THESIS

2 004 LIBRARY
5 - sir/j Michigan State
9 I ' University

This is to certify that the
thesis entitled

TELEVISED SPORTS AND GAM:
AN EXTENTION OF THE GENERAL AGGRESSION MODEL

presented by

DAVID KEITH WESTERMAN

has been accepted towards fulfillment
of the requirements for the

MA. degree in COMMUNICATION

 

Major Professor’s Signature

la-g'OB

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

L

SEN 2% E006 ll

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

 

 

TELEVISED SPORTS AND GAM:
AN EXTENTION OF THE GENERAL AGGRESSION MODEL

By

David Keith Westerman

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirement
For the degree of

MASTER OF ARTS
Department of Communication

2003

ABSTRACT

TELEVISED SPORTS AND GAM:
AN EXTENTION OF THE GENERAL AGGRESSION MODEL

By

David Keith Westerman

Previous research attempting to answer the question of why exposure to media
violence leads to subsequent aggression has led to the formation of the General
Aggression Model (GAM) (Anderson & Bushman, 2002a). This paper attempts to
extend the findings of the GAM by testing it using televised sports. More specifically, it
examines the effects that violent sports have on the formation of a hostile expectancy
bias, as tested previously with video games (Bushman & Anderson, 2002). It also
examines the impact of the scripted nature of the sport on this relationship, as
differentiated previously by Depalma and Raney (2002).

One hundred and thirty-nine undergraduates from an introductory communication
course participated in this study. Participants were exposed to one of four sports clips:
scripted violent (professional wrestling), scripted non-violent (pair’s figure skating), non-
scripted violent (boxing), or non-scripted non-violent (baseball). After exposure, each
subject was given three incomplete story stems to complete (as used by Bushman &
Anderson, 2002), as well as a series of other questionnaires. The data were consistent
with the hypothesis that violent sports lead to a greater formation of a hostile expectancy
bias than nonviolent sports, and the scripted nature of the sport was found to moderate the
relationship. The robustness of these relationships is also discussed. Limitations and

implications of the study are then discussed, as are suggestions for future research.

ACKNOWLEDGMENTS

Rarely, if ever, is a project like a masters’ thesis completed solely by the author.
This thesis is no exception. There are many people I would like to thank for helping me,
either directly or indirectly, in finishing this project.

The first thanks go out to my committee members, Dr. Bradley Greenberg and Dr.
Charles Atkin. Your insight and comments helped make this a better paper than I could
have hoped to write by myself.

Thanks also go to my advisor, Dr. Ronald Tamborini. Thank you for all the
support and help both in starting and completing this manuscript. Thank you also for all
the time you put into this project, and for all the time you have put into me as a person.

Thanks and love go to my family and friends, the Lynch Hall All-Stars. Without
your constant support and care through all these years, I would never have even been in
the position to complete thins. Further thanks go out to Dr. Mark Bamer and Dr. James
Wittebols for encouraging me in the right direction.

Special thanks go out to Kelli Asada, Sarah Crook, Dr. Ken Lachlan and Emily
Moyer for taping televised sports for me since I didn’t have a television. Without your
help, this project would not have been possible.

This is in no way an exhaustive list of all the people who should probably be
thanked for contributing to this thesis or to me, but one more extra special thanks that
needs to be included is for Cat Kingsley. Thank you for putting up with me when this
project was being completed and for continuing to do so now that it is done. I know it

was probably not easy, and I want to say that I really appreciate it. Thank you.

iii

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................. vi
LIST OF FIGURES ............................................................................................ vi
INTRODUCTION ............................................................................................... 1
The General Aggression Model .................................................................... l
GAM and Media Violence ............................................................................ 3
Research on Media Violence ......................................................................... 5
Televised Sports and Aggression .................................................................. 6
“Scripted” vs. “Unscripted” Sports ..................................................... 7
METHOD ........................................................................................................... 11
Overview ...................................................................................................... 11
Participants ................................................................................................... 12
Procedure ..................................................................................................... 12
Materials ....................................................................................................... 13
Outcome Measures ...................................................................................... 15
Control Measures ......................................................................................... 16
RESULTS ................................................................................. . ......................... 19
The Main Effect of Sports Violence ........................................................... 19
The Moderating Effect of Scriptedness ...................................................... 20
DISCUSSION .................................................................................................... 23
APPENDICES ................................................................................................... 29
Appendix A. Consent Form ........................................................................ 29
Appendix B. Debriefing Statement ............................................................. 30
Appendix C. Pilot Test Items for Stimulus Material .................................. 31
Appendix D. Media Use and Demographic Items ...................................... 32
Appendix E. The Revised Eysenck Personality Questionnaire .................. 34
Appendix F. Story Stems ............................................................................. 36
Appendix G. Coding Scheme ...................................................................... 38
Appendix H. Buss Perry Aggression Scale ................................................ 39
Appendix 1. Experimental Script ................................................................. 42
Appendix J. The General Aggression Model .............................................. 44
Appendix K. Sample Hostile Responses .................................................... 45
Appendix L. Means and Standard Deviations ............................................ 46
REFERENCES .................................................................................................. 50

iv

LIST OF TABLES

Percentage of Hostile Responses by Condition ................................................ 21

Total Number of Hostile Responses by Condition ........................................... 22

LIST OF FIGURES

The General Aggression Model ......................................................................... 45

Vi

INTRODUCTION

Although the question of whether or not exposure to violent media makes people
more aggressive may be unresolved in the public’s eye, among some in the scientific
community, it has been answered affirmatively (Bushman & Anderson, 2001a). Recent
efforts to explicate processes thought to govern this effect have resulted in the formation
of the General Aggression Model (GAM) (Anderson & Bushman, 2002a). GAM
suggests that exposure to violent media alters internal processes related to hostility that
govern audience behaviors, and results in aggressive outcomes. Support for the model
has been found in studies using fictional content conventionally studied in media research
including violent film (Anderson, 1997) and video games (Anderson & Dill, 2000,
Bushman & Anderson, 2002). While valuable within this limited context, the application
of GAM to non-fictional media holds value not only for its capacity to show broad
relevance, but also for its potential to identify variables that might moderate the model’s
use. This paper extends the use of GAM to televised sports violence and attempts to
determine the extent to which the scripted or unscripted nature of televised sports
violence moderates the outcome of GAM processes after controlling for trait aggression,
viewer enjoyment, and several other potential competitor variables.
The General Aggression Model

Anderson and his colleagues (Anderson, 1997; Anderson, Anderson, Dill, &
Deuser , 1998; Anderson, Deuser, & DeNerve, 1995; Anderson, Anderson, & Deuser,
1996; Anderson & Dill, 2000; Bushman & Anderson, 2001b; Lindsey & Anderson, 2000)
have developed and constantly refined a model of human aggression called the General

Aggression Model (GAM). The most recent version of GAM (Anderson & Bushman,

2002a) integrates five overlapping domain specific theories including: Berkowitz’s
cognitive neoassociation theory (Berkowitz, 1990), social learning theory (Bandura,
2001), script theory (Huesmann, 1998; Schank & Abelson, 1977), excitation transfer
theory (Zillmann, 1983), and social interaction theory (Tedeschi & Felson, 1994).

The model explains human aggression as a process involving a series of routes
connecting cognitive inputs and several aggression related outcomes. Inputs are made up
of variables related to both the person and situation. A variety of characteristics of a
person or their situation are said to influence an outcome by changing the person’s
internal states. This internal change takes place through one or more of three routes:
cognition, affect, or arousal, and leads to an immediate appraisal of the situation at hand.
If there are insufficient resources to appraise the situation or the outcome is not
important, an impulsive action will be enacted to respond to the input variable. If the
outcome is important, a reappraisal process will occur and re-occur until a thoughtful
action is enacted. In this manner, GAM suggests that aggression occurs because personal
traits and situational variables interact to create a change in the present internal state of a
person, which then causes the person to appraise the situation and decide on a course of
action to take to deal with the present state.

The predictions of GAM have been supported in tests using several different
situation and person inputs, including pain, (e.g., Anderson, Anderson, Dill, & Deuser,
1998), heat (e.g., Andserson, Deuser, & DeNeve, 1995), pictures of weapons (e.g.,
Anderson, Anderson, & Deuser, 1996), and trait aggression (e.g., Dill, Anderson,
Anderson, & Deuser, 1997). Of particular interest here are studies using different forms

of violent media like video games (Anderson & Dill, 2000; Bushman & Anderson, 2002)

and film (Anderson, 1997). These studies show GAM is applicable to several areas of
narrative fictional media violence; however, the question remains as to whether or not
GAM applies to other violent media forms.

GAM and Media Violence

Anderson (1997) tested GAM using violent film as a situational variable and trait
aggression as a personal variable. He found that participants exposed to a violent film
clip (from Karate Kid III) expressed greater affective hostility, as measured by a state
hostility questionnaire, than those exposed to an equally arousing, but non-violent clip
(from Gorillas in the Mist). Surprisingly, no significant difference was found for
cognitive hostility, as measured by a word reaction time activity. A second study found
that exposure to a violent film clip does lead to an increase in aggressive thoughts, but
only for people low in trait hostility.

In two studies on video games, Anderson and Dill (2000) found evidence that
violent video games also can impact internal states related to aggression. In the first
study, they demonstrated that playing a violent video game leads to faster recognition of
hostile words than playing a non-violent game. Consistent with GAM, this evidence
suggests that the violent video games can act to prime hostility and produce a change in
present internal states through routes of cognition. The second study produced similar
results with behavioral outcome indicators. Anderson and Dill (2000) measured
aggression by recording the intensity and length of a noise blast participants attempted to
deliver to opponents in a competitive reaction time task. In the experiment, if the subject
lost, their opponent was able to give them a noise blast of variable intensity and length.

In turn, if they won, they determined the intensity and length of a noise blast delivered in

return. In reality, there was no actual opponent; participants were simply told they had
lost on different trials, and a computer randomly generated the duration and intensity of
the noise blast. The researchers theorized that the combination of both the priming of
aggression through video games use and the immediate provocation of the noise blast
from the supposed opponent would be necessary to increase participants’ aggression
levels. Consistent with this belief, they found that after trials where the subject
supposedly won, there was no difference in aggression based on video game condition.
However, in trials alter the subject supposedly lost, there were main effects showing that
both violent video game use and trait irritability facilitated aggressive response. All of
these results are consistent with what would be predicted by GAM.

In another study using video games, Bushman and Anderson (2002) tested the
impact of violent video games on the creation of a hostile-expectancy bias, or the
tendency to believe that others will react to potential conflicts in an aggressive manner.
In research asking participants to write the end to incomplete story-stems, they found that
participants playing a violent video game predicted that characters in these story stems
would behave more aggressively in responses to the circumstances presented in the story
than those who played a non-violent game. This is consistent with GAM predictions
suggesting that hostile situational cues (e. g., the video game violence) result in
differential outcomes (e.g., a hostile expectancy bias).

These studies on film and video games show that GAM can be an effective model
to help explain how some forms of media violence impact aggression. While it is
reasonable to expect that other forms of media violence should follow similar patterns, it

is possible that additional features of these media violence forms might moderate the

manner in which they work within the model. One feature of violence closely studied in
media studies but not yet considered by GAM researchers is the fictional nature of media
violence. Of particular interest in this study is the fictional or realistic representation of
violence in mediated sports, a feature that Depalma and Raney (2002) distinguish as
scripted or unscripted sports violence.
Research on Media Violence

Questions concerning the effect of exposure to violent media on subsequent
aggression are as old as violent media. While historically this issue has been the concern
of social critics, over the last fifty years, considerable empirical research has focused on
the outcomes from exposure to television and other forms of media violence. Although
the effects of media violence on aggression may still be debated in the public arena,
scientific evidence clearly shows that the two are positively linked. Meta-analysis of
scientific studies on media-related aggression shows that the average correlation between
violent media and aggressive behavior is around .20 (Bushman & Anderson, 2001 a).
Evidence shows violent media related to aggression across research using multiple
methods of measurement, across longitudinal and cross-sectional studies, and across field
and lab settings. Subsequent meta-analyses of all these types of research show a positive
relationship between exposure to violent media and subsequent aggression (Anderson &
Bushman, 2002b).

Although the research included in meta-analytic research contains over two
hundred different studies involving over 50,000 participants (Anderson & Bushman,
2002b), few studies have looked at the effect of televised sports violent on subsequent

aggression. The little research that does exist suggests that violent sports, like other

violent media, lead to increased aggression; however, the exact nature of that relationship
in terms of GAM processes remains decidedly unclear.
Televised Sports and Aggression

Given the great popularity of televised sports, it is surprising that more research
has not focused on the effect of viewing mediated sports violence. This popularity is
most obvious on cable television, where several cable networks are dedicated solely to
providing 24 hour-a-day coverage of sports programming (e.g., ESPN, ESPN2, Fox
Sports, etc.), but the popularity of mediated sports is not just a cable phenomenon. In
1998, over one-third of all programming on broadcast network television was devoted to
sports (Nielsen Media Research, 1998). A common week of Nielsen TV ratings today
shows broadcast sports ofien earning ratings in the two’s and three’s, signifying that
millions of people are watching (“Nielsen Sports TV Ratings”, 2003; “Top 10 US
Television Sports Programs”, 2003). In all, sports appear ubiquitous on television.
People continue to watch and fuel continued broadcasting. The sheer amount of exposure
demonstrates the need to learn more about the impact of watching televised sports.

The few studies that examine the effects viewing violent sports suggest that
exposure impacts subsequent aggression. For example, experimental studies have found
that viewing a violent sports clip can increase aggressive mood (Celozzi, Kazelskis, &
Gutsch, 1981; Russell, Di Lullo, & DiLullo, 1988) and aggressive behavior among
angered participants (Russell, Di Lullo, & Di Lullo, 1988). In cross-sectional correlation
research, Brown, Sumner, and Nocera (2002) found that viewing contact sports predicted
men’s sexual aggression toward women. Similarly, using time series analysis, Philips

(1983) linked exposure to televised boxing matches with an increase in national homicide

rates. Thus, existing evidence links aggression with exposure to televised sports
violence. Yet the research in this area is scattered and not based on any coherent model.
As such, little is known about the processes that govern the link between televised sports
violence and outcomes that should be expected. The extension of GAM to this area of
media violence is one way to advance understanding here.

Logic suggests that the processes governing exposure to other forms of media
violence are related to those governing televised sports violence. As such, outcomes
from exposure to televised sports violence should mirror those found with other research
on the GAM. For example, research on GAM using video games as a situational variable
found that violent games increased the hostile expectancy bias found in users (Anderson
& Bushman, 2002). If I extend the GAM to sports, I might expect televised sports
violence to act as a situational variable that produces a hostile expectancy bias.

Based on GAM and the findings of previous research on exposure to violent
sports, the first hypothesis of this paper is offered:

H1: Exposure to a violent sports clip will lead to a greater hostile expectancy bias

than exposure to a non-violent sports clip.

“Scripted " vs. “Unscripted ” Sports. Depalma and Raney (2002) note that the
terms “sports violence” and “aggressive sports” are catch-all phrases that fail to
distinguish violence not only in terms of the level portrayed, but also in terms of several
important contextual features known to play a part in moderating the effect of exposure.
While contextual features like the motive for violent acts, their associated rewards, the
graphicness of their portrayal, and their realism are generally held to heighten exposure’s

effect (Wilson, Kunkel, Linz, Potter, Donnerstein, Smith, et al., 1997), research on sports

violence fails to consider these features. Depalma and Raney (2002) herald the need for a
more sophisticated definition of mediated sports violence that considers the type of
contextual features known to moderate its impact. In particular they discuss the realism of
sports violence, a feature they distinguish along lines associated with the “scripted” or
“unscripted” portrayal of violence in sporting events. Based on evidence showing that
presentational features of violence related to its characterization as real versus fictional
can alter its impact on subsequent aggression (Potter, 1997, 1999), they argue that the
scripted nature of sports violence is an important moderating feature governing the
impact of its exposure. In research comparing the enjoyment of watching “scripted”
(professional wrestling) and “unscripted” (boxing) sports, they found that enjoyment and
mood were influenced by this contextual feature. People watching a scripted sports clip
enjoyed it less and were in a worse mood after exposure, especially females.

While Depalma and Raney (2002) make no claims concerning the manner in
which the scripted or unscripted characteristics of violent sports relate to aggressive
outcomes, it seems logical to assume that they might. In arguing that realism and other
contextual features are as relevant to outcomes from sports violence as they are to
outcomes from media violence in general, Depalma and Raney (2002) lay the foundation
for extending the logic associated with these moderators to similar contextual features of
violent sports media associated with its scripted or unscripted nature. One might argue
that the unscripted nature of some sport violence (or violence in general) might account
for increased aggression known to result from realism in portrayals of media violence. If

research on scripting was found consistent with earlier studies on realism, it might

suggest that the unscripted (and, thus somewhat less predictable) nature of realistic
violence in media is responsible, at least in part, for its increased impact on aggression.

In order to better understand this possibility, clear explication of what is meant by
“scripted” and “unscripted” violence in sports is necessary. Depalma and Raney (2002)
formally define the two in a simple form as follows: “unscripted sports aggression refers
to presentations of violence that naturally occur in the course of the traditional aggressive
sports” (p. 5), while “scripted sports aggression, on the other hand, will refer to the
violent depictions found in the increasingly popular sports entertainment arena” (p. 5). In
their discussion of these differences, however, they distinguish between the two by
considering the utility of violence in different sports. They argue that violence and
aggression are encouraged in some sports while they are prohibited in others, and that
this differentiates the role of violence in various sports. Unscripted sports like football,
boxing, and hockey encourage aggression, and understand itas part of an “instinctual
competitive drive” necessary for the participant to perform his or her best. By contrast,
instead of encouraging aggression, scripted sports like inline roller derby, Slamball, and
professional wrestling mandate the use of violence. Here, the use of “excessive force or
physical harm” is portrayed as a means to a successful end (i.e., winning or entertaining)
by intentionally violating the rules. Though not stated directly, I might infer from this that
the different roles for violence ascribed to scripted versus unscripted sports (i.e.,
instinctive competitive drive versus intentional exploitation of force) are typically (if not
naturally) confounded with fictional and realistic forms of sport violence.

The present study focuses on scripted versus unscripted violence based not only

on the logic extending the notion of realism, but also on empirical evidence by Depalma

and Raney (2002) showing the usefulness of this distinction in predicting affective
outcomes form exposure to mediated sports violence. Consistent with the meanings
represented in Depalma and Raney (2002) I define scripted sports violence in terms of
function and intentionality.

For the present study, “scripted” sports violence will be defined as excessively
forceful movements and actions choreographed ahead of time and intentionally
performed in sporting events for the purpose of entertainment. In other words, the exact
movements of the competing athletes are practiced before the actual competition, and the
intentional exploitation of force is included in performance primarily for the approval of
spectators. “Non-scripted” sports violence is defined as the instinctive use of excessively
forceful movements and actions in a sporting event performed in an attempt to win the
event or as a reaction to unplanned instigating actions occurring during the event. While
some of these actions might have been practiced, the exact movements between
combatants are not choreographed ahead of time. Notably missing from this definition is
any requirement that violence must be represented as an intentional violation of rules
performed to win the sporting event. While I do not doubt that this is a common feature
in scripted sports violence, I consider scripted violence to include any rehearsed violence
that intentionally exploits force for the approval of spectators.

Based on these definitions of “scripted” and “unscripted,” the issue of how
scripted violence moderates the impact of televised sport on the hostile expectancy bias is
investigated. While the logic governing hypotheses in this investigation might suggest
certain outcomes, conflicting findings in relevant literature make it difficult to offer

specific hypotheses. If I equate “unscripted” sports violence with “realistic” violence,

10

evidence from research on exposure to realistic violence suggests that realism (i.e.,
unscripted violence) increases aggressive responses (Potter, 1997, 1999). However, in
research comparing the enjoyment of exposure to “scripted” (professional wrestling) and
“unscripted” (boxing) sports, Depalma and Raney (2002) found that people who viewed
an unscripted violent sport exhibited less hostile responses. Specifically, they enjoyed it
more and were in a better mood after exposure. Given findings that are seemingly
inconsistent, research questions are appropriate for investigating the manner in which the
scripted nature of sports is related to aggressive outcomes. The following research
question is offered in this regard:

RQl: Will differences in “scripted” vs. “unscripted” features of violence

moderate the impact of sport violence on levels of hostile expectancy bias?

Method

Overview

An experiment was designed with the intent of replicating and extending
Bushman and Anderson’s (2002) procedures for studying the impact of exposure to
violent media on hostile expectancy bias. Their protocol using violent video games was
applied to the study of violent television sports. Materials used to manipulate exposure to
televised sports violence were pilot tested to for levels of violence and scriptedness. In
the main experiment, respondents in a 2 X 2 factorial design watched one of four video
clips varying sports violence (violent, non-violent) and scriptedness (scripted, non-
scripted). The impact of several contextual factors including trait aggression and
individual differences in the enjoyment of sports was measured in order to control for

extraneous influence.

ll

Participants

A sample of 139 undergraduate students enrolled in an introductory
communication course at the same university took part in the experiment. The age of
participants ranged from 18-40, with a mean age of 20.5. Of the 139, 70 (50.4%) were
male, and 69 (49.6%) were female. The racial breakdown of the sample is as follows:
Caucasian, 65.5%, African-American, 15.8%, Asian-American, 8.6%, Hispanic, 2.9%,
Mixed, 2.2%, Other, 2.2%, Native American, 0.7%, and 2.2% did not report their race.
Students received course credit in exchange for their voluntary participation.
Procedure

Participants were tested in groups of five to ten. They were told that they were
taking part in an experiment to determine peoples’ reactions to watching sports
programming. Before beginning the study, participants were asked to give their consent
by reading and signing an informed consent form (see Appendix A). After providing their
consent, participants were randomly assigned to view a sports video clip that represented
either the non-violent/non-scripted condition (baseball, N=31), the non-violent/scripted
condition (pair’s figure skating, N=34), the violent/non-scripted condition (boxing,
N=35), or the violent/scripted condition (professional wrestling, N=39). The
experimenter sat in the experimental room the whole time while the clip was showing
providing a presence intended only to keep discussion during the clip to a minimum.

After watching the clip, participants completed three ambiguous story stems, used
previously by Bushman and Anderson (2002). This served as the primary outcome
measure in this study. Next they evaluated the video clip they saw on measures of

enjoyment, arousal, perceived scriptedness and boredom used to control for effects due to

12

unintended differences in the experimental materials. Next, each participant answered a
group of media use questions. Finally, they completed the Buss Perry Aggression
Questionnaire (Buss & Perry, 1992), the Eysenck Personality Questionnaire-Revised
(Eysenck, Eysenck, & Barrett, 1985) and several demographic measures used to control
for extraneous variance due to individual differences. After filling out these measures,
participants were debriefed (see Appendix B). For a more complete version of the
experimental script, please see Appendix 1.
Materials

Baseball, pair’s figure skating, and professional wrestling were chosen to
replicate the operational procedures used to define these three conditions in earlier
research on scripted and violent sports (DePaIma & Raney, 2002). The fourth condition,
non-violent/scripted condition was added in order to fully cross scriptedness with
violence. Pair’s figure skating was selected to represent this condition because it was a
non-violent and scripted (choreographed) sport that involved multiple participants, and
thus matched the other conditions on this multiple-participant attribute. Each sports video
clip was about twenty-minutes long without commercials. Several separate short clips
from each genre were combined to create each 20-minute stimulus. The pro wrestling
clips came from episodes of Monday Night Raw and Smackdown. The figure skating
clips came from the Four Continents Championship and the 2002 Winter Olympics. The
boxing clips come from recent bouts on Fox Sports. The baseball clips come from recent
Major League Baseball games.

Before the main study was conducted, a pilot study was conduced to guarantee

that the experimental materials varied on violence and scriptedness as intended. First, the

13

four video clips were checked for the number of violent acts each clip contained. The
NTVS (Wilson etal., 1997) definition of a violent act was used for this purpose. A coder
familiar with the definition watched each video and coded for the number of violent acts.
The baseball and figure skating clip each contained no violent acts, the professional
wrestling video contained 199 violent acts, and the boxing clip contained 522 violent
acts. Chi-square tests demonstrate that there were significantly more aggressive acts in
the violent/non-scripted condition than in the violent/scripted condition, x2 (1,
N=721)=144.7, p<.001.

Second, the video clips were pilot tested to make sure they varied on scriptedness
as intended and to check for unintended differences on enjoyment, arousal, and boredom.
The pretest measures for scriptedness as well as those for arousal and boredom were
created specifically for this study. These measures appear in Appendix C. Similarity
across levels of enjoyment was measured using a scale adapted from previous
entertainment research (Depalma & Raney, 2002). This scale also appears in Appendix
C.

A sample of 61 undergraduates from a large Midwestern University watched one
of four video clips selected to represent the four conditions varying sports violence and
scriptedness, and rated them on measures of scriptedness as well as on enjoyment,
arousal, and boredom. None of the participants in the pilot sample took part in the main
experiment. Of the 61 pilot participants, 18 viewed a baseball video clip (the non-
violent/non-scripted condition); 14 viewed figure skating (the non-violent/scripted
condition); 15 viewed boxing (the violent/non-scripted condition); and 14 viewed

professional wrestling (the violent/scripted condition).

14

A manipulation check performed using one-way between subjects AN OVA on
the measure of scriptedness produced a significant effect, F (3,57) = 86.92, p<.001.
Subsequent Newman-Keuls analyses provided evidence that the clips varied as intended.
Both professional wrestling (M= 4.38, SD=1.13) and figure skating (M=4.86, SD=.36)
were perceived as more scripted than boxing (M=1.80, SD=.60) or baseball (M=1.41,
SD=.67) at p<.05.

Additional AN OVA’s on enjoyment, arousal, and boredom produced significant
differences for enjoyment, F(3,57) = 7.80, p<.001, arousal, F(3,57) = 8.57, p<.001, and
boredom, F (3,57) = 11.02, p<.001 (for means and standard deviations of these and other
variables used in this study, please see Appendix L). Subsequent Newman-Keuls
analyses demonstrated that boxing, professional wrestling and figure skating were
perceived as more enjoyable, more arousing, and less boring than baseball at p<.05. Due
to these observed differences, enjoyment, arousal, and boredom were included as control
variables in the main study.

Outcome Measures.

The hostile expectancy bias outcome response was measured using procedures
established by Bushman and Anderson (2002). Participants were provided with a set of
three incomplete story stems about a car accident, persuading a friend, and going to a
restaurant. Two matching versions of the story stems were created with either male or
female characters. This was done to control for possible gender effects in the stories (see
Appendix F). Participants were asked to read the story stem and then list up to 20
different things that they think the person in the story would do or say, think or feel.

Seventy-two participants completed form A, and 67 completed form B.

15

Trained research assistants coded participant responses to the story stems for the
number of total responses and the number of hostile responses. Using Bushman and
Anderson’s (2002) protocol, they independently tabulated the number of aggressive
thoughts, behaviors and feelings subjects reported when completing the story stem (see
Appendix G). Again, following Bushman and Anderson’s procedures, the percentage of
hostile response was computed for each participant to control for individual differences in
the total number of responses provided by participants. Percentage of hostile response
scores served as the outcome measure in all analyses.

Two steps were taken to examine the quality of the outcome measure. First,
responses from the two coders were inspected to establish intercoder reliability. Each
coder was given the story stems for the same 14 (out of 139) participants. The two
coders had a 99.5% agreement rate on what was a response, coming up with a total of
440 responses. They had a 91.8% agreement rate on whether a response was hostile or
not. Based on this evidence of high intercoder agreement, one coder was used to code the
entire sample. Second, a one-way ANOVA was conducted to identify unintended
differences in hostile response outcomes associated with forms A and B. Since there was
no significant difference between forms A and B, F (1 ,137)=.169, ns, the two were
combined in subsequent analyses.

Control Measures

The Revised Eysenck Personality Questionnaire (REPQ) was used to help control
for the affects of personality (Eysenck, Eysenck & Barrett, 1985). The scale contains 48
forced-choice items with 12 statements measuring each of four personality "types"

including extroversion, neuroticism, psychoticism, and social desirability. Participants

16

respond to each item in a yes or no manner. Scores are a count of the number of
statements for which one chooses the designated response. The scale has proven reliable
in several studies (e. g., Eysenck et al., 1985) and been demonstrated as a comparable
alternative to the long form (Barrett & Eysenck, 1992). (see Appendix E).

Extroversion (e. g., “Do you like mixing with people?”) is said to tap an
individual’s level of social adaptability. Neuroticism (e.g., “Do you suffer from
“nerves?”) involves an individual’s level of anxiety, emotionality, and social isolation.
Psychoticism items (e. g., “Have you ever cheated at a game?”) assay an individual's
inclination toward a "lack of restraint, responsibility, need for cognitive structure, and
willingness to live by society's rules and mores (socialization)" (Zuckerman, Kuhlman, &
Camac, 1988, p.104). Social desirability (e. g., “Are all your habits good and desirable
ones?”) measures an individual’s likelihood to participate in acts that may be deemed
socially inappropriate. Reliabilities of these subscales in this study were coefficient alpha
equal .84, .85, .45, and .62, respectively.

The Buss-Perry Aggression Questionnaire (BPAQ) (Buss & Perry, 1992) was
used to measure the effects of trait aggression. This questionnaire contains 29 standard
Likert-type items that measure four subtraits of aggression: physical aggression (e.g., If
somebody hits me, I hit back), verbal aggression (e.g., I can’t help getting into arguments
when people disagree with me), anger (Some of my fiiends think I’m a hothead), and
hostility (At times I feel I have gotten a raw deal out of life). Buss and Perry (1992)
found the BPAQ to have a coefficient alpha of .89 and a test-retest reliability of .80.
They also demonstrated a significant relationship between peer reports of aggressiveness

and higher scores on these four scales for male college students (see Appendix H).

17

Reliabilities for these subscales in this study were alpha equal .85, .76, .78, and .70,
respectively

The demographic and media use measures included single item measures of age,
sex, race and enjoyment of 12 different sports including soccer, basketball, baseball, golf,
auto racing, hockey, football, figure skating, roller derby, boxing, tennis, and professional
wrestling. Additional items measured the number of hours per week spent watching
sports in general, watching each of the 12 sports specified above, visiting intemet sports
cites, listening to sports radio, reading sports books and magazines, and reading the sports
pages in the newspaper. Acceptable reliability levels were not found for items
measuring enjoyment of different sports, or for the enjoyment of different violent sports
despite several attempts at dropping various combinations of items. Therefore, these
items were not used in further analyses. The items measuring general sports media
consumption (items 2, 4, 5, 6, 7) had a reliability of alpha equal .85, and were summed
across to create a new variable called “sports media consumption” to be used as a control,
in place of enjoyment of sports.

The same items used in the pilot test to measure boredom, arousal and enjoyment
resulting from exposure to the video clips were included as measures to control for
effects due to unintended differences in the experimental materials. Boredom (e.g., “The
footage was boring”) and arousal (e. g., “What I saw was exhilarating”) were each
measured using three five-point Likert-type response items. Enjoyment (e.g., “How
much would you like to see the entire clip?”) was measured with ten seven-point Likert-
type response items. Reliabilities for these measures were .92, .90, and .94, respectively.

Items for these scales are reported in Appendix C.

18

Results

In order to test the hypothesis and research question under consideration, both the
total number of hostile responses and the percentage of hostile response scores were
subjected to a 2 (violence/nonviolence) x 2 (scripted/unscripted) ANOVA. An additional
analytical procedure known as “destructive testing” (Anderson & Anderson, 1996) was
conducted also to test the robustness of the observed relationships.

The Main Effect of Sports Violence

Hypothesis one suggested that a greater hostile expectancy bias would be found
following exposure to a violent sports clip than exposure to a non-violent sports clip. In
line with this prediction, the 2 X 2 AN OVA on total hostile responses produced a
significant main effect of violence, F (1,138) = 4.08, p<.01, eta squared=.03‘.
Participants watching a violent sport reported a greater total number of hostile responses
(M=6.97, SD=4.20) than those watching a non-violent sport (M=5.57, SD=3.64). Thus,
the data is consistent with the hypothesis that watching a violent sport would lead to a
greater hostile expectancy bias than watching a non-violent sport.

To help control for the effects of number of total responses (both hostile and non-
hostile), another 2 X 2 ANOVA was run with percentage of hostile responses as the
dependent variable. The 2 X 2 ANOVA on percentage of hostile response scores
produced a significant main effect of violence, F(1,138) = 7.31, p<.01, eta squared = .05.
Participants watching a violent sport reported a greater percentage of hostile responses
(M=.23, SD=.13) than those watching a non-violent sport (M=.18, SD=.09). Thus, the
data are again consistent with the hypothesis that watching a violent sport would lead to a

greater hostile expectancy bias than watching a non-violent sport.

l9

After demonstrating the impact of the sports violence manipulation, the
robustness of the main effect was examined using the destructive testing approach
(Anderson & Anderson, 1996). In this approach, the predicted relationship is first tested.
If it is found to exist, control variables are entered into a regression model to see if the
relationship is broken. The goal of this test is to identify the strength of the relationship
by determining how many control variables it can withstand before it breaks.

Toward this end, a block linear regression was performed on total hostile
responses and then percentage of hostile responses adding a new set of control variables
with each block until the relationship was no longer significant. The order of the blocks
was determined by assumptions about which would be the strongest control variables.
Thus, after violence was entered into the first block, the Buss-Perry subscales and the
Eysenck subscales were entered into the second block, sports media consumption was
entered into the third block, boredom was entered into the fourth block, arousal was
entered into the fifth block, and enjoyment of the clip was entered into the sixth block. A
linear regression was run to see at what point the relationships that were found previously
break. This analysis showed that a significant main effect of violence on total hostile
responses still existed up to the end of the last block, F (13,138)=2.26, p<.05. It also
showed that the significant main effect of percentage of hostile responses still existed up
to the end of the fifth block, F (12,138)=1 .88, p<.05. Thus, the relationship between the
violence of the clip and the number of hostile responses held even when adding all of the
above control variables, and the relationship between violence and percentage of hostile

responses held when adding all of the control variables except enjoyment of the clip.

20

The Moderating Effect of Scriptedness

Research question one asked if the scripted nature of sports would moderate effect
of exposure to a violent sports clip on hostile expectancy bias. In this regard, results of
the 2 X 2 AN OVA on total hostile responses showed no main effect of scriptedness,

F (1,138)=.036, ns. The percentage of hostile response scores also produced no main
effect of scriptedness, F (1,138)=.312, ns. No significant violence X scriptedness
interaction was found for total hostile responses, F(1,l38)=1.97, ns. However, a
significant violence x scriptedness interaction was found for percentage of hostile
responses, F (1,138) = 4.38, p< .05, eta squared = .03. Thus, the data show possible
evidence of a moderator effect.

Subsequent Newman-Keuls test were conducted to inspect the interaction among
the four conditions. These tests showed that percentage of hostile response scores in the
violent/scripted condition were significantly higher than those in the non-violent/scripted
condition. No other differences were observed. (See table 1).

Table 1.

Percentage of Hostile Responses by Condition

 

 

Sciptedness
Violence Non-scripted Scripted
M=.203b M=.17a
Non-violent
SD=.1 1 SD=.07
M=.213b MZ-26b
Violent
SD=. 10 SD=.15

 

 

 

 

Note. Means that do not share the same subscripts differ at p<.05.

21

After examining the specific pattern represented by the interaction, destructive
testing was again applied to test the robustness of the interaction. To test this
relationship, violence, scriptedness and the interaction term for violence X scriptedness
were entered into the first block. Variables entered into subsequent blocks followed the
same sequence as described in tests on hypothesis 1. This analysis showed that a
significant interaction effect on percentage of hostile responses still remained at the end
of all the blocks, F(15,138)=1 .91 , p<.05. Thus, evidence of the moderator effect
associated with the interaction effects was not broken by the destructive tests.

Furthermore, although a significant violence X scriptedness interaction was not
found for total hostile responses, the pattern of means is similar to the pattern for
percentage of hostile responses.

Table 2.

Total Number of Hostile Responses by Condition

 

 

Sciptedness

Violence Non-scripted Scripted
M=6. 13 M=5 .06

Non-violent
SD=4.29 SD=2.89

. M=6.54 M=7.36

Violent

SD=3.70 SD=4.62

 

 

 

 

Note. No means differ at p<.05.

22

Discussion

This study was designed to extend the General Aggression Model developed by
Anderson and his colleagues (e. g., Anderson & Bushman, 2002a) by applying its logic to
a type of content that has received limited attention, sports, and by considering other
variables that may moderate any relationship found. Data were consistent with the first
hypothesis that viewing violent sports would lead to a greater creation of the hostile
expectancy bias among participants than viewing nonviolent sports. They also indicated
that the scripted nature of sports moderated this effect. One interpretation of these data is
that media violence may increase aggressive responses only if the violence is scripted.
Support for this interpretation can be derived from the fact that participants in the
violent/scripted condition had the highest mean percentage of hostile responses. Further
support is found in the fact that the violent/non-scripted condition did not significantly
differ from either non-violent condition. Thus it is possible that the conventionally
accepted inducing effect of violent media on hostile reactions may occur only for
“scripted” violence. Obviously, much more research needs to be done in this area before
we can have the necessary confidence in this claim, but the findings here are consistent
with this type of understanding. At the same time, the data are also consistent with the
notion that scripted/non-violence actually produces less hostility, but the lack of a logical
explanation for this interpretation makes it much less plausible.

This finding suggests something may be different between sports and fictional
media. Although it is generally thought that media violence will increase aggressive
responses resulting from exposure, perhaps this is not always the case in sports. The

findings suggest that in sports media, violence per se may not be enough to increase a

23

hostile expectancy bias. Only when that violence was scripted was a significant increase
found for the hostile expectancy bias. This suggests that sports may be processed and
attended to differently than other forms of violent media. In this study, if the “non-
scripted” boxing is equated with “realistic”, the findings are in contrast with other
findings from violent media, which generally find that realism increases aggressive
outcomes (Potter, 1997, 1999). This implies that there may be something about the
nature of sports and “realistic” sport violence that make it different from other forms of
media violence. One possible difference is the mandated nature of the violence. In
sports, a greater proportion of the violence that occurs is sanctioned by rules allowing the
use of violence in pursuit of a goal; winning a game or match. Perhaps the sanction
associated with “realistic” sports violence also acts to instill clear rules in viewers’ minds
concerning the social boundaries that delimit the use of hostility to certain arenas and
prohibit it in others. Though the factors leading to these differences are not clear,
reactions to the forms of scripted violence found in professional wrestling seem different.
Notably, these reactions seem more in line with the effects observed in traditional media
violence research - research usually conducted with the type of narrative fictional
violence that might be called scripted. If this is the case, one might argue that the
findings for wrestling mirror those found in most media entertainment research on
violence and hostility. Under this interpretation the unique observation here is that non-
scripted violence (specifically, mandated non-scripted violence) did not increase hostility.
It is possible that this mandated nature of sports violence somehow causes realistic (or
non-scripted) violence to cause less aggressive responses in viewers than fictitious (or

scripted) violence. However, this is merely speculation, and much more research would

24

be needed in this area before claims can truly be made. Whatever the reasons behind this,
the evidence supports the claim that realism works differentially for sports and other
forms of mediated violence, and should be taken into account in future studies in this
area.

There are several possible limitations to this study. One is the adequacy of the
measurement used. Although most measures used were at least minimally adequate
(alpha>.70), both the psychoticism (alpha=.45) and social desirability (alpha=.62)
subscales of the Eysenck questionnaire fell below this level. Especially troublesome is
the psychoticism scale, which is known to be an important potential control variable for
effects of watching violent media (Zillmann & Weaver, 1997). It is possible that better
controls for these individual differences would have produced a different outcome in the
destructive testing.

Another limitation of this study is the choice of sports for each condition. Only
one sport was included in each of the four conditions. It is possible that each chosen sport
contained aspects beyond those identified and controlled that produced the results found.
For example, although efforts were made to equate clips of the different sports, verbal
aggression is often more pervasive in professional wrestling telecasts than in other sports.
In fact, it would be difficult, if not entirely impossible; to show a representative 20-
minute clip of professional wrestling that did not include verbal aggression. Thus, it is
possible that verbal aggression is responsible for some of the heightened hostility found
in the scripted/violent condition.

Related to this problem are concerns about the unknown impact of natural

confounds associated with different forms of sports and violence. For example, while

25

differences in enjoyment, arousal and boredom across the four sports selected did not
impact the outcomes of the present study; the pilot study showed that the clips differed
considerably along these dimensions. The baseball clip was especially distinct, being far
less enjoyable or arousing and the more boring than the other sports. It is possible that
the baseball clip used here was unusual on these dimensions, but it is also possible that
baseball is just different in these regards. It is a sport with constant breaks and rest times
(unlike the other three used). Future studies in this area should consider using basketball
or some other non-violent, non-scripted sport to see if this makes a difference. The
constant action and motion common in basketball should make it more similar to other
sports used here in terms of the arousal and boredom it creates.

Similar concerns stem from differences in enjoyment of the different clips used in
this study. Low enjoyment of sports clips has been linked with negative states, like bad
moods (DePaIma & Raney, 2002). It is possible that these negative states could have a
different impact on hostile expectancy bias. One problem with enjoyment levels in this
study is the skewed distribution level of enjoyment. The median level of enjoyment
overall is this study was only 2.90 and the mean was only 3.26 (SD=1.41) on a seven-
point scale. This means that in general, the participants in the main study found the clips
they saw less enjoyable than average. No doubt this was, in part, a function of the
research setting. Still, it may be interesting to study the differential effects of watching
enjoyable vs. unenjoyable clips of the same sport.

A conceptual problem with this study and research on sports and violence in
general lies in our limited knowledge of characteristics that distinguish sports violence,

and the lack of broadly shared definitions of attributes that distinguish different types of

26

sports. For example, under the definition of scripted sport used here, both figure skating
and professional wrestling are classified the same. However, there are great differences
between these two sports; besides the verbal aggression mentioned above. Along with
being scripted, professional wrestling is staged with outcomes are predetermined. By
contrast, while figure skating may be choreographed, the outcomes are not
predetermined. Thus, more precise definitions are needed for “scripted” and other
distinguishing attributes of media sport. Future research may also want to consider the
role of scriptedness according to GAM logic by comparing movies and other forms of
“scripted” television violence to sport violence. This could help differentiate between the
scripted and unscripted nature of the violence, and may provide a better test of its effect.
Clarification of this concept and other attributes that distinguish sports violence will not
only help us to better understand aggressive response associated with mediated sports,
but will give us greater insight on factors that govern the impact on viewer hostility that

result from all forms of media violence.

27

Footnotes
lBoth dependent variables (total hostile responses and percentage of hostile
responses) were also entered into a 2 (violence) X 2 (scriptedness) X 2 (gender)
ANOVA. There were no significant main effects or interaction effects involving gender
for total hostile responses or percentage of hostile responses. Thus, gender was not used

as an independent variable.

28

Appendix A

Evaluating Televised Sports
CONSENT FORM

You are being asked to participate in a study conducted by the Department of
Communication Michigan State University. This research is designed to help determine
how you react to watching televised sporting events. You will watch a 20 minute clip of
a televised sporting event and then be asked to fill answer some questions. The total time
this should take is around 45 minutes.

Your participation in this study is completely voluntary. You may refuse to participate in
any procedures or answer any questions you object to; you may also discontinue your
participation at any time without penalty. Your privacy in this research will be protected
to the maximum extent allowable by law.

If you have any questions or concerns about this study, please contact:

David Westerman
Primary Investigator
355-5162

westemrflgmsucdu

 

If you have any questions regarding your rights as a study participant, or are dissatisfied
at any time with any aspect of this study, you may contact-anonymously, if you wish:

Ashir Kumar, M.D.

Chair of the University Committee on Research Involving Human Subjects
(517)-355-2180

uclrrisfigmsucdu

202 Olds Hall

East Lansing, MI 48824

 

Your signature below indicates your voluntary agreement to participate in this study.

 

Participant‘s Signature Date

 

Participant‘s Name (Please print clearly)

Thank You

29

Appendix B
Debriefing

After completing the final portion of the study, subjects will be verbally debriefed
by the experimenter, who will read the following statement:

“Now that you have finished that portion, I have one more question to ask you. At any
time during the study did you think you knew what the researcher was trying to test?”

The experimenter will give the subject a chance to respond, and then continue by saying:

“Thank you for participating in this study. Now before you go, I have one more thing
that I need to tell you. In addition to what I have already told you about the purpose of
this study, I would like you to know more about our interest in peoples’ reactions to
watching sports. We are specifically interested in knowing whether or not viewing
different types of sporting events creates a hostile expectation bias in viewers. As such,
we are looking to see if the manner in which people complete the incomplete story stem
is influenced by watching different sporting events. We are sorry if this incomplete
disclosure upsets you in any way. We would like to remind you that your name will not
be connected with your responses, however, if you are disturbed by any of what has
happened I will dispose of your responses now and it will not be included in any further
part of our study.”

The experimenter will give the subject a chance to respond, comply with the subject’s
request, and then continue by saying:

“Do you have any further questions about this study?”

The experimenter will answer any questions the subject may have, and then finish by
saying:

“I would like to once again thank you for participating in this study. We ask that you do
not discuss the details of this study with anyone other than me and the primary
investigator. If you have any questions or concerns in the future, feel free to contact
Dave Westerman, the primary investigator. His e-mail address is westemi4@msu.edu.
Now, unless you have anything further, we are finished. Thank you for your time and
have a great day.”

30

Appendix C
Pilot-Test Items for Stimulus Material

To pretest the stimulus materials, the following scale was used to measure
enjoyment (Depalma & Raney, 2003). This was filled out by participants after
completing the incomplete story stems. Each item uses a 7-point scale from 1 (“not at
all”) to 7 (“extremely”).

1. How exciting did you think the clip was?

2. How well-played was the action?

3. How high did you find the energy level of the clip?
4. How good was the footage?

5. How enjoyable did you think the commentating was?
6. How much did you like watching?

7. How entertaining did you find the clip?

8. How much would you like to see the entire clip?

9. How suspenseful was the clip you watched?

10. How much did you enjoy what you watched?

The following questions were also asked as part of the pilot test, as well as after
the enjoyment questions in the major study. The first three questions measured arousal,
the next three measure scriptedness, and the last three measure boredom. All nine were
answered according to the following scale:

l--strongly disagree

2--disagree

3—neither agree nor disagree

4—agree

5—strongl y agree

1. What I watched was exhilarating.

2 The clip I saw was exciting.

3 The action I watched was energizing.

4. It seemed like the actions were pre-planned.

5. I think what I saw was choreographed.

6 The action is this clip was probably rehearsed ahead of time
7 The clip I saw was interesting.

8 What I saw held my attention.

9 The footage was boring.

31

Appendix D

Media use and Demographic Items
1. How old are you?
2. Please circle your gender: Male Female
3. Please circle the choice that best describes your race:
Caucasian African-American Asian-American Hispanic Native American
Mixed Other
1. On a scale from one to ten, where one means you do not enjoy at all and
ten means you enjoy as much as you possibly could, how much do you
enjoy each of the following sports?

 

a. Soccer g. Football

b. Basketball h. Figure Skating

0. Baseball i. Roller Derby

(1. Golf j. Boxing

e. Auto Racing k. Tennis

f. Hockey 1. Professional Wrestling

For questions 2 through 7, please refer to the following scale

0—not at all
l—about 1 hour
2—about 2 hours
3—about 3 hours
4—about 4 hours
5—about 5 hours
6—more than 5 hours

2. On average, how many hours a week do you watch sports?
3. On average, how many hours a week do you watch each of the following

sports?

a. Soccer g. Football

b. Basketball h. Figure Skating

c. Baseball i. Roller Derby

(1. Golf j. Boxing

6. Auto Racing k. Tennis

f. Hockey 1. Professional Wrestling

4. On average, how many hours a week do you spend on the intemet, looking
at sports related sites?

5. On average, how many hours a week do you devote to reading books or
magazines about sports?

6. On average, how many hours a week do you spend reading the sports page
of the newspaper?

32

7. On average, how many hours a week do you listen to sports-related
programming on the radio?

33

Appendix E

The Revised Eysenck Personality Questionnaire

Answer the following questions "yes" or "no". Circle the answer that represents how

you feel.
1. Does your mood often go up and down?
2. Do you take much notice of what people think?
3. Are you a talkative person?
4. If you say you will do something, do you always
keep your promise no matter how inconvenient it might be?
5. Do you ever feel “just miserable” for no reason?
6. Would being in debt worry you?
7. Are you rather lively?
8. Were you ever greedy by helping yourself
to more than your share of anything?
9. Are you an irritable person?
10. Would you take drugs that may have strange or dangerous effects?
1 1. Do you enjoy meeting new people?
12. Have you ever blamed someone for doing something
you knew was really your fault?
13. Are your feelings easily hurt?
14. Do you prefer to go your own way rather than act by the rules?
15. Can you usually let yourself go and enjoy yourself at a lively party?
16. Are all your habits good and desirable ones?
17. Do you often feel “fed-up”?
18. Do good manners and cleanliness matter much to you?
19. Do you usually take the initiative in making new friends?
20. Have you ever taken anything (even a pin or button)
that belonged to someone else?
21. Would you call yourself a nervous person?
22. Do you think marriage is old-fashioned and should be done away with?
23. Can you easily get some life into a rather dull party?
24. Have you ever broken or lost something belonging to someone else?
25. Are you a worrier?
26. Do you enjoy cooperating with others?
27. Do you tend to keep in the background on social occasions?
28. Does it worry you if you know there are mistakes in your work?
29. Have you ever said anything bad or nasty about anyone?
30. Would you call yourself tense or “highly-strung?”
31. Do you think people spend too much time
safeguarding their future with savings and insurance?
32. Do you like mixing with people?
33. As a child were you ever cheeky (talked back) to your parents?
34. Do you worry too long after an embarrassing experience?
35. Do you try not to be rude to people?
36. Do you like plenty of bustle and excitement around you?
37. Have you ever cheated at a game?
38. Do you suffer from “nerves?”

34

YES
YES
YES

YES
YES
YES
YES

YES
YES
YES
YES

YES
YES
YES
YES
YES
YES
YES
YES

YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES

YES
YES
YES
YES
YES
YES
YES
YES

NO
NO
NO

NO
NO
NO
NO

NO
NO
NO
NO

NO
NO
NO
NO
NO
NO
NO
NO

NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO

NO
NO
NO
NO
NO
NO
NO
NO

39.
40.
41.
42.
43.

45.
46.
47.
48.

Would you like other people to be afraid of you?
Have you ever taken advantage of someone?
Are you mostly quiet when you are with other people?

Do you often feel lonely?

Is it better to follow society’s rules than go your own way?
Do other people think of you as being very lively?

Do you always practice what you preach?

Are you often troubled about feelings of guilt?
Do you sometimes put off until tomorrow what you ought to do today?

Can you get a party going?

35

YES
YES
YES
YES
YES
YES
YES
YES
YES
YES

NO
NO
NO
NO
NO
NO
NO
NO
NO
NO

Appendix F
Story Stems

Instructions: We would like you to do this story completion task. Don’t spend too much
time on this task. For each story, you only need to list 20 things in all, not 20 in each
column. When you are done, please on to the next part of this questionnaire. (There
were two versions of this form used in the study. The changes for form B are in
parentheses).

THE CAR ACCIDENT

Todd (Jane) was on his (her) way home from work one evening when he (she)
had to brake quickly for a yellow light. The person in the car behind him (her) must have
thought Todd (Jane) was going to run the light because he (she) crashed into the back of
Todd’s (Jane’s) car, causing a lot of damage to both vehicles. Fortunately, there were no
injuries. Todd (Jane) got out of his (her) car and surveyed the damage. He (She) then
walked over to the other car,

What happens next? List 20 things that Todd (Jane) will do or say, think, and feel
as the story continues.

PERSUADING A FRIEND

Janet (Mark) has worked all summer long, and now, a couple of weeks before
school started, she (he) felt she (he) deserved a holiday. After a bit of thought, she (he)
decided a vacation to the coast would be ideal. After all, what could be better than sun
tanning and swimming in the ocean? The problem was that she (he) did not want to go
alone. She (He) knew her (his) best friend Shannon (Shane) would go if she (he) could
but Shannon (Shane) had been saving her (his) money to buy a new stereo. Janet (Mark)
decided to go over to Shannon’s (Shane’s) place and try to convince her (him) to come
to the coast.

What happens next? List 20 things that Janet (Mark) will do or say, think, and
feel as the story continues.

GOING TO A RESTAURANT

Jane (Todd) had worked hard all day long cleaning her (his) apartment. She (He)
was tired but decided to reward herself (himself) with a meal in one of the restaurants
down the street. Upon entering the restaurant, Jane (Todd) decided upon a Caesar salad,
French onion soup, and filet mignon. Some 15 minutes later, a waiter (waitress) came
around to take her (his) order. Time slowly passed and Jane (Todd) was getting hungrier
and hungrier. Finally, about 45 minutes after her (his) order had been taken, Jane (Todd)
was about to leave when she (he) saw the waiter (waitress) approaching with her (his)
food.

What happens next? List 20 things that Jane (Todd) will do or say, think, and feel
as the story continues.

36

Response Sheet for Story Stems

Remember, you only need 20 total actions, thoughts or feelings for each story. What will
the main character:

Do or Say

Think

Feel

37

Appendix G
Coding Scheme

For the purposes of this study, an aggressive response will be regarded as any response
having a hostile, injurious, or destructive connotation, as defined by Bushman and
Anderson (2002), and as told to the secondary investigator in personal communication
with Bushman (2003). In order to train coders, the secondary investigator will explain
the definition of an aggressive response to the coders. Then, sample responses created by
the secondary investigator will be used to help train the coders to find aggressive
responses. When the secondary investigator feels that the coders have been properly
trained, they will each code a sample group of responses independently, and inter-coder
reliability will be determined for this sample. Finally, if inter-coder reliability is
acceptable with this sample, each coder will be given ten percent of the real sample to
code independently and inter-coder reliability will be determined for this sample. Again,
if inter-coder reliability is acceptable, then the rest of the sample will be randomly
assigned to the two coders to be coded for aggressive responses. Coders will be blind to
the condition from which each set of responses belongs.

38

Appendix H
BUSS PERRY AGGRESSION SCALE

The following questions deal with your beliefs about the way the world works. Using the
scale that follows, please circle the number that best describes how you feel about each
statement.

1 = Extremely moharacteristic of Me
2 = Somewhat U_ncharacteristic of Me
3 = Neutral

4 = Somewhat Characteristic of Me

5 = Extremely Characteristic of Me

1) When people are especially nice,
I wonder what they want.
1 2 3 4 5

2) Given enough provocation,
I may hit another person.
1 2 3 4 5

3) I flare up quickly but get over it quickly.

1 2 3 4 5
4) My friends say that I'm
somewhat argumentative.
1 2 3 4 5

5) If somebody hits me, I hit back.

1 2 3 4 5
6) I am suspicious of overly friendly strangers.

1 2 3 4 5
7) I am an even-tempered person.

1 2 3 4 5

8) I can think of no good reason
for ever hitting a person.

1 2 3 4 5

39

9) I sometimes feel like
a powder keg ready to explode.

l

10) I get into fights a little more than
the average person.
I

l l) I often find myself
disagreeing with people.
1

12) I sometimes feel that people are
laughing at me behind my back.

1

13) Other people always seem
to get the breaks.
1

14) Once in a while I can't control
the urge to strike another person.

1

15) When people annoy me, I may
tell them what I think of them.

1
16) There are people who pushed me
so far that we came to blows.

1

17) I wonder why sometimes
I feel so bitter about things.

1
18) When frustrated,
I let my irritation show.
1

19) I am sometimes eaten up
with jealousy.
1

40

20) Some of my friends think
I'm a hothead.
1

21) I tell my friends openly
when I disagree with them.
1

22) I have become so mad that
I have broken things.
1

23) Sometimes I fly off the
handle for no good reason.
1

24) I have threatened
people I know.
1

25) At times I feel I have gotten
a raw deal out of life.

1

26) If I have to resort to violence
to protect my rights, I will.

1

27) I lmow that "friends" talk
about me behind my back.

1

28) I can't help getting into arguments
when people disagree with me.

1
29) I have trouble controlling

my temper.
1

41

Appendix I
Experimental Script

This was the script that was followed for each group when conducting this
research. What the experimenter said is in quotes. What the experimenter did is in
italics.

1.“Hello, everyone. Before we start, I need you all to look over and sign the consent
form”.
The consent form was then handed out and collected.

2.”Now that you all have read over the consent form, you can see this is a study about
your reactions to televised sports. In order to do this, you must first watch a sports clip.
The clip will take about twenty minutes. After that, I have some things for you to
respond to. Those will take about 25 minutes to complete.”

The clip was then started. After it was finished, while handing out the story stems, the
experimenter said:

3.”Ok. Now that you have all seen the clip, I have a few things for you to fill out. The
first is a group of incomplete story stems. What these are. . .there are three of them. . .are
a paragraph setting up a story that is incomplete. It is your job to complete each story by
listing things you think the main character would do or say, think, or feel as the story
continued. Now, don’t take to long on any one of these, because we are only going to
take ten minutes to do all three of them. And one more thing. I know it said to come up
with twenty things. . .if you can’t, that’s ok. Just come up with as many as you can.”
When there was five minutes left and two minutes left, the experimenter let the
participants know. When time was up, while handing out the questions about the clip, the
experimenter said:

4.”Ok. I need you to finish writing whatever phrase you are currently working on. If you
turn those over on your desk. The next thing I have is a group of questions about the clip
itself. Please pay attention to the instructions for each page, as the way to respond
changes fo each. When you are done, if you turn them over so I can tell when everyone
is done, that would be great.”

When everyone had finished, while handing out the rest of the questionnaire, the
experimenter said:

5.”Ok. Now the next part is a group of questions about you. The first set is about your
sports media usage in general, and the second part is about you in general. Again, the
directions change throughout, so please be wary of that. Also again, when you are
finished, if you turn these over on your desk, so I can tell when everyone is done and we
can move on to the next part, that would be great”

Once everyone had finished, the experimenter said:

42

6.”Ok. Now I have a question for you. If you could answer it on the back or the bottom

of the last page, that would be great. The question is’What was the purpose of this
study?’”

After all participants had a chance to write an answer, they were debriefed and allowed to
leave.

43

Appendix J

The General Aggression Model*

A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Inputs I Person Situation
* Social
[Encounter]
Present Internal State: A
Routes ’Afiect‘
Cognition ----- Arousal
I Thoughtful_
Appraisal & Action
Outcomes Decision
Processes Impulsive _____w
Action

 

 

 

*This model was taken from Anderson, CA. & Bushman, B.J. (2002a). Human aggression. Annual Review
of Psychology. 53, 27-51.

44

Appendix K
Sample Hostile Responses

Below is a sample of actual hostile responses from people who had watched one of the
violent sports:

THE CAR ACCIDENT

Do/Say:

-What the hell are you doing?

-Are you OK? Well, my car’s not!

-Physically assault her if she was angry enough.
-Punches out window.

-Administer middle finger.

-Fuck you, buddy!

-Calls police or beat up the person.

-Hey motherfucker, what the fuck?

-I should smash your face in.

Think:

-Shit! Fuck!

-That he wants to hit this guy.

-This person’s stupidity makes me mad.
-What an idiot!

-Asshole.

-I want to hit this person for wrecking my car.
-I am going to kill this person.

-What a bitch!

-This motherfucker is dead.

-I’m a beat his ass to a pulp.

Feel:

-Angl'y

-Irritated

-A desire to punch the lady.
-Mad

-Upset

-Hostile

-Pissed

-Rage

45

PERSUADING A FRIEND

Do/ Say:
-Fuck your stereo.

Think:

-Shane is an idiot.
-Shane is lame.
-What a wimp, he needs a vacation just as much as I do.
-If he doesn’t come with me he’s not a good friend.
-Asshole.

-This guy is an idiot to not want to go.

Feel:
-Angry
-Upset
-Agitated
-Annoyed
-Frustrated

GOING TO A RESTAURANT

Do/Say:

-Yell at the waitress.

-Throw food at her.

-Leave without paying.

-Did you have to kill the cow first?
-Tips her low.

—Curses at whoever is in charge.
-That was ridiculous.

Think:

-This is bullshit, I want a free meal.

-Hurry up woman!

-About time! My stomach has almost eaten itself.
-Ha-you’re getting no tip.

-What an idiot waitress.

-Yell at waiter asking why it took so long.

-I’m never coming back here.

Feel:

-Angry
-Like throwing food in her face.

46

-Outraged
-Frustrated
-Infuriated
-Pissed
Unhappy
-Ticked
-Mad
-Annoyed
-Frustrated

47

Appendix L

Means and Standard Deviations

Pilot Study
Mean Standard Deviation

Enjoyment 3.04 1.54
Arousal 2.34 1.25
Boredom 3 .46 1 .25

Main Study

Mean Standard Deviation

% of hostile responses 21 12
Enjoyment 3.26 1.41
Arousal 2.48 1.05
Boredom 3.27 1.19
Hostility 2.69 .72
Physical Aggression 2.44 .87

Verbal Aggression 3.07 .86
Argumentativeness 2.45 .76
Psychoticism .78 .14
Extraversion .2 1 .24
Neuroticism .55 .31

Social Desirability .75 .18

Sports Media Consumption 1.51 1.26

48

REFERENCES

Anderson, CA. (1997). Effects of violent movies and trait irritiability on hostile feelings
and aggressive thoughts. Aggressive Behavior, 23, 161—178.

Anderson, CA & Anderson, KB. (1996). Violent crime rate studies in philosophical
context: A destructive testing approach to heat and southern culture of violence
effects. Journal of Personality and Social Psychology, 70, 740-756.

Anderson, CA, Anderson, K.B., & Deuser, W.E. (1996). Examining an affective
aggression framework: Weapon and temperature effects on aggressive thoughts,
affect, and attitudes. Personality ans Social Psychology Bulletin, 22, 366-3 76.

Anderson, CA & Bushman, B.J. (20023). Human aggression. Annual Review of
Psychology. 53, 27-51.

Anderson, CA. & Bushman, B.J. (2002b). The effects of media violence of society.
Science, 295, 2377-2379.

Anderson, CA, Deuser, W.E., & DeNerve, K. (1995). Hot temperatures, hostile affect,
hostile cognition, and arousal: Tests of a general model of affective aggression.
Personality and Social Psychology Bulletin, 21 , 434-448.

Anderson, C .A. & Dill, KB. (2000). Video games and aggressive thoughts, feelings, and
behavior in the laboratory and in life. Journal of Personality and Social
Psychology, 78, 772-790.

Anderson, K.B., Anderson, CA, Dill, K.E., & Deuser, W.E. (1998). The interactive
relations between trait hostility, pain, and aggressive thoughts. Aggressive
Behavior, 24, 161-171.

Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of
Psychology, 52, 1-26.

Barrett, P.T., & Eysenck, SB. (1992). Predicting EPQ-R full scale scores from the short
form version. Personality & Individual Diflerences, 13(7), 851-853.

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

Brown, T.J., Sumner, K.E., & Nocera, R. (2002). Understanding sexual aggression

against women: An examination of the role of men’s athletic participation and
related variables. Journal of Interpersonal Violence, 17, 937-952.

49

Bryant, J ., Comisky, P., & Zillmann, D. (1981). The appeal of rough-and-tumble play in
televised professional football. Communication Quarterly, 29, 256-262.

Bryant, J. & Raney, AA. (2000). Sports on the screen. In D. Zillmann & P. Vorderer
(Eds.), Media Entertainment: The psychology of its appeal (pp. 153-174).
Mahwah, NJ: Erlbaum.

Bushman, B.J. & Anderson, CA. (2001 a). Media violence and the American public:
Scientific facts versus media misinformation. American Psychologist, 56, 477-
489.

Bushman, B.J & Anderson, CA. (2001b). Is it time to pull the plug on the hostile versus
instrumental aggression dichotomy? Psychological Review, 108, 273-279.

Bushman, B.J. & Anderson, CA. (2002). Violent video games and hostile expectations:
A test of the general aggression model. Personality and Social Psychology
Bulletin, 12, 1679-1686.

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

Celozzi, M.J., Kazelskis, R., & Gutsch, K.U. (1981). The relationship between viewing
televised violence in ice hockey and subsequent levels of personal aggression.
Journal of Sport Behavior, 4, 157-162.

Depalma, A.J. & Raney, AA. (2002, November). The effect of viewing varying levels of
aggressive sports programming on enjoyment, mood, and perceived violence.
Paper submitted to the Mass Communication Division at the annual convention of
the International Communication Association, San Diego, CA.

Dill, K.E., Anderson, CA, Anderson, K.B., & Deuser, W.E. (1997). Effects of
aggressive personality on social expectations and social perceptions. Journal of
Research in Personality, 31, 272-292.

Eysenck, S.B.G., Eysenck, H.J., & Barrett, P. (1985). A revised version of the
psychoticism scale. Personality and Individual Differences, 6(1), 21-29.

Lindsay, J .J . & Anderson, CA. (2000). From antecedent conditions to violent actions: A
general affective aggression model. Personality and Social Psychology Bulletin,
26, 533-547.

Huesmann, LR. (1998). The role of social information processing and cognitive schema
in the acquisition and maintenance of habitual aggressive behavior. In R. Geen &
E. Donnerstein (Eds), Human aggression: Theories, research and implications
for policy (pp. 73-109). New York: Academic Press.

50

Nielsen Media Research. (1998). 1998 report on television. New York: Author.

Nielsen Sports TV Ratings. (n.d.). Retreived March 25, 2003, from
http://www.top5s.com/tvsports.htm.

Phillips, DP. (1983). The impact of mass media violence on US. homicides. American
Sociological Review, 48, 560-568.

Potter, W. J. (1997). The problem with indexing risk of viewing television aggression.
Critical Studies in Mass Communication, 14, 228-248.

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

Russell, G.W., Di Lullo, G.W., & Di Lullo, D. (1988). Effects of observing competitive
and violent versions of a sport. Current Psychology: Research and Reviews, 7,
312-321.

Schank, R.C. & Abelson, RP. (1977). Scripts, plans, goals and understanding: An
inquiry into human knowledge structures. Hillsdale, NJ: Erlbaum.

Sargent, S.L., Zillmann, D., & Weaver, J .B. (1998). The gender gap in the enjoyment of
televised sports. Journal of Sport and Social Issues, 22, 46-64.

Tedeschi, J .T. & Felson, R,B. (1994). Violence, aggression, and coercive actions.
Washington, DC: American Psychological Association.

T 0p 10 US Television Sports Programs. (2003, March 19). Retrieved March 25, 2003,
from http://www.pazsaz.com/topsports.html.

Wilson, B. J ., Kunkel, D., Linz, D., Potter, W. J ., Donnerstein, E., Smith, S. L.,
Blumenthal, E., & Gray, T. (1997). Violence in television programming overall:
University of California, Santa Barbara study, National Television Violence Study
(Vol. 1, pp. 3-268). Newbury Park, CA: Sage.

Zillmann, D. (1983). Arousal and aggression. In R. Geen & E. Donnerstein (Eds),
Aggression: Theoretical and empirical reviews, Vol. 1. New York: Academic.

Zillmann, D. & Weaver, J .B., 111. (1997). Psychoticism in the effect of prolonged
exposure to gratuitous media violence on the acceptance of violence as a preferred

means of conflict resolution . Personality and Individual Differences, 22(5), 613-
627.

Zuckerman, M., Kuhlman, D.M., & Camac, C. (1988). What lies beyond E and N? Factor

analyses of scales believed to measure basic dimensions of personality. Journal of
Personality and Social Psychology, 54, 96-107.

51

 

ilglijgjijjijijjjil

 

 

ram—‘2‘}. ..__‘