not a
significant main-effect, between-subjects factor, F (2, 86) = 2.02, p =
.14. The time at which subjects responded to the what would you do
scenarios had a significant main effect as a within-subjects factor, F (1,
86) = 10.23, p < .01. Most importantly, in support of H3 and H4, time of
response and group/condition interacted significantly to predict
within-subjects change in responses to the what would you do scenarios, F
(2, 86) = 5.26, p < .01. Those exposed to the control television program
(Seventh Heaven) experienced a slight increase on the what would you do
scale, from M = 8.91 (SD = 3.09) to M = 9.40 (SD = 4.15). Those exposed to
the fantasy-based, violence without HM television program (Buffy the
Vampire Slayer) experienced a larger increase, from M = 8.58 (SD = 3.20) to
M = 10.75 (SD = 4.34). Finally, those exposed to the realistic, violence
with HM television program (The Sopranos) experienced a still larger
increase, from M = 9.41 (SD = 5.09) to M = 12.69 (SD = 3.83). Therefore, H3
is supported because The Sopranos exposure resulted in a larger increase in
state aggression than Buffy the Vampire Slayer exposure. Yet, H4 is not
supported in this case because members of the control group experienced a
slight

Hypermasculinity, Aggression, and Television Violence
increase in aggression in responses to the what would you do scenarios, and
the hypothesis predicted that they would not.
        For the what would you like to do additive scale, group/condition was not
a significant main-effect, between-subjects factor, F (2, 86) = .06, ns. On
the other hand, the time at which subjects responded to the what would you
like to do scenarios approached significance as a main effect,
within-subjects factor, F (1, 86) = 3.39, p = .07. Most importantly, time
of response and group/condition interacted significantly to predict
within-subjects change in responses to the what would you like to do
scenarios, F (2, 86) = 10.11, p < .001. Those exposed to the control
television program (Seventh Heaven) experienced a slight decrease on the
what would you like to do scale, from M = 14.09 (SD = 6.48) to M = 13.00
(SD = 5.33). Those exposed to the fantasy-based, violence without HM
television program (Buffy the Vampire Slayer) experienced a small increase,
from M = 12.67 (SD = 6.28) to M = 13.75 (SD = 6.27). Finally, those exposed
to the realistic, violence with HM television program (The Sopranos)
experienced a larger increase, from M = 11.72 (SD = 2.59) to M = 14.31 (SD
= 2.99). Thus, due to the significant interaction between time of response
and group/condition, H3 and H4 are supported in this case.
        In order to address the research question and discover whether exposure to
the stimulus that contained violence and HM (The Sopranos) would be
associated with an increase in self-reports of HM, paired t-tests were run
for those assigned to this condition (N = 32). Pre-test and post-test
reports on each of the three dimensions of HM were entered as paired
variables. Results show that after exposure, subjects had higher
self-reports for two of the three dimensions of HM, danger as thrilling
(pre-test, M = 5.19,
Hypermasculinity, Aggression, and Television Violence
SD = 1.31; post-test, M = 6.28, SD = 1.97; t = -4.35, p < .001) and
violence as manly (pre-test, M = 4.09, SD = 0.78; post-test, M = 5.25, SD =
0.98; t = -7.40, p < .001). For the remaining dimension, callous attitudes
toward sex, subjects reported no such increase (pre-test, M = 4.50, SD =
0.80; post-test, M = 4.47, SD = 0.76; t = .44, ns).
Discussion
Conclusions
The results point first to the complexity of the concepts of
aggression/hostility and hypermasculinity, in that the dimensions of both
often operated differently from one another in relationships with other
factors. In general, however, the study did find evidence that the
pre-existing aggression and hypermasculinity of male television viewers has
the potential to predict the degree of aggressiveness in responses to
violent television.
In the case of pre-existing aggression and hostility, the dimensions of
anger and physical aggression predicted changes in the state aggressiveness
of subjects' answers to the question "what would you like to do" in
response to the constructed scenarios. The AH dimension of anger predicted
such changes in answers to the "what would you do" question. One
explanation for the addition of physical aggression in the "what would you
like to do" case, is that, in these questions, subjects were asked to
report their preferred response to the frustrating scenarios if they knew
there would be no consequences. Perhaps subjects needed to be assured that
there would be no negative repercussions in order for their physical
aggression tendencies—arguably a less socially acceptable response than
anger— to surface as a predictor of their responses. Nonetheless, results of

Hypermasculinity, Aggression, and Television Violence
the study support the past finding that pre-existing aggression is an
important predictor of aggressive responses to violent media (Bushman,
1995, 1996; Huesmann et al., 1984).
Pre-existing hypermasculinity, conversely, predicted change in responses to
"what would you do", but not to "what would you like to do" in relation to
the frustrating scenarios. One possible explanation for this result is that
males high in HM may have been bold enough to report little difference
between their responses to what they would do and what they would like to
do if reassured of no negative repercussions. In other words, perhaps the
tough, macho orientation described in the personality constellation (Mosher
& Sirkin, 1984) contributes to the notion that these individuals may say
that they will do what they please and are unconcerned about potential
consequences. Overall, the study at hand reinforces past research that has
linked HM with aggression (Gold et al., 1992), as well as those past
studies that have determined that HM predicts aggressive responses to media
(XXXX, 2001a; Beaver, Gold, & Prisco, 1992; Russell, 1992).
The results also point fairly clearly to the conclusion that responses to
violent television stimuli will differ depending upon the nature of the
violent depiction (Atkin, 1983; Berkowitz & Alioto, 1973; Jo & Berkowitz,
1994). The more realistic television stimulus—that also combined HM and
violent portrayals—resulted in larger increases in aggressive responses to
the constructed scenarios compared to the less realistic television
stimulus that contained violence but no evidence of HM. Indeed, those
subjects who watched The Sopranos experienced the largest increases between
pre- and post-test aggressiveness in both the "what would you do" and the
"what would you like to do" questions compared to those who watched the
other two programs. Results were
Hypermasculinity, Aggression, and Television Violence
relatively linear, as predicted, with The Sopranos outscoring Buffy the
Vampire Slayer, which, in turn, outscored Seventh Heaven in predicting
increases in state aggression, with one exception. Members of the control
group (Seventh Heaven) actually experienced an unpredicted increase in
aggressiveness in their "what would you do" responses to the constructed
scenarios. This could be due to the sentiment expressed by some of the
group members as they left the experiment, that they had disliked the
program, finding it preachy and contrived.
The final major conclusion, and perhaps the most unique contribution of the
study, is that there is some evidence to suggest that television exposure
can contribute to hypermasculinity. Those who were exposed to a television
stimulus that contained HM male characters (The Sopranos) reported higher
estimates in their self-reports of their own HM tendencies after exposure
compared to before. Two of the three dimensions of HM, danger as thrilling
and violence as manly, increased from the pre- to post-test among these
subjects, whereas callousness toward sex remained the same. The failure of
self-reports on the latter dimension to increase is curious because the
television stimulus modeled this dimension. The episode featured main
character Tony Soprano becoming very angry at his mistress, threatening
her, and then nearly strangling her. Perhaps subjects were actually
sensitized to the issue of violence against women from watching the fairly
realistic and potentially disturbing scene, and therefore, their
self-reports of callousness toward sex did not increase. In any case, the
study represents the first known attempt to determine whether television
exposure is a causal contributor to HM, and results suggests tentative
support for two of the three dimensions of HM.
Hypermasculinity, Aggression, and Television Violence
Theoretical Implications
        The study at hand examines two individual differences in television
audience members—their hypermasculinity and their aggression—and measures
the extent to which each determines the likelihood and the magnitude of an
aggressive response to television violence. It is argued here that
neo-associationism (Berkowitz, 1984) and the mental models theory
(Roskos-Ewoldsen, Roskos-Ewoldsen, & Dillman Carpentier, 2002) suggest that
those inclined toward HM and aggression may have greater accessibility to
aggressive frames to use to define and respond to new situations. In this
study, such new situations were constructed for subjects in the form of a
series of scenarios detailing frustrating circumstances that many young men
may encounter, similar to past measures used by Cahoon and Edmunds (1984).
The study at hand tests the idea that those subjects with prior high levels
of HM and aggression already have a vast framework of past experiences and
attitudes about aggression to use to make sense of provoking and
frustrating new situations, compared to those lower in HM and aggression.
Furthermore, neo-association theory suggests that a violent television
stimulus is the prime that evokes those stored ideas and brings them front
and center in one's mind after viewing (Jo & Berkowitz, 1994). Thus, as is
shown in the data gathered in this study, one's pre-existing personality
tendencies and past experiences interact with exposure to a violent
television stimulus to shape how one responds to a new situation.
Furthermore, the theoretical framework used in this study suggests that
high aggressive and high HM individuals are more likely to apply aggressive
mental models to
Hypermasculinity, Aggression, and Television Violence
new situations if they perceive a link between themselves and the
television stimulus (Berkowitz, 1984; Roskos-Ewoldsen, Roskos-Ewoldsen, &
Dillman Carpentier, 2002). A viewer who is strongly identifying with
aggressive characters while viewing television will have more accessible
aggressive thoughts in mind immediately after viewing (Jo & Berkowitz,
1994). Priming of aggressive responses and/or the use of aggressive mental
models to respond to a new situation is, therefore, more likely if there is
a high degree of realism in the television depiction (Atkin, 1983;
Berkowitz & Alioto, 1973) because identification and involvement are
facilitated by realism (Jo & Berkowitz, 1994). The results of this study
support the central role of perceived similarity with characters and
realism of the portrayal in predicting aggressive responses to violent
media in that the realistic stimulus (The Sopranos) triggered higher
aggressive responses than the fantasy-based stimulus (Buffy the Vampire
Slayer).
Limitations and Future Research
        There are certainly limitations to the study that should be kept in mind
when interpreting its results. First, the possibility exists that although
there were false purpose items in the pre- and post-test questionnaires,
subjects in the treatment conditions may have determined that, in essence,
the study was about violent television and aggression, which may have
impacted their responses. Second, as with nearly experiment, the study is
constrained by artificiality in that viewing occurred in a room with the
author present. Indeed, since the author is female, it is possible that
subjects' responses to the callousness about sex or women hypermasculinity
questions were affected by her presence. Strategies that were undertaken to
minimize this possibility were the assurance of anonymity and
Hypermasculinity, Aggression, and Television Violence
strong encouragement for subjects to be honest and candid. Finally, using
The Sopranos to represent a somewhat realistic television program with HM
and violence and Buffy the Vampire Slayer to represent a fantasy-based
program with violence and little to no HM was an appropriate means of
investigating the factors associated with the depiction of violence.
However, in future research, multiple programs can be used in order to
eliminate the conflation of the reality/fantasy factor and the violence
only/violence and HM factor that occurred in this study.
Overall, this study attempts to advance knowledge in a number of areas—the
individual differences that predict aggressive responses to violent
television, the factors in the depictions of the violent television itself
that predict aggressive responses, and the potential for television
exposure to also serve as a cause of HM in young males. It is the author's
view that in media effects research, variables should no longer necessarily
be viewed as only causes (independent variables) or effects (dependent
variables). That somewhat narrow perspective can be replaced with a more
sophisticated and complex view of human behavior that acknowledges that
some psychosocial forces can be at once causes and effects. Thus, future
research should continue to examine the relatively novel concept of
hypermasculinity as not only a pre-existing condition that may shape
responses to media, but also as a dynamic concept that may be triggered by
media exposure as well. Such a development in the study of television
violence is necessary to extend and further understand this
well-researched, yet always socially significant, topic.


Hypermasculinity, Aggression, and Television Violence
References

XXXX, 2001a
XXXX, 2001b
        Atkin, C. (1983). Effects of realistic TV violence vs. fictional violence
on aggression. Journalism Quarterly, 60, 615-621.

Bandura, A., Ross, D., & Ross, S.A. (1963). Imitation of film-mediated
aggressive models. Journal of Abnormal and Social Psychology, 63, 575-582.

        Beaver, E.D., Gold, S.R., & Prisco, A.G. (1992). Priming macho attitudes
and emotions. Journal of Interpersonal Violence, 7, 321-333.

Berkowitz, L. (1970). Aggressive humors as a stimulus to aggressive
responses. Journal of Personality and Social Psychology, 16, 710-717.

Berkowitz, L. (1984). Some effects of thoughts on anti- and prosocial
influences of media events: A cognitive-neoassociation analysis.
Psychological Bulletin, 95, 410-427.

Berkowitz, L. (1990). On the formation and regulation of anger and
aggression: A cognitive-neoassociation analysis. American Psychologist, 45
(4), 494-503.

Berkowitz, L. (1993). Aggression: Its causes, consequences, and control.
New York: McGraw Hill.

Berkowitz, L., & Alioto, J. (1973). The meaning of an observed event as a
determinant of its aggressive consequences. Journal of Personality and
Social Psychology, 28, 206-217.

Berkowitz, L., & Rawlings, E. (1963). Effects of film violence on
inhibitions against subsequent aggression. Journal of Abnormal and Social
Psychology, 66, 405-412.

Broude, G.J. (1990). Protest masculinity: A further look at the causes and
the concept. Ethos, 18(1), 103-122.

Bushman, B.J. (1995). Moderating role of trait aggressiveness in the
effects of violent aggression. Journal of Personality and Social
Psychology, 69 (5), 950-965.



Hypermasculinity, Aggression, and Television Violence

Bushman, B.J. (1996). Individual differences in the extent and development
of aggressive cognitive-associative networks. Personality and Social
Psychology Bulletin, 22, 811-820.

Buss, A.H., & Durkee, A. (1957). An inventory for assessing different kinds
of hostility. Journal of Consulting Psychology, 21 (4), 343-346.

Buss, A.H., & Perry, M. (1992). The aggression questionnaire. Journal of
Personality & Social Psychology, 63(3), 452-459.

Cahoon, D.D., & Edmonds, E.M. (1984). Guns/no guns and the expression of
social hostility. Bulletin of the Psychonomic Society, 22(4), 305-308.

Cahoon, D.D., & Edmonds, E.M. (1985). The weapons effect: Fact or artifact.
Bulletin of the Psychonomic Society, 23(1), 57-60.

Collison, M. (1996). In search of the high life: Drugs, crime,
masculinities and consumption. British Journal of Criminology, 36(3), 428-444.

Comstock, G. & Scharrer, E. (1999). Television: What's on, who's watching,
and what it means. San Diego, CA: Academic Press.

Donnerstein, E., & Berkowitz, L. (1983). Effects of film content and victim
association on aggression behavior and attitudes. Unpublished study,
University of Wisconsin, Madison.

Donnerstein, E., Linz, D., & Penrod, S. (1987). Changes in viewers'
attitudes and beliefs. The impact of violent pornography. Is it the sex or
is it the violence? The question of pornography (pp. 74-136). New York: The
Free Press.

Geen, R.G., & George, C. (1969). Relationship of manifest aggressiveness to
aggressive word associations. Psychological Reports, 25, 711-714.

Gold, S.R., Fultz, J., Burke, C.H., Prisco, A.G., & Willett, J.A. (1992).
Vicrious emotional responses of macho college males. Journal of
Interpersonal Violence, 7(2), 165-174.

Govia, J.M., & Velicer, W.F. (1985). Comparison of multidimensional
measures of aggression. Psychological Reports, 57, 207-215.

Herr, P. (1986). Consequences of priming: Judgment and behavior. Journal of
Personality and Social Psychology, 51, 1106-1115.


Hypermasculinity, Aggression, and Television Violence

Huesmann, L.R., Eron, L.D., Lefkowitz, M.M., & Walder, L.O. (1984).
Stability of aggression over time and generations. Developmental
Psychology, 20 (6), 1120-1134.

Iyengar, S., & Kinder, D. (1987). News that matters: Television and
American opinion, Chicago, University of Chicago Press, 167-85.

Jo, E., & Berkowitz, L. (1994). A priming effect analysis of media
influences: An update. In J. Bryant and D. Zillmann (Eds.), Media effects:
Advances in theory and research. Hillsdale, NJ: Erlbaum.

Josephson, W.L. (1987). Television violence and children's aggression:
Testing the priming, social script, and disinhibition predictions. Journal
of Personality and Social Psychology, 53 (5), 882-890.

Levine, S.R., Wyer, R.S., & Schwarz, N. (1994). Are you what you feel? The
affective and cognitive determinants of self-judgments. European Journal of
Social Psychology, 24 (1), 63-77.

Maguire, K. & Pastore, A.L. (Eds.). (1995). Sourcebook of criminal justice
statistics. US Department of Justice, Bureau of Justice Statistics,
Washington, DC: US Government Printing Office.

Mendelsohn, M. (1996). The media and interpersonal communications: The
priming of issues, leaders, and party identification. The Journal of
Politics, 58 (1), 12-26.

Menon, G., Raghubir, P., & Schwarz, N. (1995). Behavioral frequency
judgments: An accessibility-diagnosticity framework. Journal of Consumer
Research, 22 (2), 212-228.

Mosher, D.L., & Sirkin, M. (1984). Measuring a macho personality
constellation. Journal of Research in Personality, 18, 150-163.

Mosher, D.L., & Tompkins, S.S. (1988). Scripting the macho man:
Hypermasculine socialization and enculturation. Journal of Sex Research,
20(1), 60-84.

Moss, H.E., Ostrin, R.K., Tyler, L.K., & Marslen-Wilson, W.D. (1995).
Accessing different types of lexical semantic information: evidence from
priming. Journal of Experimental Psychology: Learning, Memory and
Cognition, 21 (4), 863-874.

Oliver, M.B. (2002). Individual differences in media effects. In J. Bryant
& D. Zillmann (Eds.), Media effects: Advances in theory and research.
Mahwah, NJ: Erlbaum, pp. 507-524.

Hypermasculinity, Aggression, and Television Violence

Potter, W.J. (1999). On media violence. Thousand Oaks, CA: Sage.

Roskos-Ewoldsen, D.R., Roskos-Ewoldsen, B., & Dillman Carpentier, F.R.
(2002). Media priming: A synthesis. In J. Bryant & D. Zillmann (Eds.),
Media effects: Advances in theory and research. Mahwah, NJ: Erlbaum, pp.
97-120.

Russell, G.W. (1992). Response of the macho male to viewing a combatant
sport. Journal of Social Behavior and Personality, 7 (4), 631-638.

Schwarz, N. (1995). Social cognition: Information accessibility and use in
social judgment. In Smith, E., Osherson, D.N. (Eds). Thinking: An
invitation to cognitive science, vol. 3 (2nd ed.). Cambridge, MA: MIT Press.

Schwarz, N. (1998). Accessible content and accessibility experiences: The
interplay of declarative and experiential information in judgment.
Personality & Social Psychology Review, 2 (2), 87-99.

Schwarz, N. & Bienias, J. (1990). What mediates the impact of response
alternatives on frequency reports of mundane behaviors? Applied Cognitive
Psychology, 4 (1), 61-72.

Schwarz, N., Bless, H., Strack, F., Klumpp, G., et al. (1991). Ease of
retrieval as information: Another look at the availability heuristic.
Journal of Personality & Social Psychology, 61 (2), 195-202.

Schwarz, N., Strack, F., Kommer, D., & Wagner, D. (1987). Soccer, rooms,
and the quality of your life: Mood effects on judgments of satisfaction
with life in general and with specific domains. European Journal of Social
Psychology, 17 (1). 69-79.

Singer, D.G., & Singer, J.L. (Eds.). (2001). Handbook of children and the
media. Thousand Oaks, CA: Sage.

Smeaton, G., & Byrne, D. (1987). The effects of R-rated violence and
erotica, individual differences, and victim characteristics on acquaintance
rape proclivity. Journal of Research in Personality, 21, 171-184.

Sparks, R. (1996). Masculinity and heroism in the Hollywood "blockbuster":
The cultural industry and contemporary images of crime and law enforcement.
British Journal of Criminology, 36(3), 348-360.

Strack, F., Martin, L.L., & Schwarz, N. (1988). Priming and communication:
Social determinants of information use in judgments of life satisfaction.
European Journal of Social Psychology, 18 (5), 429-442.

Hypermasculinity, Aggression, and Television Violence

Sullivan, J.P., & Mosher, D.L. (1990). Acceptance of guided imagery of
marital rape as a function of macho personality. Violence and Victims, 5,
275-286.

Turner, C., & Berkowitz, L. (1972). Identification with film aggressor
(covert role taking) and reactions to film violence. Journal of Personality
and Social Psychology, 33, 755-763.

U.S. Department of Health and Human Services. (2001). Youth violence: A
report of the Surgeon General. Rockville, MD: U.S. Department of Health and
Human Services, Centers for Disease Control and Prevention, National Center
for Injury Prevention and Control, Substance Abuse and Mental Health
Services Administration, Center for Mental Health Services, and National
Institutes of Health, National Institute of Mental Health.

Weir, J.A., & Branscombe, N.A. (1990). Discriminating between males who do
and do not possess a proclivity for sexual aggression. Paper presented at
the annual meeting of the Southeastern Psychological Association.

Wilson, B.J., Kunkel, D., Linz, D., & Potter, W.J. (1998). Violence in
television programming overall, National Television Violence Study Volume
2. Thousand Oaks, CA: Sage Publications.

Wyer, R., Jr., & Srull, T. (1981). Category accessibility: Some theoretical
and empirical issues concerning the processing of information. In E.
Higgins, C. Herman, & M. Zanna (Eds.), Social cognition (Vol. 1, pp.
161-197). Hillsdale, NJ: Lawrence Erlbaum.

Zaitchik, M.C., & Mosher, D.L. (1993). Criminal justice implications of the
macho personality constellation. Criminal Justice and Behavior, 20 (3),
227-239.

Zillmann, D. (1991). Television viewing and physiological arousal. In J.
Bryant & D. Zillmann (Eds.), Responding to the screen (pp. 103-133).
Hillsdale, NJ: Lawrence Erlbaum.

Zillmann, D., & Bryant, J. (1991). Responding to comedy: The sense and
nonsense in humor. In J. Bryant & D. Zillmann (Eds.), Responding to the
screen (pp. 261-279). Hillsdale, NJ: Lawrence Erlbaum.





Table 1. Repeated measures ANOVA, with pre-existing trait aggression and
exposure to violent television stimulus as independent variables, and
change in state-based aggression dependent variables, N = 56.
                                                State aggression
                                   Time 1                           Time 2
DV Exposure                     M (SD)  M (SD)          M (SD)  M (SD)
_______________________________________________________________________What
would you do?
        Low physical agg.       8.69 (5.05)                     11.31 (3.94)
        High physical agg.      9.67 (2.92)                     12.76 (4.38)
                                Change over time, F (1, 54) = 27.29, p < .001
                                Between subjects differences, F (1, 54) = 1.38, ns
                                Change over time X Phys. Agg., F (1, 54) = .18, ns

        Low verbal agg. 9.38 (5.43)                     11.92 (4.13)
        High verbal agg.        8.77 (3.24)                     11.80 (4.37)
                                Change over time, F (1, 54) = 27.36, p < .001
                                Between subjects differences, F (1, 54) = .13, ns
                                Change over time X Verbal agg., F (1, 54) = .22, ns

        Low anger               9.06 (5.08)                     10.86 (3.98)
        High anger              9.33 (2.73)                     13.52 (3.72)
                                Change over time, F (1, 54) = 39.94, p < .001
                                Between subjects differences, F (1, 54) = 2.02, p = .12
                                Change over time X Anger, F (1, 54) = 6.36, p = .01

        Low hostility           9.46 (5.43)                     11.62 (4.04)
        High hostility          8.70 (3.22)                     12.07 (4.27)
                                Change over time, F (1, 54) = 27.41, p < .001
                                Between subjects differences, F (1, 54) = .03, ns
                                Change over time X Hostility, F (1, 54) = 1.32, ns

What would you like to do?
        Low physical agg.       11.31 (2.41)                    12.60 (2.82)
        High physical agg.      13.48 (6.59)                    15.95 (6.15)
                                Change over time, F (1, 54) = 38.28, p < .001
                                Between subjects differences, F (1, 54) = 5.49, p <.05
                                Change over time X Phys. Agg., F (1, 54) = 3.83, p < .05


        Low verbal agg. 11.50 (2.66)                    13.50 (2.96)
        High verbal agg.        12.67 (5.67)                    14.17 (5.72)
                                Change over time, F (1, 54) = 33.20, p < .001
                                Between subjects differences, F (1, 54) = .59, ns
                                Change over time X Verbal agg., F (1, 54) = .68, ns

        Low anger               11.51 (2.62)                    12.94 (3.08)
        High anger              13.14 (6.56)                    15.95 (6.10)
                                Change over time, F (1, 54) = 41.90, p < .001
                                Between subjects differences, F (1, 54) = 3.79, p = .05
                                Change over time X Anger, F (1, 54) = 4.45, p < .05

        Low hostility           12.84 (6.20)                    14.46 (6.03)
        High hostility          11.50 (2.21)                    13.33 (2.93)
                                Change over time, F (1, 54) = 31.91, p < .001
                                Between subjects differences, F (1, 54) = 1.09, ns
                                Change over time X Hostility, F (1, 54) = .13, ns





























Table 2. Repeated measures ANOVA with categorized hypermasculinity as the
independent variable and changes in state aggression as the dependent
variables, N = 32.

State aggression
                                   Time 1                           Time 2
DV Exposure                     M (SD)  M (SD)          M (SD)  M (SD)
_______________________________________________________________________What
would you do?
        Low HM          8.94 (6.51)                     10.67 (2.91)
        High HM         10.00 (2.39)                    15.29 (3.31)
                                Change over time, F (1, 30) = 20.20, p < .001
                                Between subjects differences, F (1, 30) = 4.79, p < .05
                                Change over time X HM, F (1, 30) = 5.22, p < .05
What would you like to do?

        Low HM          11.00 (2.43)                    13.28 (2.97)
        High HM         12.64 (2.59)                    15.64 (2.53)
                                Change over time, F (1, 30) = 87.03, p < .001
                                Between subjects differences, F (1, 30) = 4.96, p <.05
                                Change over time X HM, F (1, 30) = 1.63, ns





















Table 3. Repeated measures ANCOVA with treatment condition as the
independent variable and changes in state aggression as the dependent
variables, controlling for overall television exposure and amount of
violence in favorite programs, N = 91.

State aggression
                                              Time 1                        Time 2
DV Exposure                             M (SD)  M (SD)          M (SD)  M (SD)
______________________________________________________________________
What would you do?
        Seventh Heaven          8.91 (3.09)                     9.40 (4.15)
        Buffy the Vampire Slayer        8.58 (3.20)                     10.75 (4.34)
        The Sopranos                    9.04 (5.09)                     12.69 (3.91)
                                Change over time, F (1, 86) = 10.23, p < .01
                                Between subjects differences, F (2, 86) = 2.02, ns
                                Change over time X group, F (2, 86) = 5.26, p < .01

What would you like to do?

        Seventh Heaven          14.09 (6.48)                    13.00 (5.33)
        Buffy the Vampire Slayer        12.67 (6.28)                    13.75 (6.28)
        The Sopranos                    11.72 (2.59)                    14.31 (2.99)
                                Change over time, F (1, 86) = 3.39, p = .06
                                Between subjects differences, F (2, 86) = .06, ns
                                Change over time X group, F (2, 86) = 10.11, p < .001