.5;
~3.. .
2.. ;
3 :

.. n
, 3.. 9...:
st?

#Lr ......,
. 1- . .2
It

 

‘11
(‘15.
.V.

 

lull}- . 1|

1 I'IV‘AE ‘ ‘ . l ‘0 ‘

.mnrusu #:tnmmu. r; . . , ....:L..u¥.x~‘;n.....,........n maﬁa , . . a.” ._
. A ‘ .

 

LIBRARY
Michigan State
University

This is to certify that the
dissertation entitled

NEWSPAPER COVERAGE OF COLLATERAL INTIMATE

PhD

 

PARTNER HOMICIDES

presented by

EMILY M. MEYER

has been accepted towards fulﬁllment
of the requirements for the

Comm Arts & Science Media &
Information Studies

degree in

 

7/ ’
//,/ ~/

7'

L Major Profegéor’s Signature

8/20/08

 

Date

MSU is an afﬁnnative-action, equal-opportunity employer

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K.lProj/Acca.Pres/ClRC/DateDue.Indd

NEWSPAPER COVERAGE OF COLLATERAL INTIMATE PARTNER HOMICIDES
By

Emily M. Meyer

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Comm Arts & Science Media & Information Studies

2008

ABSTRACT
NEWSPAPER COVERAGE OF COLLATERAL INTIMATE PARTNER HOMICIDES
By

Emily M. Meyer

The killing of intimate partners is a topic widely explored in the criminal justice and
public health Iiteratures; however there is a dearth of research on mass media reporting
patterns. To this end, this dissertation focuses on collateral intimate partner homicides
(IPH), or the killing of individuals exogenous to the intimate relationship and how these
killings are processed and framed by Michigan newspapers. A mixed-method content
analysis of articles published between 1990-2007 is conducted (n=215) as a ﬁrst step to
determining reporting trends in collateral IPH. The database for this project is drawn
from a collection of newspaper clippings assembled by the Michigan Press Reading
Service and is the most comprehensive in the nation.

This research addresses a critical void in the literature on the use of news frames in
intimate partner homicide reporting. Thus, two IPV theories are tested for their impact
on framing: family violence and coercive control. The quantitative portion of the study
was completed by two independent coders and subjected to a rigorous reliability analysis.
Qualitative data analysis consisted of theoretical, open coding as set forth by Glaser &
Straus (1967) and Altheide (1996). The major research questions driving this study are
(1) “What are the predominant trends in collateral IPH?” and (2) “which theoretical
framework better describes the media narratives utilized to report collateral IPH”?
Results overwhelmingly indicated that a majority of collateral intimate partner homicides

are reported as incident—speciﬁc and attributed to mutual relationship conﬂict between

primary partners (e.g., the two individuals involved in either a current or previous
intimate relationship), with the collateral victim serving as an auxiliary. However,
perhaps the most signiﬁcant ﬁnding is that although IPV research indicates that women
and men both perpetrate IPH, using only the inclusion criterion that collateral IPH is
related to IPV, 100% of homicides were perpetrated by men. Despite this ﬁnding, news
articles maintain a gender-neutral tone, failing to explore relationship context. These
ﬁndings have numerous implications for translational science endeavors, so
recommendations for researchers, journalists, advocates, and policymakers conclude the

study.

DEDICATION
I would like to dedicate this dissertation to my husband, Louis Boudreau, and my three

furry children: Ziggy, Ivy, and Willow. Without their unconditional love and support, I
would not have been able to make it through this challenging phase of my life.

iv

ACKNOWLEDGMENTS

I would ﬁrst like to acknowledge my parents. Without their consistent,
unconditional support, I would have never grown into a woman who felt she was capable
of accomplishing a doctoral degree. Second, my husband, Louis, deserves tremendous
acknowledgment, as he has stood by my side for the past eleven years, quietly and
conﬁdently encouraging me throughout my academic pursuits. He has sacriﬁced so
much so that I could get a PhD; I will never be able to thank him enough.

It is also vital that I acknowledge Dr. Lori Post. Not only did she mentor my
dissertation, she also took me under her wing when I was a timid master’s student who
was ready to leave Michigan State. She convinced me to stay and provided numerous
opportunities that have contributed to the scholar I am today. If it weren’t for her
conﬁdence, tough love, and unwavering faith in me, I would have never given MSU a
second chance.

Lastly, it is important to acknowledge the individuals that helped make my
dissertation a reality. Ellen Hayse, the woman who ﬁrst initiated the clipping service that
provided the data for my dissertation, has been a mentor, conﬁdante, and amazing friend
throughout my PhD training. Also to my coders, Bryce Bernard and Brendan Brown,
two strong men who managed to analyze these articles despite the frustration and
depression that came along with them, thank you for persevering and serving as a

tremendous asset to my success. I couldn’t have done it without you!

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................... VIII
LIST OF FIGURES .................................................................................................... IX
CHAPTER 1: INTRODUCTION ..................................................................................... 1
INTIMATE PARTNER VIOLENCE (IPV) — A BRIEF INTRODUCTION ............................ 2
RATIONALE FOR DISSERTATION ........................................................................ 4
STRUCTURE OF REMAINING CHAPTERS .............................................................. 6
CHAPTER 2: LITERATURE REVIEW AND THEORETICAL FRAMEWORK ........................ 8
JUSTIFICATION OF THIS RESEARCH PROJECT ...................................................... 8
INTRODUCTION TO FRAMING ........................................................................... 8
MEDIA FRAMES ............................................................................................. l 1
Media Frames as a Dependent Variable ......................................................... l l
SELECTION AND CONSTRUCTION OF COLLATERAL IPH FRAMES .......................... l3
NEWSPAPER FRAMING OF SOCIAL PROBLEMS ................................................... 14
Newspaper Framing of Domestic Violence ...................................................... 16
Newspaper Framing of Intimate Partner Homicide ............................................ l 8
Newspaper Framing of Collateral IPH .......................................................... 20
COMPETING H’V THEORIES .............................................................................. 20
Family Violence Theory ............................................................................ 21
The Theory of Coercive Control .................................................................. 24
THEORETICAL CONTRADICTIONS .................................................................... 28
Issues of Sampling & Measurement ............................................................... 31
Theoretical Implications for this Project ......................................................... 34
SUMMARY .................................................................................................... 35
RESEARCH QUESTIONS & HYPOTHESES ............................................................ 36
CHAPTER 3: RESEARCH METHODOLOGY .................................................................. 39
SAMPLE ........................................................................................................ 39
DEFINITIONS ................................................................................................. 4|
Variables ............................................................................................. 43
Criteria for Determining Theoretical Alignment ................................................ 45
CONTENT ANALYSIS PROCEDURES .................................................................. 46
Qualitative Content Analysis ...................................................................... 47
Quantitative Content Analysis ..................................................................... 48
QUANTITATIVE CODING PROTOCOL ................................................................. 49
QUANTITATIVE DATA COLLECTION ................................................................. 5 l
QUALITATIVE DATA COLLECTION ................................................................... 52
ANALYSIS ..................................................................................................... 52
Coder Reliability Analysis ......................................................................... 53
Validity & Reliability of Qualitative Data ....................................................... 54
Qualitative Analysis Procedures .................................................................. 57
STATISTICAL TESTS & PROCEDURES ................................................................ 57
CHAPTER 4: RESULTS .............................................................................................. 61
SAMPLE CHARACTERISTICS ............................................................................ 61
NEWSPAPER INFORMATION AND GENERAL STORY CHARACTERISTICS ................ 61
Unique Collateral IPH Crime Reporting Techniques .......................................... 63

vi

RESEARCH QUESTION #1 ................................................................................. 66

 

RESEARCH QUESTION #2 ................................................................................. 69
Quantitative Results ................................................................................. 70

Qualitative Results .................................................................................. 78

SUMMARY ................................................................................................... l I 8
CHAPTER 5: DISCUSSION ........................................................................................ 120
7 GENERAL DISCUSSION .................................................................................. 120
DOMINANCE OF SYMMETRICAL EXPLANATIONS FOR COLLATERAL IPH .............. 121

DOES THE EXPLANATION FIT THE CRIME? ......................................................................... 124
WHERE DO COLLATERAL VICTIMS FIT? ..................................................... 126

THE AMBIGUOUS NATURE OF COERCIVE CONTROL IN COLLATERAL IPH ............ 128
LIMITATIONS ............................................................................................... 129
FUTURE RESEARCH ....................................................................................... 130
CHAPTER 6: TRANSLATIONAL SCIENCE .................................................................. 133
MASS MEDIA IMPLICATIONS OF COLLATERAL IPH ............................................ 134
RECOMMENDATIONS AND GUIDELINES FOR JOURNALISTS ............................... 135
PUBLIC POLICY IMPLICATIONS ...................................................................... 137
SERVICE PROVIDER IMPLICATIONS ................................................................ 139
CONCLUSION ............................................................................................... 141
APPENDIX A ........................................................................................................... 143
NEWSPAPERS INCLUDED IN DISSERTATION SAMPLE ......................................... 144
APPENDIX B ........................................................................................................... 145
COLLATERAL IPH CODING SHEET ................................................................... 146
APPENDIX C ........................................................................................................... 153
CODER RELIABILITY ANALYSIS ..................................................................... 154
APPENDIX D ........................................................................................................... 156
COLLATERAL IPH QUALITATIVE CODING SHEET .............................................. 157
APPENDIX E ........................................................................................................... 160
EXAMPLE QUALITATIVE ANALYSIS MATRICES ................................................ I61
BIBLIOGRAPHY ...................................................................................................... 162

vii

List of Tables

Table 3.1 — Conceptual & Operational Deﬁnitions ............................................. 42
Table 3.2 — Criteria for Theory Determination .................................................. 46
Table 3.3 — Numbered Variables included in Dissertation ..................................... 58
Table 4.1 — Total Crime Victims .................................................................. 66
Table 4.2 — Collateral Crime Victims ............................................................ 66
Table 4.3 —- Homicide Explanations and Portrayals ............................................. 71
Table 4.4 — Proposed Motivation & Causes of Collateral Homicides ........................ 72
Table 4.5 — Unexpected Nature of the Crime (Variable 35a) and Proposed Conﬂict
(Variable 36a) ....................................................................................... 73
Table 4.6 — Unexpected Nature of the Crime (Variable 35a) and Causes of Violence
(Variable 39) ......................................................................................... 73
Table 4.7 - Intentional Nature of the Homicide (Variable 35b) and Mention of Prior
Abuse Against Principal IPV Victim (Variable 40) ............................................ 74
Table 4. 8— Intentional Nature of the Homicide (Variable 35b) and Explicit Mention of
IPV (Variable 42) .................................................................................... 74
Table 4.9 — The Explicit Mention of IPV (Variable 42) and Prior Abuse Against Principal
IPV Victim (Variable 40) ........................................................................... 75
Table 4.10 - Victim Type Impact on Reporting Homicide as Intentional or Conﬂict-
Oriented ................................................................................ 76
Table 4.11 — Family Violence & Coercive Control Sentences .............................. 108

viii

List of Figures

Figure 3.1 — Variable Diagram .................................................................... 45
Figure 4.1 — Domestic Disputes Category ....................................................... 82
Figure 4.2 — Individual Factors .................................................................... 90
Figure 4.3 — Social Factors ........................................................................ 90

ix

Chapter 1: Introduction

A man angered after a dispute with his wife confessed to tossing his four young
children oﬂ a bridge, authorities said Wednesday as they searched murky waters
for the bodies.

Lam Luong, 3 7, who is charged with four counts of capital murder, told

authorities Tuesday night that he drove to the Dauphin Island bridge and dropped

the children from a span that reaches 80 feet in places, said Detective Scott

Rivera (Associated Press, 2008).

The media report violent homicides every day in the United States. These events,
both shocking and difﬁcult to comprehend, trigger basic emotional reactions in the
public’s consciousness, leaving many asking “why”; why would a human being do that to
another? Why would a father take the lives of his four children? The media, via
explanations offered by journalists and other sources (e.g. police, neighbors, friends, etc.)
attempt to address these questions in various capacities. For instance, in the example
provided, the murder of four children was a result of a marital “dispute” in which his wife
triggered Lam Luong’s subsequent rage. As such, this dissertation is concerned with how
journalists frame these events and what theory of intimate partner violence (IPV)
plimarily informs explanations provided by reporters. This is a salient and necessary
endeavor because it offers potential insight into the dominant beliefs and stereotypes
driving newspaper reporting of collateral IPH. In other words, how is a crime in which a
bond and / or connection between perpetrator and victim exist portrayed? Possible factors
that inﬂuence journalistic framing have been well documented in the literature, including:
(1) social norms and values; (2) organizational pressures and constraints; (3) pressures of

interests groups; (4) journalistic routines; and (5) ideological or political orientations of

journalists (Shoemaker and Reese, 1996; Tuchman, 1978). However, there may be

additional reasons, including ignorance, pressure from the editor, or word space
limitations (Shoemaker & Reese, 1996). Generally speaking, crime reporting relies
heavily on ofﬁcial sources (Ferrand-Bullock, 2008) and is a durable news commodity
(e.g., Garofalo, 1981). Media content “provides a map of the world of criminal events
that differs in many ways from the one provided by ofﬁcial crime statistics” (Sacco,
1995, p. 143), Speciﬁcally the what (cultural context), who (social context), where
(geographical context), and when (the temporal context) of the act. These four classic
contexts also impact the worthiness of crime stories (Koch, 1990). One might also
hypothesize that like any other human being, reporters and their informants are trying to
make sense of the social world around them, and that simple answers exist for complex
crimes. Regardless, there is a disconnect between what IPV researchers are ﬁnding and

what journalists are reporting (6. g. Ferrand-Bullock and Cubert 2002).

Intimate Partner Violence (IP10 - A Brief Introduction

Intimate partner violence (IPV), a concept frequently interchanged with

,9 66

“battering abuse”, “domestic violence”, and “interpersonal violence”, is a serious social
and public health problem in the United States. IPV is deﬁned by the Centers for Disease
Control (CDC) as “threatened, attempted, or completed physical or sexual violence or
emotional abuse by a spouse, former spouse, current or former boyfriend or girlﬁiend or
a dating partner” (Black & Breiding, 2008). This deﬁnition focuses on the existence of

an intimate relationship rather than the current relationship status or living arrangements.

Furthermore, this deﬁnition is gender neutral.

IPV is pervasive in US. society, with signiﬁcant deleterious outcomes. Each year
IPV contributes to 2 million injuries among women and nearly 600,000 injuries among
men (Black & Breiding, 2008). In addition to injury, IPV is frequently associated with
chronic physical and somatic illnesses (Ahmad et al., 2007; Bonami et al., 2006;
Campbell & Soeken, 1999; Demaris & Kaukinen, 2005; Kramer, Lorenzon, & Mueller,
2004; Stover, 2005; Thompson et al., 2006), and mental health conditions (Bogat et al.,
2003; Campbell & Soeken, 1999; Cascardi & O’Leary, 1992; Demaris & Kaukinen,
2005; Houskamp & Foy, 1991; Huth-Bocks, Levendosky, & Bogat, 2002; Kessler et al.,
2001; Vitanza, Vogel, & Marshall, 1995). Women are more likely to suffer abuse than
men, as two to four million women are assaulted by their partners every year, making it
the number one cause of injury for women ages 15-44. Partners and ex-partners murder
at least 1,200 women each year, and experts estimate that at least 5 million children
witness an assault (Bancroft, 2002; Black & Breiding, 2008). In a landmark national
survey of the prevalence, incidence, and consequences of violence against women (the
National Violence Against Women Survey, or NVAW), Tjaden and Thoennes (2000a)
uncovered an alarming rate of IPV across the lifespan (22.1% total) (also see Tjaden &
Thonnes, 2000b). In line with this, the latest survey conducted by the Centers for Disease
Control (CDC) found a lifetime prevalence rate of approximately 25%.

As such, journalists frequently report on victimizations that occur within a
familial or intimate context, including child abuse, sexual assault, and IPV. While
journalists report collateral IPH killings, to date none have been coined as such and no
research exists on this topic. A collateral victim is connected to the IPV victim and/or

perpetrator, but is not a direct recipient of ongoing abuse. The killing of a collateral

victim is usually the result of a triggering event that occurs within the context of the
intimate relationship - either calculated or unplanned. Collateral victims include parents,
siblings, children, and new romantic partners among others. Studying this particular event
is salient because (1) it has never been investigated before in the IPV literature and (2) it
provides media researchers with a more comprehensive understanding of the types of

ﬂames used to report multiple victim IPV.

Rationale for Dissertation

Media scholars have devoted signiﬁcant attention to how social issues and
problems are reported to the public (e. g. Best, 1995; Hartman and Husband, 1974; Horsti,
2003; McNair, 1988), especially within a political context. Likewise, researchers have
debated media impact for nearly a century, ﬂuctuating between the strong effects
hypothesis (i.e., the media have a strong inﬂuence on public opinion) and minimal effects
hypothesis (i.e., the media have little to no impact on the opinions or perceptions of its
audience)(for a brief history, see Scheufele, 1999). However, there is reason to believe
that the media have an effect on public perceptions and attitudes, with some scholars
claiming, “the news media exert signiﬁcant inﬂuence on our perceptions of what are the
most salient issues of the day” (McCombs & Reynolds, 2002, p. 1). Zillmann (2002)
cautions, “However, the capacity to reach large audiences carries with it the risk of
misleading the public in case the disseminated information proves to be distorted and
inaccurate or simply in error'” (p. 21). He goes on, “Media institutions committed to

providing veridical accounts of phenomena of consequence thus should take some

 

I . .
EmphaSIs IS author’s own

responsibility for their case aggregations, ensuring that the reported cases yield correct
rather than distorted perceptions of the phenomena” (p. 21).

Despite a lack of agreement in the weak-strong media effects argument,
journalists play a unique role in the information dissemination process since they act as
primary gatekeepers, presenting news messages that have potential to affect public
perceptions, attitudes, and behaviors (McCombs & Reynolds, 2002). Therefore, how
newsmakers ﬂame collateral IPH has an impact on the public’s view of IPV as either the
outcome of a relational dispute or part of a pattern of abuse. What the mass media
communicate to the public about collateral IPH (and IPH generally) has the potential to
sway opinion and perpetuate myths about IPV (Bems, 2004; Ferrand-Bullock & Cubert,
2002; McManus & Dorfman, 2003; Meyers, 1994; Maxwell et al., 2000).

This study explores the ﬂaming of collateral IPH in Michigan newspapers. A
clipping service assembled the database; so all available years are analyzed (1990-2007).
Additionally, two competing theories of IPV, family violence and coercive control, will
be tested to determine which theoretical ﬂamework more ﬂequently describes the media
narratives and subsequent ﬂames. This is of particular importance because unlike
traditional applications of ﬂaming theory, the given news context does not revolve
around electing a candidate, changing an existing law or policy, or arguing for / against a
war (e. g. Allen etal., 1994; Nelson & Oxley, 1999); instead, this is the ﬁrst step towards
determining the types of ﬂames used to describe this social problem in the mass media,

and thus the potential impact on people’s opinions and perceptions of IPV.

Structure of Remaining Chapters

The literature review and theoretical ﬂameworks driving the current study are
presented in Chapter 2. The review is broken into two major sections: (1) ﬂaming in the
media and (2) competing IPV theories. Section One provides an introduction to
journalistic ﬂaming as well as a discussion of ﬂaming social problems: speciﬁcally
intimate partner homicide (IPH). Section Two introdr‘rces the major theories associated
with IPV scholarship: coercive control (an explication of feminist theory) and family
violence (an explication of systems theory). After a description of each, the section closes
with a review of the primary disagreements and ongoing debate associated with each
theoretical perspective. These two theories will be tested via quantitative and qualitative
analysis to determine which better informs media ﬂaming of collateral IPH and which
theory reﬂects the acts portrayed in the media. As such, relevant research questions will
close the chapter.

Chapter 3 provides a detailed explanation of the study’s research methodology
and design, including a brief introduction to the primary methods utilized in this project:
content analysis and mixed methodological analysis. After a brief description of the
selected sample (including exclusionary criteria), conceptual, operational, and variable
deﬁnitions are provided. Next the coding protocol will be introduced as well as how the
data were collected. Analysis procedures end the chapter.

Major results will be presented in Chapter 4 according to their corresponding
research questions and hypotheses. Since this is a mixed method endeavor, results for

qualitative and quantitative components will be described simultaneously as appropriate,

but care will be taken to delineate ﬁndings for clarity. Chapter 5 will synthesize and
present implications of the study’s results, including limitations and future research.
Lastly, Chapter 6 will provide recommendations and guidelines for journalists,

policymakers, and advocates.

Chapter 2: Literature Review and Theoretical Framework

Justiﬁcation of this Research Project

This dissertation is the ﬁrst of its kind. It assesses whether media ﬂaming of
collateral IPH more closely reﬂects coercive control or family violence theory, an
endeavor that will lead to a greater understanding of the mass media’s portrayal of the
causes and consequences of IPV. By testing these theories, which have been at odds for
over thirty years, the mass media and IPV literatures will beneﬁt ﬂom the insights gained
into how this social problem is constructed and reported to the public. Therefore, two
major areas of research: (1) media ﬂaming, including its application to social problems
and IPH and (2) competing IPV theories (coercive control theory and family violence
theory) are explored in this chapter. These topics form the foundation of research
questions as well as the conceptualization and operationalization of subsequent

hypotheses and variables.

Introduction to Framing

Frames “organize the world for both journalists who report it and, in some
important degree, for us who rely on their reports” (Gitlin, 1980, p. 7). Thus, ﬂaming is a
concept with two primary meanings: (1) it is the way in which media content is shaped
and contextualized according to some familiar point of reference or latent structure of
meaning and (2) it is the way in which an audience adopts the ﬂames of reference offered
by the mass media, leading them to see the world in a different way (McQuail, 2005). It

refers to the typical manner in which journalists shape news content, and subsequently,

how the audience adopts these ﬂames and sees the world in a way similar to journalists
and their informants (McQuail, 2005; Tuchman, 1978). This is a salient concept for this
dissertation because how IPH articles are ﬂamed by the mass media can play a role in
audience perceptions and attitudes. Tuchman (1978) notes, “the news ﬂame organizes
everyday reality, and the news ﬂame is part and parcel of everyday reality. ..[it] is an
essential feature of news” (p. 193). Media ﬂames encompass the working routines of
journalists, allowing them to quickly identify and classify information and package it “for
efficient relay to their audiences” (Gitlin, 1980, p. 7).

Erving Gofﬂnan (1974) and Gregory Bateson (1972) often receive credit as the
“pioneers” of ﬂaming research, with Goffrnan coming ﬂom a sociological perspective
and Bateson ﬂom anthropology. Each adopted a social-psychological perspective to
investigate how people make sense of their everyday realities (Reese, 2001 ). However,
the concept of ﬂaming can be traced back to the early 19303 and has been used in many
other disciplines including cognitive psychology (Bartlett, 193 2), economics (Kahneman
& Tversky, 1979), linguistics (Tarmen, 1979), social-movements research (Snow &
Benford, 1988), policy research (Schon & Rein, 1994), communication science
(Tuchman, 1978), political communication (Gitlin, 1980), public relations research
(Hallahan, 1999) and health communication (Rothman & Salovey, 1997). However, the
sociological approach proposed by Gamson (1989; 1992) is more typically applied to
ﬂaming in the ﬁelds of media and communication. From this vantage point, scholars set
out to determine how issues are constructed, how discourse is structured, and meaning
developed. The word frame traditionally represents an active process and the production

of a speciﬁc result (Reese, 2001).

Entrnan (1993), a scholar widely believed to lay the groundwork for most modern
ﬂaming research (McQuail, 2005), refers to it as a “ﬂactured paradigm”. He notes that
“to ﬂame is to select some aspects of a perceived reality and make them more salient in a
communicating text, in such a way to promote a particular problem deﬁnition, causal
interpretation, moral evaluation, and/or treatment recommendation in the item described”
(Entrnan, 1993, p. 52). His declaration that ﬂaming is a “scattered conceptualization”
that requires multiple theoretical perspectives to explain its occurrence has stimulated
much debate among mass media scholars. He publicly challenged its status as a theory,
criticizing its inconsistent deﬁnitions and use of key terms (i.e., frame, ﬂamework, and
ﬂaming). Entrnan (1993) concluded that the only way to master the ﬂaming construct is
to articulate one single theory of ﬂaming, a notion that has met with some resistance (see
D’Angelo, 2002).

Therefore, according to Entrnan (1993), ﬂaming consists of selection and salience
in that frames deﬁne problems, diagnose causes, make moral judgments, and suggest
remedies. This particular deﬁnition is especially useful in assessing the reporting of
crimes as it aligns with concrete steps involved in the criminal justice process.
Additionally, these “diagnostic ﬂames” provide a glimpse into the proposed motivation
and cause for the collateral IPH: a result of conﬂict (e.g., family violence) or an
intentional killing (e. g., coercive control). As such, this construct plays a central role in
this study because it provides a theoretically guided lens in which to investigate

journalistic sense-making of collateral IPH.

10

Media Frames

Framing studies are typically classiﬁed as “individual” or “media” oriented. An
individual, or audience ﬂame, is deﬁned as “mentally stored clusters of ideas that guide
individuals’ processing of information” (Entrnan, 1993, p. 52), while a media, or
audience ﬂame is a “central organizing idea or story line that provides meaning to an
unfolding strip of events. . .The frame suggests what the controversy is about, the essence
of the issue” (Gamson & Modigliani, 1987, p. 143). Scheufele (1999) provides a
comprehensive review of the difference between these ﬂames, deﬁning media ﬂames as
devices embedded in political discourse, while individual (i.e., audience) ﬂames are
conceptualized as internal structures of mind (Kinder & Sanders, 1990). Although both
types of studies share the common assumption that ﬂames link social / cultural realms
with everyday understandings of social interaction (Friedland & Zhong, 1996), individual
ﬂames highlight information-processing schemata, whereas media ﬂames focus on the
“attributes of the news itself” (Entrnan, 1991, p. 7). In sum, although each are related to

the other, the present dissertation focuses exclusively on media ﬂames.

Media Frames as a Dependent Variable

It is important to brieﬂy review studies that measure media ﬂames as dependent
variables since the present research utilizes this model to determine types of media
ﬂames found in newspaper reports of collateral IPH. Seminal research focusing on the
media ﬂame as a dependent variable can be found in Gans (1979), Shoemaker & Reese
(1996), and Tuchman (1978). These authors investigate the extrinsic and intrinsic factors

inﬂuencing the production and selection of news, and collectively determined ﬁve factors

11

that tend to impact media ﬂames: (1) social norms and values, (2) organizational
pressures or constraints, (3) pressures of interest groups, (4) journalistic routines, and (5)
ideological or political orientations of journalists (Shoemaker & Reese, 1996; Tuchman,
1978). Each of these factors play a role in how a collateral IPH is potentially ﬂamed, as
access to relevant information and preexisting routines may impact reporting. Recent
studies reﬂect these factors in terms of what impacts media ﬂames. For instance,
individual experience, daily routines (i.e. demands of the newspaper, the reactive nature
of news, assigrrment-speciﬁc writing / reporting), organizational issues, extra-media
effects (i.e. current events occurring outside the story), and ideology (Campos-McCoy,
2007) all play a role in how racial issues are ﬂamed. In their piece on international social
problem construction, Benson and Saguy (2005) found that cultural repertoires, legal
constraints, journalistic ﬁeld relations to the state market and competition among
journalistic outlets, and global position of nation-states impact media ﬂames.
Interestingly, in their content analysis of newspaper articles about the women’s
movement, Terkildsen and Schnell (1997) identiﬁed ﬁve media ﬂames that emerged as a
result of public opinion and issue perspectives held by political elites. The following
media ﬂames emerged: (l) the gender roles ﬂame, (2) the anti-feminist ﬂame, (3) the
economic rights / workplace ﬂame, (4) the political roles ﬂame, and (5) a feminism /
gender equality ﬂame. In a similar vein, Misra, Moller, and Karides (2003) assessed
magazine articles about the welfare movement (1929-1996) and found that the dominant
media ﬂame of “dependency” is a result of racialized and gendered images of welfare

recipients.

12

Obviously, the particular media ﬂame applied by journalists is related to the issue
being reported. As such, the current project, in line with the studies cited above, is at a
stage in which no issue-speciﬁc ﬂames exist; rather, ﬂames are expected to emerge
throughout analysis, subsequently categorized in accordance with two major IPV
theories. Therefore, Entman’s (1993) conceptualization of ﬂaming will provide a general
template to document the types of generic ﬂames implemented by reporters, while issue-
speciﬁc ﬂames of coercive control and family violence will be illuminated throughout
analysis. This dissertation will cover all of Entman’s (1993) generic frames: (1) problem
construction, (2) proposed cause of crime, (3) moral implications, and (4) potential
remedies, although the ﬁrst two are expected to be the most apparent. These generic
ﬂames will guide coding and interpretation of results, leading to issue-speciﬁc ﬂames
associated with coercive control and family violence theories. For instance, the
quantitative protocol captures proposed causes, motivations, and portrayals of the

homicide(s), while the qualitative is segmented according to these ﬁve ﬂame categories.

Selection and Construction of Collateral IPH Frames

The constructionist media ﬂame paradigm proposed by D’Angelo (2002) posits
that journalists create “interpretive packages” that are heavily inﬂuenced by their sources
or sponsors. These “packages” reﬂect the “issue culture” of a given topic (Gamson &
Mogdiliani, 1987, 1989). D’Angelo (2002) states that researchers can work across
paradigms, and presents studies that exemplify this synthesis (McLeod and Detenber,

1999; Iyengar, 1991; Price et al., 1997; Gamson, 1992).

13

The issue of information selection and subsequent construction of news forms the
foundation of the social constructivist media effects model. The media “actively select
the ﬂames of reference that readers or viewers use to interpret and discuss public events”
(Tuchman, 1978, p. ix). Some scholars have gone so far as to say that journalists give the
story a “spin” (N eurnan, Just, & Crigler, 1992); a result of the inﬂuential factors
discussed earlier (i.e. organizational constraints, personal judgments, etc.) This is
important to keep in mind as one interprets the types of ﬂames used to describe collateral
IPH, as well as the dominant “reasons” for presenting them in a certain way (e.g., conﬂict
or argument-related as opposed to coercive or abusive explanations). For example,
according to the constructionist paradigm, journalists can only perceive a portion of
reality. Therefore, they select (via internal psychological processes) certain pieces of
information in an effort to make sense of the whole. As such, a reporter might interpret a
neighbor’s testimony that a couple was experiencing “domestic problems” as evidence
for a symmetrical, conﬂict-oriented homicide cause. In fact, constructionism emphasizes
the interactive process of constructing social reality (Scheufele, 2000). As such, reporters .
have a tendency to emphasize certain information that they see as salient (i.e., an
impending divorce, inﬁdelity, or separation) in order to attract the audience, while
neglecting others, such as an outstanding personal protection order or history of violence

(Entrnan, 1993).

Newspaper Framing of Social Problems

The literature presented to this point has focused on the theoretical underpinnings

of ﬂaming. However, this dissertation highlights collateral IPH, an occurrence considered

14

a persistent health, social, and political problem in the United States (see Bems, 2004;
Stark, 2007). Before moving on to a discussion of ﬂaming in this particular context,
historical and recent deﬁnitions of “social problems” will be explored to provide a
foundation for the current analysis.

When reﬂecting on social problems in the United States, crime (including
intimate partner violence and homicide), discrimination, and poverty immediately come
to mind. However, conceptualizing a “social problem” is difficult, as sociologists have
been debating appropriate deﬁnitions for years (Spector & Kitsuse, 2000). As such, the
majority of theoretical scholarship on social problem construction comes ﬂom the work
of Blumer (1971), Becker, (1963), Spector and Kitsuse (1977), Gusﬁeld (1981),
Schneider (1985), and Best (1995). Some deﬁnitions are as basic as: “social problems
are trouble spots within society — social arrangements that do not work properly” (Best,
1995, p. 3), while others deﬁne a social problem as “a social condition that has been
found to be harmful to individual and/or societal well-being” (Bassis, Gelles, & Levine,
1982, p. 2). Others have been slightly more descriptive, claiming that a social problem is
(1) is widely regarded as undesirable or as a source of difﬁculties; (2) caused by the
actions or inactions of people or of society; and (3) affects, or perceived to affect large
groups of people (Farley, 1987). Best (1995) deﬁnes social problems as events that
emerge when a claimsmaker (e. g. a journalist) brings an issue to the public. Spector and
Kitsuse (2000), however, offer a more recent conceptualization that views social
problems as (1) constructed and (2) part of an ongoing process.

We have proposed to conceive all social problems to be activities of individuals

or groups making assertions about perceived social conditions, which they

15

consider unwanted, unjust, immoral, and thus about which something should be

done. This deﬁnition proposes that any such claim may become a social problem,

and focuses research on the process by which claims are assembled and asserted

by the claimants (p. xi).

Unfortunately, recent scholarship on Speciﬁc social problems does not provide
substantive conceptualizations, as authors primarily rely on the readers’ commonsense
and guidelines set forth by seminal theorists. However, inquiry into the nature and
impact of social problems is widely investigated. For example, in a recent study on the
struggles of the US underclass, J argowsky & Yang (2006) deﬁne urban poverty as a
situation that produces “degradations in. . .quality of life due to extant social conditions”
(p. 58). Additionally, studies of life in shuns (Pokhariyal, 2005), the feminization of
poverty (Brady & Kall, 2008), and HIV/AIDS in sub—Saharan Aﬁica (Agadjanian &
Menjivar, 2008) discuss the implications of their selected social problem. IPV and IPH
are also considered social problems (e.g., Garcia etal., 2007), so this study contributes to
the social problem literature by deﬁning and exploring collateral IPH as a distinct social

problem in US. society.

Newspaper Framing of Domestic Violence

A few studies have been published on the newspaper ﬂaming of domestic
violence, with most concluding that journalists provide a distorted view of the crime
(Bems, 1999; Consalvo, 1998; Meyers, 1994). Meyers (1994), through a textual analysis
of newspaper articles, notes that news coverage of domestic violence is typically ﬂamed

to support the status quo. In other words, coverage blames the victim while relieving

l6

society of any obligation to the incident. Similarly, Maxwell et al. (2000) found that
articles cover speciﬁc incidents, and tend to place blame or solutions with individual
victims and perpetrators, failing to offer broader solutions for the problem.

In her book, Framing the Victim: Domestic Violence Media and Social Problems,
Nancy Bems (2004) proclaims that this social problem focuses only on the victims; either
celebrating their courage for leaving a violent relationship or blaming them for letting the
abuse continue. Frames ﬂequently utilize accusations; accusing the victim of insti gating
the abuse or holding them responsible for ending the abuse. She warns that although
some of these ﬂames have helped foster support for victims (i.e. developing legislation or
building shelters), they have done little to “develop public understanding of the social
context of violence and may impede social change that could prevent violence” (p. 3).
She goes on to warn that as a result of these ﬂames, the violence perpetrated by abusers is
lost; domestic violence is not seen as a social problem originating with the abusers Since
they are excluded ﬂom the problem deﬁnition portion of ﬂame development. The
consequences of this method are serious; domestic violence is not ﬂamed as a social
problem resonating in a social and cultural context that tolerates and fosters abuse.

In sum, despite the attempts made by activists, advocates, professionals, and
researchers, the overall context of domestic violence has not resonated in the public
consciousness. Bems (2004) points to a void in the literature that this dissertation will
address, speciﬁcally the inﬂuence of context and principal relationship information on
the ﬂaming of intimate partner violence. She states: “However, the information these
people can provide us (i.e. advocates, professionals, etc) has not seriously inﬂuenced the

general public’s understanding of domestic violence and our society’s typical public

17

policy response to the problem” (p. 3). In other words, the crime is presented as
“incident” speciﬁc rather than a part of a continuing cycle of violence (see Ferrand-
Bullock & Cubert, 2002). As such, Bem’s (2004) ﬂaming of domestic violence in the

media provides a starting point for this study.

Newspaper Framing of Intimate Partner Homicide

Intimate partner homicide (IPH) is the most extreme outcome of IPV. However it
is also the most objective data source since it does not rely on self-reports (Cohen,
Llorente, & Eisdorfer, 1998; Garcia et al., 2007; Straus & Smith, 1990). While the events
leading up to collateral IPH remain outside the scope of this study (only newspaper
articles reporting the crime post-homicide are available), it is important to brieﬂy
introduce empirically validated risk factors for IPH in order to establish a context for the
present topic. For instance, the greatest risk factor for IPH is history of abuse (Bailey et
al., 1997; Campbell et al., 2003; Kellermann & Heron, 1999). Additionally those
(Bachman & Saltzrnan, 1995; Gauthier & Brankston, 2004; Rennison & Welchans, 2000;
Tjaden & Thoennes, 2000a) who are divorced, separated or never married are more likely
to be victims of IPH. These statistics are important to keep in mind as the reader explores
research methodology and results in Chapters 3 and 4.

F errand-Bullock and Cubert (2002) analyzed Washington newspaper coverage of
domestic violence fatalities, including the accuracy of victim portrayals and overall
attention to the broader social issue of domestic violence. They discovered that coverage
was often distorted (a ﬁnding echoed by scholars investigating domestic violence),
supporting common stereotypes and misconceptions of domestic violence. Taylor and

Sorenson (2002) conducted a content analysis of homicides reported in the Los An geles

18

Times between 1990-1994. They found that coverage of homicide was episodic, factual,
and unemotional in its tone. Additionally, homicides perpetrated by intimates were
covered differently than other homicides; they were less likely to be “opinion
dominated”, emotional, and more likely to begin with a “hook”.

Wozniak, (2007) investigated gender issues in the reporting patterns of IPH. As a
result of a newspaper content analysis she conﬁrmed Best’s (1995) hypothesis that the
media attempt to sensationalize the news by highlighting the worthiness of the victim and
its role in how stories are reported. Although her study was fairly small, (n=100), she
found that the broader issue of domestic violence (or IPV) was not mentioned in 99% of
the sampled articles. When context was reported, however, it was generally used to
describe the rarity of female perpetrated IPV, leaving the reader to wonder about the
ﬂaming of the perpetrator.

Although additional articles on newspaper reporting of IPH are unavailable, other
databases have been analyzed in order to investigate this crime’s causes and
consequences; most notably police records. For instance, Block and Christakos (1995)
investigated IPH in Chicago over a twenty-nine year period. They conducted the study in
order to differentiate between expressive homicides (typically associated with female
perpetrators) and instrumental homicides (typically associated with male perpetrators),
with expressive homicides being those in which the motive is to hurt the other person,
while instrumental homicides are perpetrated in order to obtain something, such as
money or property. Their sample was not drawn ﬂom a comprehensive newspaper
database; instead, police records were utilized. Similarly, Bowman and Altman (2003)

looked at wife murder ﬂom 1910-1930 and used police records to answer research

19

questions in lieu of newspapers. Each of these studies explored gender issues, however
they were not resolved. Recently, Biroscak and Smith (2005) assessed how many cases of
IPH that occurred in Michigan between 1999-2003 (determined via death reports, police
reports, etc) actually ended in newspapers. To their surprise, only 75% were printed in
newspapers. Thus, it is safe to conclude that Bem’s (2004) recent book on the ﬂaming of
domestic violence in mass media is among the ﬁrst to explain the intersection of [PV and
journalistic ﬂaming. The literature remains sparse although many advocacy groups have
developed guidelines for reporters on how to cover IPH (e. g., Ryan, Anastario, &

DaCunha, 2006).

Newspaper Framing of Collateral IPH

As of the current dissertation, no research exists on the newly developed concept
of collateral IPH. Additionally, no study has focused on media coverage of collateral
killings; only domestic violence / intimate partner violence and traditional IPH (one
victim and one perpetrator) have been investigated. Social science scholars have devoted
signiﬁcant attention to the consequences / risks of IPV for children and other family
members, most notably the impact of witnessing IPV (e. g. Coulbom Faller, 2003; Devoe
& Smith, 2002; Edelson, 1999; Ganley & Schechter, 1996; Graham-Bermann, 2001); but

even this stream of research falls outside this new area.

Competing IPV Theories

Family violence and coercive control theories are incorporated in this dissertation
because they represent two divergent perspectives of intimate partner violence. The crux

of the debate centers on the importance of patriarchy in domestic violence (Anderson,

20

1997). Feminist researchers contend that IPV is a result of gender and power (Dobash &
Dobash, 1979; Stark & F litcraft, 1991; Yllo, 1993), while family violence theorists argue
that patriarchy is just one variable in a complex web of causes (Straus, 1979).
Additionally, family violence scholars believe that IPV is a result of natural familial
conﬂict, and that violence is used to deal with stress, lack of resources, and to settle
arguments (see Straus, Gelles, & Steinmetz, 1980; 2007). On the other hand, feminist
theorists argue that IPV is a result of structural gender inequality and the use of male
power and control within intimate relationships. If one were to dichotomize these
perspectives, discussion would primarily revolve around the assumption of gender
symmetry in the perpetration of IPV. As such, coercive control (a concept explication of
feminist theory) and family violence theorists maintain different etiologies and outcomes
of intimate partner violence. For example, each theory regards gender and its place in
IPV research differently. Coercive control posits IPV is a gendered phenomenon (i.e.
men are more likely to perpetrate IPV than women), while family violence researchers
see its occurrence as gender-neutral; men and women are equally likely to become
perpetrators and victims of abuse. As such, this section provides a brief overview of each
theory, including its roots and development, closing with a summary of the primary

conceptual and methodological differences.

Family Violence Theory

Family violence theory emerged ﬂom the work of sociologists Murray Straus and
Richard Gelles at The Family Research Laboratory at the University of New Hampshire
(Straus, 1992). These scholars, with their roots in conﬂict theory (Adams, 1965; Coser,

1956; Dahrendorf, 1959; Scanzoni, 1972; Simmel, 1955; Sprey, 1969) and resource

21

theory (Allen & Straus, 1980; Homing, McCullough, & Sugimoto, 1981; O’Brien, 1975),
assert that violence is an inevitable part of “all human association” (Straus, 1979, p. 75).
Sociologists ﬂom this tradition are interested in general systems concepts, such as
“conﬂict of interest”, “aggression”, and “hostility”, and how they play out in the
American family (e.g. how “intrafamily conﬂict” occurs and is resolved) (Straus, 1979).

The national survey developed and utilized by family violence researchers to
measure IPV and other forms of abuse is referred to as the National Family Violence
Survey (NFVS) (Straus, Gelles, & Steinmetz, 1980). It uses the Conﬂict Tactics Scales
(CTS), which measure the methods implemented in response to a conﬂict situation
(Straus, 1987). The CTS situates intimate partner violence (or intrafamilial violence)
within conﬂict theory (e. g. IPV is a natural outcome of relational conﬂict). The CTS was
ﬁrst launched in 1979 to capture the “taken-for-granted” violence that occurs in daily life.
Straus (1979) used Coser’s (1956) conceptual deﬁnition of conﬂict, which is deﬁned as
the “means or behavior used to pursue one’s interest” (p. 76). Thus, IPV is a result of
interpersonal conﬂict. Straus (1979), in designing the CTS, articulated three primary
control “tactics”. They are:

1. The use of rational discussion, argument, or reasoning, or an intellectual approach
to a dispute (deﬁned as “reasoning”). This is the most adaptive means in which to
settle an interpersonal disagreement.

2. The use of verbal / nonverbal acts that symbolically hurt the other, or threaten to

hurt the other (deﬁned as “verbal aggression”). This occurs when one partner

threatens the other via physical or psychological means.

22

3. The use of physical force as a means of resolving conﬂict (deﬁned as “violence”).

In this case, hitting, pushing, slapping, or hitting with an object are used as the

means to settle a dispute.

The CTS, then, consists of a list of actions that a family member might implement
when in conﬂict with another. Therefore, when intimate partners’ agendas differ (e. g.
what television program to watch or how to handle the family’s ﬁnances) or they are
dealing with a loss / Shift of resources (e. g. loss of employment, economic troubles,
decreased occupational status of one spouse), any of these tactics can be used. However,
according to family violence theorists, these actions are the inherent result of conﬂict.
The CTS was revised in 1996 with the creation of the CTSZ (Straus et al., 1996). This
version is still used today.

Therefore, a certain amount of family violence is normal. In fact, “most of the
violent acts that occur in the family are so much a part of the way family members relate
to each other that it is not considered violence” (Straus, Gelles, & Steinmetz, 2007, p.
39). Family violence theorists, then, delineate between “normal” and “abusive” violence.
Normal violence, which is a result of a conﬂontation that takes place between intimates,
can cause physical pain and injury that range ﬂom slight pain to murder. Abusive
violence, on the other hand, is “an act which has the potential for injuring the person
being hit” (Straus, Gelles, & Steinmetz, 2007, p. 22). In other words, the “abuser” sets
out to injure, maim, or even kill the other. These two concepts are difﬁcult to delineate,
as one must pinpoint a motive or context to determine if the act is abusive or normal
violence. AS a result, additional concepts, including “aggression” and “physical

aggression” are used to supplement the two.

23

The Theory of Coercive Control

Feminist theorists argue that IPV is part of a system of coercive control in which
men maintain societal dominance over women (Dobash & Dobash, 1979; Martin, 1976;
Stark & F litcraft, 1996). The theory of coercive control is a recent explication in the ﬁeld
of IPV research. Prominent pioneers that contributed to the development of this theory
include Kersti Yllo, Michele Bograd (e. g. Yllo & Bograd, 1988), Susan Brownmiller
(1978), Carrie Yodanis (see Yodanis, 2004), Jacquelyn Campbell (e.g. Campbell &
Soeken, 1999), Donald Dutton (see Dutton, 2006, 2007), Richard Felson (e. g. F elson &
Messner, 2000), Ann Jones (e. g. Jones, 2000), G. Ann Bogat (e. g. Bogat et al. 2003,
2005), Michael Johnson (e. g. Johnson, 1995), and Kristin Anderson (e. g. Anderson,
1997) among others. These scholars focus on the gendered nature of IPV, with many
directly challenging the family violence theorists. Using the feminist ﬂamework as a
theoretical foundation, Evan Stark (2007) recently articulated a theory that encompasses a
majority of the tenets proposed by feminist IPV scholars, including the centrality of
gender and power. Four additional areas of feminist IPV scholarship have also informed
Stark’s (2007) work: (1) the explanatory utility of the constructs of gender and power; (2)
family as a historically Situated social institution; (3) the crucial importance of
understanding and validating women’s experiences, and (4) employing scholarship for
women, rather than simply about them (p. 14).

Of primary importance in this evolving theory is the concept of coercion.
Coercion is the use of force or threats to either compel or dispel a certain response in

another human being

24

No matter how many punches or injuries or instances of depression are
catalogued, the cage (i. e. the environment in which a battered woman lives)
remains invisible as long as we omit the strategic intelligence (i. e. the
manipulative, calculated nature of IP IO that complements these acts with
structural constraints and organizes them into the pattern of oppression that gives

them political meaning (p. 198).

Therefore, coercive control theory implies the systematic subjugation and
methodical action perpetrated by an abuser. Feminist IPV scholars acknowledge the
importance of this concept, noting that battering takes place within a context of coercive
control, (e.g. Campbell & Soeken, 1999). However, this is the ﬁrst theoretical
explication to place this concept at the foreﬂont. Stark (2007), unlike early feminist
scholars, directly challenged the notion that women crumble under the tyranny of abuse,
suffering ﬂom classic disorders such as Battered Women’s Syndrome (BWS) and post-
traumatic stress disorder (PTSD). He claims that female victims of coercive control react
in ways similar to hostages, with the exception that “they are the formal equals of men
who oppress them, not their subjects. Their subjugation occurs against a background of
entitlement as well as inequality...” (p. 199). He notes that if men want to maintain their
structural hierarchical superiority they “must do so directly and personally in each
relationship. These qualities mark coercive control as deliberate and malevolent, setting
the stage for them to be deﬁned as criminal” (pp. 199-200). Since the theory of coercive
control refutes the assumption that women suffer ﬂom the maladies of battered wives,
women may actively respond to their abuser’s calculated attacks on their personal

ﬂeedom and liberty (e. g. monitoring mileage on her car or making her keep a log of daily

25

activities) in ways not traditionally accepted (i.e. they engage is self-preserving behaviors
and defense mechanisms rather than succumbing to the violence and abandoning hope).

The groundwork for a theory of battering as coercive control can be traced back to
the 19505 and 19608. However, the feminist movement ﬁrst gained public attention in
the 19703 with a candid discussion of rape and domestic violence (e. g., Brownmiller,
1978). One scholar (Jones, 2000) expanded the Amnesty International “chart of
coercion” to illustrate the numerous control tactics employed by abusers, including
isolation, threats, degradation, and “total destruction of will”. Along with other pioneers
in the battered woman’s movement, she argued for a deﬁnition of woman abuse that
encompasses controlling behavior that is created and maintained by an imbalance of
power between abuser and victim. Adams (1988) conceptualized battering as
“controlling behavior” with a violent act being anything that causes the victim to do
something she doesn’t want or causes her to be aﬂaid, regardless of the use of physical
violence. Bancroft (2002), a counselor that works with abusive men, also documented
the coercive behavior aspect of IPV; speciﬁcally the rewards batterers gain ﬂom
employing coercive control. Many therapists working in the ﬁeld of IPV adopt the
Duluth Model (e.g., Pence & Paymar, 1993; van Wormer & Bednar, 2002), an
intervention stemming ﬂom the feminist-critical paradigm that clearly identiﬁes male
power and control as primary causes of violence against women. The Duluth Model is a
comprehensive community-based intervention program, focused on victim safety and
accountability of perpetrator (www.duluth-model.or;g, 2008).

To review, coercive control theory links women’s struggles at home to a larger

discourse of human rights through an analogy of hostage taking or kidnapping (Stark,

26

2007). This theory has tremendous heuristic value because it illuminates dimensions of
IPV that are underrepresented in the literature because of a lack of association with
physical assault or diagnosable mental health problems. This perspective also encourages
innovative research designs that are able to capture subtle tactics of abuse. As such, the
intent of coercive control in an intimate relationship is to:

Usurp and master a partner ’s subjectivity — in its scope of its deployment, its

individualized and personal dimensions, and its focus on imposing sex stereotypes

in everyday life. The result is a condition of unfreedom (what is experienced as

entrapment) that is “gendered” in its construction, delivery, and consequence (p.

205)

In sum, a perpetrator using coercive control intends to punish, hurt, or control the
IPV victim. The act is cumulative, ﬂequently resulting in injury or death. Most
importantly, coercive control is a gender strategy with three major dimensions: (1) a
basic set of beliefs / values of what it means to be a man or woman in today’s society (i.e.
a gender ideology); (2) the resources, tools, techniques, and tactics to implement one’s
gender ideology (i.e. a gender technology); and (3) an action plan (Stark, 2007).
Coercive control is general in its scope (i.e., it encompasses a diverse set of intimate
situations); however, the perpetration of IPV is its primary outcome variable of interest. It
provides researchers with the necessary tools to understand and explain the killing of
collaterals as a means to subjugate and control an IPV victim. As a result, the present
dissertation will be the ﬁrst to test Stark’s (2007) explication as it relates to journalistic

ﬂaming of collateral IPH.

27

Theoretical Contradictions

Family violence and coercive control theories contradict in two major areas: (1)
the conceptualization of gender and its impact on IPV and (2) the appropriate population /
methodology for capturing IPV. These differences have numerous consequences for IPV
research, as each theory proposes contrasting reasons for violence, including its causes,
consequences, and cures. It is important to keep in mind, however, that this dissertation
is testing collateral IPH, which unlike traditional IPV scholarship, focuses on peripheral
victims, or those immediately outside the primary intimate relationship (e. g., a child,
parent, new romantic partner). This is a situation that has never received attention in the
literature, although scholars ﬂom both theoretical camps have investigated the causes of
child abuse and child witnessing of domestic violence. Since these two theories are used
ﬂequently in IPV scholarship, their core assumptions and disagreements will be. reviewed
and subsequently utilized in determining the types of ﬂames used to present collateral
IPH.

Feminist theory (and later coercive control) maintains the centrality of gender,
patriarchy, and power in IPV. While scholars in this tradition ask, “why do men beat
their wives”, family violence theorists inquire, “what triggers violence in the family”?
Family violence researchers assume that “particular structural arrangements within
families produce stress and conﬂict, that the family system responds to the dynamics and
conditions of the larger society, and that, mediated through socialization and learning,
violence is one response to structural and situational stimuli” (Bograd, 1988, p. 18; also
see Straus, 1973). This is not to say that gender does not have an impact in family

violence scholarship. In reporting results ﬂom two national surveys, Straus and Gelles

28

(1986) note, “One of the most ﬁrndamental reasons why some women are violent in the
family, but not outside the family, is that the risk of assault is greatest in her own home”
(p. 471; also see Straus, Gelles, & Steinmetz, 1980). They conclude, “Nonetheless,
violence by women against their husbands is not something to be dismissed because of
the even greater violence by husbands” (p. 471). Thus, family violence scholars simply
explore a different battering context that incorporates concepts such as “couple violence”.
Researchers of both disciplines acknowledge that both sexes have the potential to be
violent; however family violence theorists assume IPV is gender symmetrical or equally
likely to occur between both sexes. Feminist theorists have not accepted this conclusion.
According to Das Dasgupta (2002), a feminist researcher, “In this particular instance,
men and women’s behaviors are perceived as the same and both are termed battering due
to the obliteration of contexts” (pp. 1377). Additionally, situations in which violence is
minor or the woman is acting defensively / protectively are incorporated in family
violence surveys, as are instances in which initial aggressive behaviors lead men to
respond. Feminist researchers would argue that this type of violence is not IPV. In her
study, Das Dasgupta (200) found that women who assault their partners are distinct ﬂom
men who engage in battering behaviors. She concluded that men and women’s violence
toward their heterosexual intimate partners is historically, culturally, motivationally, and
situationally dissimilar ﬂom each other. One must look at the context (i.e., motivation,
cause, etc) when deﬁning intimate partner violence.

As such, family violence researchers adopt a “gender-neutral” approach to
investigating IPV. Models of familial violence incorporate sexual inequality as one

among many factors of interest (e. g. Gelles, 1985; Straus, Gelles, & Steinmetz, 2007). In

29

fact, family violence theorists ﬂequently point to the inadequacy of feminist research, as
it tends to focus exclusively on gender or patriarchy. Other variables such as income,
unemployment, and age are not always considered (Gelles, 1985). They argue that these
variables may explain the perpetration of [PV in both men and women (Gelles, 1993;
Gelles & Straus, 1988). However many feminist researchers such Jackie White (2002)
and Evan Stark (2007) have moved beyond a single variable focus, developing complex
ecological, meta-theoretical models. Yet, the debate continues: “Rancorous debates
between scholars who identify gender and power as the key processes in partner violence
and those who view gender as just one component of the problem have dominated the
literature during the past decade” (Anderson, 2005, p. 853).

Thus, the role of gender in theoretical development and its subsequent impact on
research methodology and measurement is of tremendous importance. Feminist scholars
argue that family violence research represents an individualistic perspective because it
removes the context and consequences of abuse ﬂom the equation (Anderson, 2005). For
example, Kimmel (2002) argues that sex-symmetry in IPV is individualistic because
family violence researchers do not conduct a thorough analysis of gender and how one’s
gender identity and ideology (i.e. the cultural deﬁnitions of masculinity and femininity)
play a role in researchers’ conceptualization, operationalization, and measurement
(Kimmel, 2002) of the construct. If gender is only considered one property of individual
reasons for abuse, there is little to no relationship between gender and partner violence.
Additionally, an individualist approach suggests that gender is not an important predictor
of IPV, or that IPV is not a gendered phenomenon (see Dutton, 1994; F elson, 2002;

Straus, 1993). However, family violence theorists would argue that gender is not

30

excluded ﬂom analysis per say, it just plays out differently (i.e., husband-beating vs.
wife-beating). Straus (2006) acknowledges that husband violence is more dangerous,
injurious, repeated, and does more damage. Additionally, a majority of wife violence is
in self-defense or the response to an assault. They conclude that despite these ﬁndings,
IPV research must go beyond controlling assaulting husbands.
It seems that violence is built into the very structure of society and the family
system itself... (wife-beating) is only one aspect of the general pattern of family
violence, which includes parent-child violence, child-to-child violence, and wife-
to-husband violence. To eliminate the particularly brutal form of violence known
as wife-beating will require changes in the cultural norms and the organization of
the family and society which underlie the system of violence on which so much of

American society is based (p. 44).

Issues of Sampling 8. Measurement

Different methodological strategies and study populations emerge as a result of
the centrality of gender and power in analysis. For instance, the National Family
Violence Survey, a general population survey (GPS), which includes the CTS (and its
later version, the CTSZ) is an instrument given to households in which a married couple
currently resides. Although researchers capture a representative sample of “married” US
homes, one major question remains, “is this a valid and reliable measure of IPV”? By
only focusing on “intact” households, couples that are not cohabiting (or married),
divorced, or separated are excluded ﬂom consideration. This has direct implications for

scholarly understanding of IPV, as research has previously indicated that risk of severe

31

violence and homicide dramatically increase after a woman has exited an intimate
relationship (Garcia et al., 2007). Critics of the CTS and CTSZ argue that context is
ignored. They note, “the act base measures used by family interaction researchers,
notably the Conﬂict Tactics Scales, consider such acts out of context and neglect their
consequences” (Dobash et al., 1992, p. 651). If the consequences of violence were
considered (i.e. severity of injury), nearly all victims would be women. The CTS, in its
quest to measure “couple violence” which is “less a product of patriarchy, and more a
product of the less gendered causal processes discussed at length by [scholars] working in
the family violence tradition” (Johnson, 1995, p. 285; Johnson, 2005) tends to focus on
the general population, not those that have been victimized by an intimate partner. Thus,
a major fault of the GPS is sample bias in that battered women tend to be excluded in
household surveys.

Feminist researchers, on the other hand, ﬂequently utilize purposive, in-depth
qualitative designs that capture victims’ experiences and struggles (e.g. Johnson &
Sullivan, 2008). This is a useful tool for understanding the details of a particular
woman’s environment, but does not guarantee an objective, representative sample of the
population. Scholars are rarely able to generalize their ﬁndings or identify regular trends
in IPV. The use of this particular methodology is ﬂequently a result of feminists’
commitment to advocacy and victim empowerment. Although rich data are generated,
the reliance on clinical samples creates a problem with sampling bias. Researchers risk
overestimating IPV and drawing inaccurate conclusions of its impact on certain

subgroups of women.

32

However, feminist theorists have recently compensated for this criticism by going
beyond clinical sampling and generating population-level surveys of their own. For
example, in order to reach a more representative sample of women while still honoring
the importance of context, general population surveys (GPS) have been used to capture
respondents who may no longer live in an abusive relationship. The National Violence
Against Women Survey (NV AW) is perhaps the most well known GPS in that it
surveyed 8,000 US men and 8,000 US women about their experiences as victims of
violence, including IPV (Post, Klevens, & Maxwell, in press; Tjaden & Thoennes,
2000b). Unlike the surveys executed by family violence researchers (i.e. the National
Family Violence Survey), eligible respondents did not have to be married or in a current
cohabiting relationship. Additionally, the NVAW, unlike the National Crime and
Victimization Survey (N CVS) (another GPS utilized by feminists) found that women are
more likely to experience IPV than men (Tjaden & Thoennes, 2000a). This may be a
result of unit of analysis, as the NCVS samples household representatives rather than an
equal number of men and women. Recently Post, Klevens, and Maxwell (in press)
conducted the largest random digit dial (RDD) survey to date.

As such, the samples (and populations) studied by family violence and feminist
theorists are inherently diﬂerent. Family violence researchers tend to study large samples
of married, cohabiting, or dating couples, while feminist researchers select samples based
on high levels of partner violence (Archer, 2000). “The sex-symmetry debate reﬂects the
failure to separate these distinct types of violence rather than the inadequate measure of
violence; family violence and feminist researchers are simply studying different

populations that engage in distinct types of intimate violence” (Anderson, 2005, p. 854).

33

For instance, Johnson and Leone (2005) separate “intimate terrorism” (a feminist
concept) and “situational couple violence” (a family violence concept) as distinct events.
Even Straus (1997, 1999), the pioneer of family violence research, notes that their

contradictory conclusions may be a result of different populations.

Theoretical Implications for this Project

Although this ongoing debate will not be resolved in the near future, feminist and
family violence theories represent the dominant lenses in which scholars, practitioners,
and even the public understand and explain IPV. If one thinks about the common
stereotypes that circulate on this topic (“She deserved it; she must have done something to
make him angry”, “he was drunk and lost control”, or “the stress of losing his job
triggered the rage”), these theoretical assumptions play a part in the views held about
intimate partner abuse and violence. Therefore, this dissertation (1) addresses the
polarization of the literature, (2) assesses the connection between IPV theory and the
media, and (3) dissects which major IPV theory informs media ﬂames. It does so by
exploring the contrasting causes and explanations for IPV by analyzing seventeen years
of newspaper reporting, which provides a glimpse into the dominant assumptions held by
the media. Although conclusions about journalistic motives or intent cannot be drawn, the
speciﬁc ﬂames used to present collateral IPH will present a much-needed understanding
of the predominant theory of IPV.

In sum, each theory offers a divergent perspective of IPV, with their own set of
strengths and weaknesses. However, they provide an excellent template to assess the

dominant ﬂaming patterns used in collateral [PH reporting. By deciphering which theory

34

ﬂames more ﬂequently align with, media scholars can determine if story construction
adequately reﬂects the complexities of intimate partner violence and homicide and if vital
information necessary for public understanding is excluded. This has direct implications
for scholars across disciplines, but most importantly, can offer insight to the journalists

and editors that produce stories on collateral IPH.

Summary

The above literature review explored two primary areas: (1) mass media ﬂaming
and (2) the two major IPV theories. To brieﬂy review, the conceptualization of ﬂaming
utilized in this dissertation is as follows: ﬂaming consists of selection and salience in that
ﬂames deﬁne problems, diagnose causes, make moral judgments, and suggest remedies
(Entrnan, 1993). This typology provides an excellent template for assessing the overall
problem, causes, and consequences of collateral IPH, and whether ﬂames tend to
resonate around family violence issues (i.e., relationship conﬂict) or coercive control
(i.e., intentional harm on one’s partner). AS demonstrated through a discussion of
ﬂarning’s developmental history, its difference ﬂom audience ﬂames, and its role as a
dependent variable, ﬂaming is a ﬂexible, dynamic construct that is ﬂequently applied in
many disciplines and studies. Although a majority of ﬂaming research highlights political
events, scholarship does exist on the ﬂaming of social problems, as indicated by Best
(1995) and Berns (2004). In fact, Berns (2004) is among the ﬁrst to focus on issues
related to violence against women.

Scant research exists on the ﬂaming of IPH, and no research exists on collateral
IPH. As a result, the predominant IPV theories described in this section will help

researchers better understand and explain the occurrence of IPV as well as why

35

journalists ﬂame news stories the way they do. These theories, in addition to

constructionist ﬂaming, will be tested throughout the remainder of this dissertation.

Research Questions & Hypotheses

Two primary research questions drive the current dissertation. These questions
test two divergent IPV theories introduced in this earlier: coercive control and family
violence theory. As a result, numerous sub-research questions and hypotheses were
generated.

The ﬁrst major research question, “what are the predominant trends in collateral
IPH”, is posited to capture the general characteristics associated with collateral IPH and
will be answered by answering the following sub-questions and corresponding
hypotheses:

RQla: Who are the primary collateral victims and perpetrators in newspaper

coverage of this topic?

H13: The primary collateral perpetrator will be the current or ex male
intimate partner.
Hlb: The primary collateral victim will be the Primary IPV Victim ’5 child.

The above hypotheses are informed by feminist theory, speciﬁcally coercive
control (Stark, 2007). According to the theory of coercive control, IPV is gender-
speciﬁc, with male perpetrators and female victims. Additionally, since the primary goal
of intimate terrorism is to control, hurt, and dominate another, it is hypothesized that the

most ﬂequently occurring collateral victim will be the primary victim’s child.

36

The second research question, “which theoretical ﬂamework better describes the
media narratives utilized to report collateral IPH” will be addressed by answering the
following sub-questions and hypotheses:

RQ2a: Why does collateral IPH occur?

H 20: Journalists more frequently report that the homicide is a result of
relationship conﬂict (i. e., family violence frame).

RQ2b: Is the collateral IPH reported as unexpected (i.e., loss of control, anger or

conﬂict-driven) or premeditated?

H 21,: Journalists more ﬁequently report collateral IPH as an unexpected
crime (i. e., family violence frame).

Rch: Does the proposed explanation of the homicide (i.e., whether it was

unexpected or premeditated) impact reporting ﬂames?

H20: If the collateral homicide is reported as unexpected, the proposed
cause will be relationship conﬂict (i. e., family violence ﬂame).

H2112 If the collateral homicide is reported as premeditated, the proposed
cause will be ongoing IPV or domestic violence (i. e., coercive control
ﬁ'ame).

RQZd: Does the type of victim (i.e. kids vs. Siblings vs. parents vs. new partner)

impact reporting?
RQ2e: Where are the collateral victims described in the article? Speciﬁcally, in
what sequence are the collateral victims introduced?
By determining this sequence, researchers can observe the priority

(whether intentional or non-intentional) given to each of the main players in

37

collateral IPH articles. Doing so will provide a glimpse into where collateral

victims fall in this complex IPV situation.

Since the dissertation data come ﬂom newspaper articles, it is difﬁcult to discern
the history and context of homicides as the investigator is looking at the homicide
through the lens of the journalist. However, the impact of gender and proposed cause of
violence can be captured, which is central to the IPV theoretical debate. If males and
females are reported to perpetrate homicides equally, with the primary cause of violence
stated as relationship conﬂict, then family violence theory will be supported. However, if
males are reported as the primary aggressors and are committing the collateral IPH to
terrorize and hurt their intimate partner (i.e., the homicide is premeditated) then coercive

control theory will be supported.

38

Chapter 3: Research Methodology

This study explores whether newspaper reporting of collateral IPV homicide
aligns more closely with coercive control or family violence theories, two perspectives
that are polarized in the literature. These theories facilitate the overarching research
questions: (1) “what are the predominant trends in collateral IPH” and (2) “which
theoretical ﬂamework better describes the media narratives utilized to report collateral
IPH”. Sub-questions and hypotheses are informed by core theoretical assumptions,
allowing for targeted measurement and analysis. As such, the present sample of
newspaper articles and subsequent mixed-method content analysis provides the tools

necessary to explore questions and support / reject hypotheses.

Sample

The Michigan Reading Press Service located in Williarnston, Michigan, collected
the IPH newspaper articles used in this study. The clipping service was unofﬁcially
initiated in 1989 when domestic violence shelters began sending newspaper articles to the
Michigan Coalition Against Domestic and Sexual Violence (MCADSV) Resource
Library. In 1991, the Michigan Domestic Violence Prevention and Treatment Board’s
(MDVPTB) Center on Sexual and Domestic Violence acquired funding to contract with a
Michigan newspaper clipping service for a more comprehensive and efﬁcient way to
collect statewide news about domestic and sexual homicides. Fearing sample bias, the
MDVPTB contracted with the clipping service to collect all IPH articles for the state of
Michigan. The selection criteria for the service were to search for and send all

newspaper articles about violent deaths occurring within domestic relationships (murder

39

or homicide of men or women involved in an intimate relationship). In addition, articles
were included if a death occurred during a sexual assault by an intimate partner or
stranger, or because a person was linked to the victim. Speciﬁc criteria included in
original purchase order ﬂom the state include:
1. “Murders of men, women, and teens by someone with whom the victim had a
romantic relationship.”
2. “Trials of men and women accused of murdering someone with whom they had a
romantic relationship.”
3. “Stalking charges and trials of intimate partners.”
4. “Domestic violence and laws, courts, police.”
5. “Agencies that address domestic violence.”

This collection is the most comprehensive available in the United States. Since
all daily or weekly newspapers were searched, any reported IPV homicide is included in
the database. This was the most efﬁcient and only way at the time to learn about
“domestic homicides” statewide, as well as determine how many occur within a year’s
time. Even Biroscak, Smith, and Post (2006) used this database for state surveillance.

Individuals not trained in the dynamics of domestic violence interpreted selection
instructions. Articles about child abuse deaths, Sibling homicides etc., were sent for
review by the Resource Library. The Resource Library director read and sorted the
articles by appropriateness and kept those that were related to intimate partner homicide
(referred to as domestic homicide in this case) as well as cases in which someone was
killed as a result of their involvement, relationship with, or proximity to an IPV victim.

The articles were sorted and stored by victims’ last names and subsequently used by the

40

Resource Library to create a yearly list of synopses of the IPV homicides reported. The
list was distributed statewide to shelters, batterers programs, trainers, students and others.

As such, a total of 48 Michigan newspapers, both daily and weeklies, are included
in the present sample, with circulation numbers ranging ﬂom 1,000 (Birch Run Herald)
to 1,215,149 (Detroit News and Free Press, Sunday morning edition). Please refer to
Appendix A for a complete list of sampled newspapers and included articles.

The entire database contains approximately 1200 articles. The sub-sample
selected for this project is based on the following criteria: (1) a collateral intimate partner
homicide is perpetrated and reported and (2) a previous / current intimate relationship is
established in the story. A total of 215 articles met this criteria and are included in the
present analysis. Thus, all newspaper articles are included in the sample where a family
member, ﬁiend, or signiﬁcant other is murdered. The principal individual involved with
the perpetrator may or may not be murdered, however at least one person other than the
primary partner is murdered in each case.

The unit of analysis in this dissertation is the article. The article was selected
since it is the most comprehensive, allowing for situations in which characteristics of
family violence and coercive control theories are presented in the same story.
Additionally, it was not unusual to come across multiple articles that report the same
collateral IPH. Each article was included even if it was the same incident described by a

different source or was continued coverage of the same incident.

Deﬁnitions

There are numerous concepts included in this study, each contributing to the

overall dissertation research questions and hypotheses. Independent variables are

41

generally exploratory and descriptive in nature. Each term is conceptualized,

operationalized, and provided with a variable deﬁnition. Thus, the major concepts are:

(1) gender of principal and collateral IPV victims (2) gender of collateral IPH perpetrator,

(5) collateral victim type (i.e. their relationship to the principal partners). Additional

variables include: (1) homicide explanation; (2) homicide portrayal, (3) perpetrator

intent; and (4) crime context. The outcome (dependent) variable of interest is the media

ﬂame. Table 3.1 summarizes the conceptual and operation deﬁnitions used in this

dissertation.

Table 3.1. Conceptual & Operational Deﬁnitions

 

 

 

 

 

 

 

 

 

 

Concept Conceptual Definition Operational Deﬁnition

Gender Complex, multilevel construct of what it Description of the perpetrator and
means to be male or female in a given victim(s)’s sex through use of names
context. and pronouns (he, she, etc.)

IPV Serious, preventable public health The violence and abuse is explicitly
problem that describes physical, sexual, described as occurring within an
or psychological harm by a current or intimate relationship in which one
former partner or spouse. individual is the abuser and the other

is the victim.

Principal IPV Partner or previous partner of the The individual identiﬁed in the

Victim individual that commits the collateral newspaper article as having been
IPH. Prior abuse may or may not be intimately involved with the
stated. perpetrator of the collateral IPH.

IPH / Collateral The individual who commits the intimate The individual identiﬁed in the

IPH Perpetrator partner homicide and / or collateral IPH article as killirg the collateral victim

Collateral Victim An individual outside of the identiﬁed The individual(s) who is / are killed
intimate relationship that is killed during during the incident. They are clearly
an IPV-related attack identiﬁed as being outside the

intimate relationship under
investigation.

Collateral IPH Collateral IPH occurs when someone Collateral IPH is measured by
known to the IPV victim and / or documenting the deaths of
perpetrator is murdered during crime individuals (including, but not
commission limited to children, family members,

new boyﬂiends/girlﬂiends) outside
the principal intimate relationship.

Homicide The proposed reasons for the collateral This concept will be measured as

Explanation IPH. follows: (1) accidental; (2) result of a

conﬂict or ﬁght in the principal
relationship; or (3) intentional /
premeditated.

 

 

42

 

Table 3.1 Conceptual & Operational Definitions Continued

 

 

 

 

 

 

 

 

 

 

Concept Conceptual Definition Operational Deﬁnition
Homicide How the collateral IPH is reported in a This concept is considered a
Portrayal given newspaper story dichotomy in the present analysis.

Either the homicide is presented as
shocking/unexpected or
deliberate/intentional. This concept
is measured by synthesizing reporter
and informant testimony.

Homicide The proposed “trigger” that led the This concept is measured by ﬁrst

Motivation and perpetrator to commit the collateral IPH documenting if the collateral IPH is

Cause related to a problem in the principal

(Perpetrator intimate relationship (i.e., divorce,

Intent) separation, inﬁdelity, etc). Second,

individual causes (i.e., mental health
or personality issues) and social
causes (i.e., unemployment, ﬁnancial
strain) are recorded.

Perpetrator’s Any background information provided by Background information related to

History of the journalist about the IPV victim, the collateral IPH. Potential

Violence (Crime collateral victim(s) or perpetrator, examples include previous attempts

Context) including past IPV or abuse in the to leave the relationship, prior
principal intimate relationship. involvement with the police, ﬁling a

personal protection order, number of
shelter stays, past domestic violence,
IPV-related arrests, etc.

Media Frames Media ﬂames deﬁne problems, diagnose Media ﬂames are determined via
causes, make moral judgments, and quantitative and qualitative analysis.
suggest remedies. They serve as They are measured according to
“interpretive packages”. Entrnan’s (1993) generic ﬂaming

criteria. Incident-speciﬁc ﬂames are
expected to emerge via qualitative
analysis.

Variables

The methodological goals are twofold: (1) describe the generic and incident-
speciﬁc media ﬂames associated with collateral IPH and (2) assess how closely story
characteristics align with family violence and coercive control theories. Implicit in these
goals is determining the proposed motivation and causal attribution for the crime(s). As
such, the primary outcome, or dependent variable, of interest is the media ﬂame. In line
with Entman’s (1993) criteria, the problem deﬁnition, potential causes, moral judgments,

and possible remedies are considered when determining ﬂames used to report collateral

43

IPH. Media ﬂames are determined by both qualitative and quantitative analysis, both of
which will be explained in a later section.

The primary exploratory, descriptive variables included in this study are: (l)
victim / perpetrator story position (i.e. the order in which perpetrator and victims are
introduced by the reporter); (2) homicide explanation (i.e., accident, result of ﬁght, or
intentional); (3) perpetrator history of violence / crime context; (4) homicide portrayal
(i.e., shocking or expected); (5) perpetrator motivation for homicide (i.e., whether it was
related to the principal intimate relationship or not); (6) violence symmetry; and (7)
causal attribution (i.e., individual or social factors). Each of these variables inform the
media ﬂames used to report collateral IPH, thus addressing the second major research
question of this study, “which theoretical ﬂamework better describes the media narratives
utilized to report collateral IPH”.

Descriptive variables are collected in order to provide general story characteristics
and document overall trends in collateral IPH (see research question #1). Therefore, (1)
crime stage, (2) crime victims, (3) occurrence of perpetrator suicide, (4) principal IPV
victim sex, (5) principal IPV victim death (i.e., if they were killed in addition to collateral
victim), (6) average number of collateral victims, (7) collateral victim age, (8) collateral
victim sex, (9) collateral victim relationship to principals, and (10) perpetrator sex are
collected. Please refer to Figure 3.1 for a visual depiction of how independent variables

cluster around the dependent variable, media ﬂame.

Figure 3.1 Variable Diagram

 

 

Story Homicide
Position Motivation

Homicide

E 1 ti Homicide
XP 3113 on \ Media Frame / Portrayal

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Perpetrator History Causal
of Violence / Crime Attribution
Context
Violence
Symmetry

 

 

 

Criteria for Determining Theoretical Alignment

It is possible that a given news article contains elements of both coercive control
and family violence theories. As such, the following table displays the key
characteristics used to determine if a story aligns more closely with tenets of family
violence or coercive control. The primary issues of consideration are the context and
cause of the homicide. The variables introduced in the previous section (see Figure 3.1)

inform the criteria located in Table 3.3.

Table 3.2. Criteria for Theory Determination

45

 

Criteria

Coercive Control

Family Violence

Relevant Variable

 

 

 

 

 

 

 

 

 

 

 

 

The perpetrator of collateral X Homicide Explanation

IPH is portrayed as a good Homicide Portrayal

person who made a bad Motivation for Homicide

decision

The perpetrator is portrayed X Homicide Portrayal

as a killer who knew what

he/she was doing

The collateral IPH is a result X Homicide Explanation

of a relational conﬂict, Motivation for Homicide

dispute

The collateral IPH is a result X Violence Symmetry

of one partner’s actions Causal Attribution

The collateral IPH was X Homicide Explanation

perpetrated to hurt principal Homicide Portrayal

IPV victim Motivation for Homicide

The collateral IPH is a result X Homicide Explanation

of anger, loss of control

The collateral IPH was X Homicide Explanation

planned, calculated, or Homicide Portrayal
premeditated Motivation for Homicide

The reporter provides quotes X Homicide Portrayal

/ testimony that the

collateral IPH was shocking,

unexpected

The reporter provides quotes X Perpetrator History of

/ testimony that the Violence / Crime Context

collateral IPH follows a

pattern, history of abuse

The collateral IPH is X Perpetrator History of

reported as an isolated Violence /

incident Crime Context

The collateral IPH is X Perpetrator History of

reported within a context of Violence /

IPV (i.e., prior violence is
noted in article)

 

 

 

Crime Context

 

Content Analysis Procedures

Utilizing both quantitative and qualitative procedures to address research

questions and hypotheses is gaining popularity in the social sciences. Numerous texts

have focused on the strengths, limitations, and practicality of mixed method research

designs (e. g., Axinn & Pearce, 2006; Mason, 2002; Tashakkori & Teddlie, 1998;

Tashakkori & Teddlie, 2003), concluding that despite their complexity, mixing methods

makes for a more comprehensive investigation. Mixed method research is rooted in

 

pragmatism, so scholars using these techniques invite diverse ways of thinking, knowing,
and valuing into their studies to better understand their phenomenon of interest. As such,
mixed method research can provide valid and credible inferences, along with
understandings that are broader, deeper, and wiser (Tashakkori & Teddlie, 2003). Mixed
method research is used in this dissertation in order to achieve more robust, contextual
ﬁndings. While quantitative statistics provide a “snapshot” of the general trends in
collateral IPH reporting, qualitative results triangulate and complement results with
textual evidence. Additionally, themes that would otherwise not be detected by a strictly
quantitative protocol are allowed to emerge, providing further insight into the

phenomenon.

Qualitative Content Analysis

Ethnographic content analysis (ECA) is the reﬂexive analysis of documents,
including newspaper articles. The strategy emerged ﬂom ethnography, (Schwartz &
Jacobs, 1979) although it was not considered a distinctive research method until the late
1980’s with Altheide’s (1987) seminal manuscript on the topic. Altheide stated that the
overall goal of ECA is discovery, as the researcher is involved at all stages of the project.
In fact, the progression ﬂom data collection, analysis, and interpretation is reﬂexive and
iterative.

Ethnographic content analysis is ﬂequently used to document and verify
theoretical relationships as well as to understand the communication of meaning in a
given set of documents. It is an interactive process, with the investigator playing a
central role in all phases of research. ECA is systematic and analytical, but not rigid. For

instance, theoretical assumptions and guidelines typically drive protocol design,

47

preliminary code categories, and variables, but other themes are allowed to emerge
throughout analysis (Altheide 1987; 1996). This is a distinctive strength because the
researcher is not restricted in their exploration, allowing the study to change directions at
any time.

As such, constant discovery and constant comparison are of paramount
importance in ECA, with the researcher paying close attention to situations, settings,
styles, images, meanings, and nuances (Glaser & Strauss, 1967). Scholars utilizing this
method collect both numerical and narrative data, all while testing theoretical claims.
ECA has the potential to reveal critical questions and issues that may otherwise go
unnoticed (Altheide, 1987; 1996). ECA is not oriented towards theory development;
rather its goal is to generate general descriptions or deﬁnitions compatible with the
selected media (Altheide, 1987; 1996). As such, it is utilized in this dissertation to
document theoretical themes that may be missed with a purely quantitative approach,
allowing the author to identify situations that align with the IPV theories under scrutiny:

coercive control and family violence theory.

Quantitative Content Analysis

Despite Altheide’s (1987) strong support for ECA, he states, “structured data
collection based on a protocol combined with ethnographic ﬁeld notes supports a
theoretically informed account of media content” (p. 74). Therefore, this dissertation
research also includes a quantitative content analysis (QCA), which has it roots in the
positivistic assumptions of objectivity (Berelson, 1966). Units of space are seen as

countable, thus measurable. In the words of Starosta (1984),

48

Content analysis translates frequency of occurrence of certain symbols into

summary judgments and comparisons of content of the discourse... whatever

‘means ’ will presumably take up space and/or time; hence, the greater that space

and/or time, the greater the meaning ’s significance (p. 185).

QCA is used in this dissertation to test hypothesized relationships (Krippendorf,
1980). Following the standardized design, research moves ﬂom category construction to
sampling, data collection, data analysis, and interpretation (Altheide, 1987). Individual
coders are then trained to ﬁnd, record, and count instances for each unit of analysis. The
overall emphasis for this portion of the study is reliability, with numerical data driving

analysis and interpretation.

Quantitative Coding Protocol

The quantitative coding protocol is located in Appendix B. It was constructed to
capture each of the concepts and variables introduced at the beginning of the chapter (see
section on deﬁnitions). A total of 44 major variables were measured in the instrument,
and divided into ﬁve major sections:

1. Story Identification and Newspaper Information — Basic identifying
information was collected in this section. Additionally, perpetrator and victim
story position were documented.

2. General Story Characteristics — Basic descriptive information regarding the
principal IPV victim, collateral victim(s), and perpetrator were captured in this
section, including the crime stage, types of crime victims, and occurrence of

suicide.

49

. Principal Victim Portrayals — This brief section collected data about the
principal IPV victim. Since they are not the primary focus of analysis, only sex
and whether he/she was killed in the story were collected.

. Collateral Victim Portrayals — This portion of the coding protocol
accommodates up to ﬁve collateral victims. The total number killed, their name,
age, and sex were noted. For each collateral victim, their speciﬁc relationship
with the principal victim and/or perpetrator was collected, as well as whether an
explanation was provided for the homicide(s).

. Homicide Portrayal — This section collected not only descriptive information on
the collateral IPH perpetrator (their sex), but their history of violence against the
principal and collateral victims as well. Additionally, how the homicide was
portrayed (shocking vs. expected) and whether the motivation was related to the
principal relationship was documented.

. Violence Symmetry - Stemming ﬂom family violence theory, the variable in this
section documented whether violence is described as a symmetrical (e. g. conﬂict
is apparent between the male and female principal partners) or asymmetrical (e. g.
dominated by one sex).

. Violence Cause — This section documented the proposed cause of violence (i.e.
individual or social factors), or the presence of causal attribution. Individual
factors include: relationship conﬂict, anger, mental health issues, or drug abuse.

Social factors include unemployment and ﬁnancial stress/ strain.

. Crime Context Information — Crime context information was collected in the

ﬁnal section of the coding protocol. In this area coders documented if the

50

journalist or any informants mention prior violence against the principal or

collateral victim. Additionally, if IPV (or any related term) was explicitly

mentioned in the story, coders marked “yes” and provided a relevant quote.

Lastly, coders made a ﬁnal judgment as to whether the collateral IPH was

presented as an isolated event or a continuation of past violence.

Each section was designed around the concepts and variables of interest. In
Appendix B the coding protocol is clearly segmented according to the above headings.

This format allows for intuitive coding and data analysis.

Quantitative Data Collection

Once the sample was assembled and sorted according to year and newspaper
source, two independent coders collected the quantitative data for this project. After
completing a rigorous training that included a thorough background of the study,
explanation of variables, and introduction to the coding protocol, two group-coding
sessions occurred. These sessions allowed the coders to talk through their thought
processes and resolve any disagreements immediately. Once these trainings were
complete, an iterative, rigorous reliability analysis was conducted in order to achieve
inter-coder reliability. The two coders underwent three rounds of testing. Because of the
small total sample (n=215), articles were randomly selected ﬂom all available
newspapers, as stratiﬁcation was not a viable option.

The reliability test included all potential study variables in which coders could
disagree. The variables were tested using simple percentage of agreement and Scott’s Pi
to correct for chance agreement. All variables that obtained 100 % agreement were

excluded ﬂom further analysis. The remaining study variables were expected to obtain

51

high levels of agreement, speciﬁcally .85 and above. This statistic is the expected norm
in the social and communication science communities (Schiff & Reiter, 2004; Zwick,
1986). However, due to small reliability sample size, those variables that failed to meet

.85 were not thrown out, but reviewed insteadz.

Qualitative Data Collection

Qualitative data were collected ﬂom the same dataset as quantitative, with a ﬁnal
sample size of n=215. However, instead of following a structured protocol designed to
capture concrete variables, a semi-structured data collection instrument was utilized to
capture theoretical-oriented ﬂames (i.e., family violence or coercive control theory).
Frames that ﬁt with Entman’s (1993) generic typology and emergent themes were also
documented. Data were recorded in the forms of quotes, and ranged ﬂom brief phrases
to entire paragraphs. The data collection process for this phase of the dissertation took
approximately three weeks, as it required multiple reviews of each article. Please refer to

Appendix D to review the qualitative data collection instrument.

Analysis

This following section introduces the steps involved in conducting both
qualitative data collection / analysis and quantitative analysis. Since the same sample of
newspapers was used in each phase, it is important to clearly articulate how each segment
was analyzed. Of critical importance in the qualitative segment was achieving

trustworthiness of the data (referred to “validity” in some qualitative texts) and a clear

 

2 There were three rounds of reliability testing for this project. The sample size for rounds two and three
was n=15. Even if coders disagreed only once, Scott’s Pi fell below 0.85. As such, the author reviewed the
variable and determined that percent agreement was acceptable.

52

data analysis plan. For quantitative analysis, inter-coder reliability was a necessity for

replication and future research.

Coder Reliability Analysis

Reliability analysis and training were conducted between April and June 2008,
requiring three rounds of testing with two independent coders. These particular coders
had already completed training and coded for another IPH content analysis project, so
they were well acquainted with the complete database. In the ﬁrst round, each variable
was tested, excluding only story identiﬁcation variables. A total of n=30 articles were
randomly selected ﬂom the entire database, or 14% of the entire sample. In the second
round, a total of twenty-four problematic variables were retested, as they fell below the
accepted Scott’s pi level of 0.85. In round three, ten variables were tested. For these
ﬁnal two analyses, coders completed n=15 articles each for the sake of expediency and
prevention of fatigue. As a result of this reduced sample size, six variables fell slightly
below the 0.85 level. However, in each of these cases there was only one disagreement,
with percent agreement over 90%.

It is important to note that this particular coding protocol allowed for ﬁve
collateral victims. However, in the reliability samples there were no cases in which
collateral victims met that total. As such, these variables were excluded ﬂom analysis
and ultimately removed ﬂom the protocol (the maximum number of collateral victims
was four in this database). Next, we found that collateral variables were redundant, with
the same homicide explanation offered across victims, so percent agreement and Scott’s
Pi statistics were averaged. Thus, analysis focusing on ﬁrst collateral victims can be

inferred to subsequent killings. The Luther Jenkins homicides justify this decision. He

53

killed his ex-girlﬁiend, her aunt, and her two children, leaving only his biological child,
Latavia, alive. In each article reporting on this “massacre”, the proposed explanation for
each collateral death was identical. Lastly, if 100% agreement between coders was

achieved Scott’s Pi was not calculated. Please refer to Appendix C for results.

Validity & Reliability of Qualitative Data

An important component to qualitative research is achieving credibility,
dependability, conformability, and transferability (Lincoln & Guba, 1985) throughout
data coding and analysis. A ﬁrst step to achieving this is setting aside preconceived ideas
and biases in order to fully participate in analysis with an open mind (Porter, 1998).
Frequently referred to as reﬂexivity, this activity promotes active reﬂection of the
investigator’s life experiences, allowing him or her to explore the connections between
ﬁndings and production / interpretation of results (Taylor & Bogdan, 1998). Therefore, it
is vital that researcher bias, motivation, and background are explicitly acknowledged so
that the reader can have a clear understanding of the implicit assumptions and biases held
by the investigator. For instance, prior to becoming a researcher, the dissertation author
worked as an advocate for survivors of sexual and domestic violence, participating in
crisis intervention and providing shelter services to female victims of IPV. This advocate
identity played a Signiﬁcant role in theoretical alignment and perspectives, as a clear
allegiance to feminist theory and methodology emerged early in the graduate career. As
a result, there is a strong personal motivation for conducting the current research —- giving
voice to victims of IPV and uncovering the gendered processes at work in IPV, a goal

that stems ﬂom feminist theory. It is important to keep these biases in mind as one

54

explores qualitative results and implications. As such, prior to analysis, memos regarding
theoretical alignment and potential biasing viewpoints were generated and shared with
the dissertation adviser. These memos served as context for interpreting and presenting
results.
T rustworthiness at the Findings

Inter-coder reliability was not achieved in this study, as there was not a second
coder to review and triangulate codes and themes generated throughout qualitative
analysis. Therefore, trustworthiness of ﬁndings (i.e.,, dependability and confonnability)
was not fully achieved; leaving the possibility those qualitative results could be
discredited. In order to fully assure reliability and validity, an additional individual coder
was needed to review transcripts and analysis matrices, as bias becomes increasingly
apparent as coding progresses. However, in lieu of a second coder, the dissertation
investigator reviewed coded transcripts and emerging themes, followed by conversations
about commonalities and disagreements. Although the study investigator solely
conducted qualitative analysis, by talking through ﬁndings as they emerged, triangulation
was achieved to a certain extent. Additionally, an audit trail was maintained (see
Rodgers & Cowles, 1993) that included methodological and analytical notes. Often
referred to as transferability, establishing trustworthiness of ﬁndings requires researchers
to achieve “a degree of similarity between sending and receiving contexts” (Lincoln &
Guba, 1985, p. 316). As such, the investigator provided a detailed description of how
meanings were created, as well as how context inﬂuenced gathering and interpreting data.
Moving through this process made transferability judgments possible, although not

methodologically complete.

55

Qualitative Analysis Procedures

Qualitative data collection and analysis focused on research question #2: Which
theoretical ﬂamework better describes the media narratives utilized to report collateral
IPH? AS is the case with quantitative analysis, the unit of analysis was the entire article,
however sentences served as primary illustrative evidence. The article was selected for
this phase of analysis because it maintained context (by not segmenting pieces of the
story) and allowed for evidence of family violence and coercive control theories to
emerge in one piece. Utilizing a combination of theoretical and open coding, dominant
themes were generated and catalogued (i.e. articles are categorized according to thematic
similarities). Since ECA is used to tests and validate theory, rather than develop it
(Altheide, 1996), a signiﬁcant portion of this phase focused on capturing the textual data
and its support of either family violence (conﬂict) or coercive control theory.
Additionally, the constant comparative method of Glaser & Strauss (1967) was
implemented, allowing the investigator to analyze the dataset numerous times, identifying
examples that emerged ﬂom the data as well as connect relevant variables and
hypotheses. Lastly, the author clustered the sample according to theoretical constructs
associated with either family violence or coercive control theory. Please refer to
Appendix D for the qualitative coding sheet used to facilitate qualitative data analysis.

As such, the steps involved in qualitative analysis were:

1. Segment and organize articles according to theory - The ﬁrst step of analysis
was to segment and organize articles according to (1) irrelevant data, (2) family
violence data, and (3) coercive control data. Irrelevant data were removed since

they were outside the scope of the dissertation, while other sentences were

56

categorized according to tenets of coercive control or family violence theory.
Qualitative matrices (Bazeley & Richards, 2000) were used to track and store
theoretical codes as well as document examples and emergent codes that
encompassed operational deﬁnitions, variables, and major themes.

2. Preliminary Theme Generation — Since qualitative data analysis is reﬂective and
iterative, this phase consisted of revision, re-coding, comparing, and generating
themes. Of particular interest were media ﬂames; speciﬁcally whether ﬂames
more ﬂequently supported family violence or coercive control theory. So by
using the constant comparative method, thematic codes were generated and
reﬁned.

3. Finalization — Upon completion of quantitative coding, theoretical ﬁndings and
themes that emerged throughout analysis were presented and shared with other
coders on the project for feedback and critique. The dissertation adviser also
reviewed major themes. Upon completion, themes were given a ﬁnal title /

categorization and description for results.

Statistical Tests & Procedures

Quantitative data analysis for this project occurred in two stages: (1) conducting
inter-coder reliability and (2) generating descriptive statistics. Data collected by the two
coders were subjected to Scott’s Pi (with an accepted agreement of approximately 0.85)
and simple percent agreement. Once reliability was achieved, and the coders collected
data ﬂom the entire sample, descriptive statistics and ﬂequencies were generated. Of

particular importance were sample characteristics (i.e. type of collateral victim most

57

ﬂequently murdered, gender of perpetrator, etc) and demographics. Major trends and
relationships in the data were also reported.

The concepts measured in this dissertation were: (1) victim / perpetrator story
position (i.e. the order in which perpetrator and victims are introduced by the reporter);
(2) homicide explanation (i.e., accident, result of ﬁght, or intentional); (3) perpetrator
history of violence / crime context; (4) homicide portrayal (i.e., shocking or expected);
(5) perpetrator motivation for homicide; (6) violence symmetry; (7) causal attribution;
and (8) media ﬂames. A total of forty-four variables were measured. Research questions
and hypotheses were addressed through both qualitative and quantitative content analysis.
To maintain clarity, each variable was assigned a number in the coding protocol and
results will be presented according to that number. Table 3.4 should act as a reference for
any variable-related ﬁndings or discussion.

Table 3.3. Numbered Variables included in Dissertation

 

Variable Number / Symbol Variable Description

 

Story ID Number

 

Story Date

 

Newspaper ID Number

 

UOW>

Sex of Reporter

 

Victim / Perpetrator Story Position

 

Crime Stage

 

Crime Victims

 

Perpetrator Suicide

 

Principal IPV Victim Sex

 

Principal IPV Victim Homicide

 

Number of Collateral Victims

 

Collateral Victim #1 Name

 

Collateral Victim #1 Age

 

Collateral Victim #1 Sex

 

ﬂﬂchQM-bWNH

HO

Collateral Victim #1 Relationship to IPV
Victim and/or Perpetrator

 

Homicide Explanation

 

Type of Explanation

 

bot—hi—I
NWN

Perpetrator Sex

 

 

 

 

33 Perpetrator History of Violence

 

Table 3.3 Numbered Variables included in Dissertation Continued

58

 

 

 

 

 

 

 

 

 

 

 

 

Variable Number / Symbol Variable Description
34 Perpetrator Prior Violence
35 Homicide Portrayal
36 Proposed Homicide Motivation
37 Violence Symmetry
38 Causal Attribution
39 Speciﬁc Causes of Violence
40 Prior Harm Against Pringipal IPV Victim
41 Prior Harm Against Collateral IPV Victim
42 Explicit Mention of IPV or Domestic
Violence
44 Incident-Speciﬁc Frame

 

 

 

The two major research questions, (1) what are the predominant trends in
collateral IPH and (2) which theoretical ﬂamework better describes the media narratives
utilized to report collateral IPH were addressed by coding speciﬁc variables included in
the quantitative protocol. Research Question 1a (who are the primary collateral victims
and perpetrators) and Hypotheses 1a (the primary collateral perpetrator will be the
current or ex male partner) and lb (the primary collateral victim will be the IPV Victim 's
children) were answered by calculating ﬂequencies of collateral victim type (e.g.,
Variable l 1; see Appendix B, pg. 143) and the perpetrator’s relationship to them. This
variable was assessed primarily through quantitative means.

Research Question 2a (why does collateral IPH occur) and Hypothesis 2a
(journalists more ﬁ'equently report that the homicide is a result of relationship conﬂict)
were answered by analyzing the proposed homicide explanation (e. g. Variable l3),
homicide portrayal (Variable 35), proposed motivation for the homicide (Variable 36),
and causal attribution (Variable 39).

Research question 2b (Is the collateral IPH reported as unexpected or

premeditated) and hypothesis 2b (Journalists more frequently report collateral IPH as an

59

unexpected crime) were answered by analyzing the homicide portrayal variable (Variable
35).

Research Question 20 (does the proposed explanation of the homicide impact
reporting frames) and Hypotheses 2c (if the collateral homicide is reported as
unexpected, the proposed cause will be conﬂict) and 2d (if the collateral homicide is
reported as premeditated, the proposed cause will be ongoing IPVor coercive control
were answered by ﬁrst investigating the relationship between homicide portrayal
(Variable 35), the proposed motivation for the homicide (Variable 36, option a), and
violence attributed to individual factors (Variable 39). Second, the relationship between
homicide portrayal (Variable 35) and the mention of prior harm or abuse (Variable 40)
and an explicit statement of IPV or domestic violence (Variable 42) was analyzed.

Research Question 2d (does the type of victim impact reporting) was addressed by
breaking collateral victims into three groups: (1) children, (2) other family members, and
(3) new boyﬁiends or love interests. Then, the ﬁrst collateral victim in every article
(Variable 11) was matched up with the proposed explanation (Variable l3) and
ﬂequencies. This question was exploratory and can only provide descriptive statistics,
however provide evidence of dominant homicide explanations can be gleaned. Lastly,
Variable 1 provided insight into research Question 2e (where are collateral victims

described in the article).

60

Chapter 4: Results

Sample Characteristics

The quantitative portion of analysis included a sample size of n = 208, as a
handful of articles were excluded because they did not ﬁt protocol design. For example,
stories in which multiple perpetrators (e. g. father and his children) were implicated and a
case where the collateral child killed the perpetrator while he was beating his mother
were not analyzed because the coding protocol only allowed for one perpetrator and did
not provide an option to select the perpetrator as a victim. The entire database (n=215)
was coded in the qualitative analysis because the protocol was more ﬂexible in this phase
of the project and we wanted to include every possible story to compensate for the
rigidity of quantitative data collection. However, before addressing research questions,

general story characteristics, both quantitative and qualitative, are introduced.

Newspaper Information and General Story Characteristics

Like other homicides, the reporting of collateral IPH follows a fairly predictable
pattern. Once a murder occurs, police are the ﬁrst to investigate and speculate on the
nature of the crime. The journalist then conducts their own investigation by interviewing
police representatives, neighbors / bystanders, friends, coworkers, and family members.
Because an article’s particular focus depends on the status of the police investigation
(e. g., a suspect has been identiﬁed, arrested, or sentenced), each story varies according to

these issues.

61

Since many articles were news briefs or Associate Press (AP) articles, 37% of
journalists could not be identiﬁed by their gender. However, 36.1% were determined
male and 25% female, with only 1.9% of articles written by members of both sexes. The
predominant crime stage reported on was judicial proceedings, a phase that highlighted
events occurring ﬂom the perpetrator’s trial to sentencing, psychiatric evaluations, and
legal appeals (41.8%). These stories were short, and primarily included quotes ﬂom
prosecutors, defense attorneys, perpetrators, and other relevant witnesses, including
police and medical examiners. A ﬁrst sentence in an article on this stage might say, “A
30-year-old Detroit man accused of killing his girlﬁiend and infant daughter in Pontiac
last month is to undergo a forensic evaluation to determine whether he is competent to
stand trial.” Crime commission through arrest represented 37.5% of the articles (e.g.,
“Nikkie Sue Sewell, the wife of the suspect, and Ypsilanti Wayne Jones, 32, were both
found shot to death shortly after 2am in a Mott Road Mobile home”), while “aftermath”
accounted for 20.7%. An article was considered aftermath if it discussed the implications
or outcomes of the collateral IPH, including family reactions to sentencing or funeral
services.

Aftermath is an incredibly complex crime stage, as it encompassed a diversity of
testimony and witnesses, with journalists infusing text with graphic and / or emotional
descriptions of the crime and its impact on those affected. Qualitative examples,
organized in an ecological structure, illustrate this category.

0 Micro Level — Family and ﬂiends’ reactions to the homicides and its impact on
their lives. This level was typically seen in stories that report on funerals or those

written to honor the deceased. For example, in remembering Jim Watson, a man

62

killed by his daughter’s boyfriend, loved ones noted, “Our family has experienced
a horriﬁc loss. . .we ask for your prayers and understanding of wanting to grieve
privately.”

0 Meso Level — This level typically encompassed neighborhood, community, or
coworker reactions to the collateral IPH. There was signiﬁcant focus on its
impact on neighborhood children or how the homicide(s) represented an act of
community violence. As with the micro level, stories associated with the meso
level ﬂequently covered funerals, but offered community-focused responses. For
example, “Hundreds of people, some sobbing and wailing, ﬁled by ﬁve ﬂower-
laden caskets to pay their respects to the victims of a brutal mass killing that
stunned the city.”

0 Macro Level — This level of aftermath reporting introduced policy or societal
implications of the collateral IPH, inﬂequently including a discussion of IPV or
domestic violence (e.g., the interviewing of domestic violence experts or
presentation of state-level statistics). “Law is a lot better than it used to be in
Michigan, Lucas said, especially with the stalking and domestic violence laws,
but there are still improvements that need to be made to take every advantage of

those laws.”

Unique Collateral IPH Crime Reporting Techniques

Because of the shocking nature of collateral homicides (e.g., deaths of children or
older individuals), articles reporting at all crime stages tended to highlight the graphic or

emotional nature of the events. Journalists ﬂequently incorporated emotional reactions

63

ﬂom neighbors, family members, ﬁiends, and other witnesses in order to better share the
stories of collateral and / or principal IPV victims. For example, in a “startling” homicide
of ﬁve people including a 16-year-old boy, 14-year-old girl, l3-year-old girl, and 9-year-
old girl, an article ﬂom the Alpena News provided emotional depictions of neighbors:
0 “At least two women collapsed (outside the family’s home), and paramedics were
summoned.”
o “A group of ﬁve women who approached police began moaning loudly after
ofﬁcers spoke quietly to them.”

Additional examples of emotional or graphic reporting styles include the
following: [be] “. . .walked toward the burned bungalow where his grandchildren died and
where teddy bears adorned the ﬂont porch. He soon became overcome with grief and
collapsed into the arms of ﬂiends near the ﬂont porch.” The surviving child in this case,
Antonia Bailey, told police “My daddy killed me. . .he killed my whole family.”

In a follow-up story, family reactions in the courtroom provided striking detail.
Anthony Bailey, the father and murderer of three of his children, “could barely stand to
hear his wife’s cries as he stood Thursday for his arraignment, charged with killing three
of their children. The story said, “Bailey tilted his head back — his eyes welling with
tears — as he heard his estranged wife, Essie Bailey, sobbing through the sound system in
36th District Court. Family members eventually had to carry Essie Bailey out of the
courtroom.”

In another case, Daniel Franklin “slashed and stabbed” an eight and ten-year-old,
leaving their youngest sibling, four-year-old D’anajeh alive. According to the article,

D’anajeh “told investigators ‘Daddy got a knife’ and described how one sister, before she

64

died, begged to be allowed to say goodbye to her mother.” Similarly, a story on Luther
Jenkins, charged with killing four people, reported the crime scene in vivid detail. “Inside
the house were signs of a struggle and blood in the living room and upstairs. Police also
found a mop, bucket and the strong smell of ammonia — signs that a cleanup had been
attempted.” A separate article noted, “Late Sunday, two dozen stuffed animals sat on the
porch of the home, along with an unlit candle. The Sunday newspaper was untouched. A
trimmed Christmas tree stood in the living room window, and there were kids’ toys on
the porch bench. No one answered the door.”

Lastly, in the J immie Reed Jr. case, the Flint Journal incorporated descriptive
police testimony into a story on crime commission. Reed shot his girlﬂiend, and then set
her and their daughter, Arctavia Reed, on ﬁre. “Police say the Detroit man shot Markeda
Byas, 31, in the head at Byas’ apartment in Pontiac on July 20. He left for work, returned
with a can of gasoline, doused her and daughter Arctavia Reed and set them aﬁre early
the next day, police say.” Police detective Darrin McAllister went on to add, “He told us
the baby woke up screaming.”

In sum, while articles follow a typical reporting pattern that progresses ﬂom arrest
to judicial trial and aftermath, the uniqueness of this crime, speciﬁcally the deaths of
children and other important loved ones, facilitate graphic and / or emotional reporting
that ﬂequently utilize numerous witnesses and rich descriptions. Thus, these particular
examples were presented at the beginning of the chapter in order to orient the reader to
some of the primary reporting characteristics associated with collateral IPH newspaper

articles.

65

Research Question #1

The most frequently occurring type of victim (see Variable 3) was the current /
ex-partner of the perpetrator (referred to as the principal IPV victim). A crime victim is
any individual that was reported as injured and / or killed in an article. This variable was
intentionally left broad so that an overall count of the most ﬂequently reported type of
victim could be captured, however suicide was not included. Interestingly, although she
was the most predominant crime victim, she was not always killed. In fact, only 57.7% of
principal victims were killed during the commission of a collateral IPH. Table 4.2
presents all victim types and their ﬂequency of mention.

Table 4.1 Total Crime Victims

 

 

 

 

 

 

 

 

 

Victim Type N (total mentions) Percentage
Principal IPV Victim 151 53.9%
Family of IPV Victim 69 18.2%
Children 67 17.6%
New Boyﬁiend or Partner 44 1 1.6%
Friends of IPV Victim 26 6.8%
Neighbors / Bystanders 12 3.2%
Other 11 2.9%
Total 380 100%

 

 

 

 

 

Table 4.2 Collateral Crime Victims

 

 

 

 

 

Victim Type N (total mentions) Percentage
Principal IPV Victim’s Child 82 26.1%
Child of IPV Victim and 57 18.2%
Perpetrator
New Boyﬂiend or Partner 45 14.3%
Other Family Not Speciﬁed as 41 13.1%

Parent or Child (e.g., sisters,
brothers, aunts, etc)

 

 

 

 

 

Principal Victim’s Mother 29 9.2%
Friend 28 8.9%
Principal Victim’s Dad 18 5.7%
Other 14 4.5%
Total 314 100%

 

 

 

 

 

66

The ﬁrst research question of this exploratory study was: “who are the primary
collateral victims and perpetrators?” This question was intentionally broad, as no
preliminary evidence has been collected on this topic. However, existing research on
intimate partner violence and the nature of coercive control informed hypotheses, as it
was predicted that the primary collateral perpetrator will be the current or ex male
partner and that the primary collateral victim will be the principal IPV victim 's child.
The logic behind these hypotheses — men commit a majority of intimate partner violence
and inﬂict more injury (Straus et al., 2006; Tjaden & Thoennes, 2000a) than women.
Women are more likely to suffer ﬂom IPV. Additionally, according coercive control
theory, abusive men operate along a spectrum of control tactics designed to hurt and
immobilize his partner (Stark, 2007). Thus, at the extreme end of this spectrum is IPH
(see Garicia et al., 2007). However in this case the IPH is directed at the person his
partner values most — her child. This extension of IPH and line of reasoning have yet to
be empirically tested, so exploratory results are presented below.

In this study, principal IPV victims were all women (100%), while all collateral
IPH perpetrators were men (100%). This ﬁnding contradicts that of general population
surveys (e. g. Garcia et al., 2007; Straus et al., 2007) and typical media coverage of IPV
and IPH (Biroscak, Smith, & Post, 2006). This result represents a rare case in which
there is a clear gender divide between perpetrators and victims; even ferrrinist surveys do
not uncover such strong gender divisions. However, principal IPV victims were not
always killed, and it was not a criterion to be included in this statistic. This individual

was simply someone who had a preexisting relationship with the man that committed the

67

collateral homicide and were described as attacked and / or injured in the article. Thus,
there is a salient difference between IPV and IPH — IPV indicates violence or abuse in the
principal relationship (not resulting in death), while IPH is the killing of an intimate
partner (and in this study a collateral individual). Evidence of intimate partner violence,
however, was not always clear in news stories as only 24% of the articles explicitly
mentioned prior violence, with the two most common issues being violence against the
principal IPV victim (n=49) and the issuance of personal protection orders (n=11).

Interestingly enough, perpetrators rarely committed suicide after the murder(s),
with only 18.3% taking their own lives. This is also an unusual ﬁnding, as a large
percentage of men kill themselves in traditional instances of IPH (Campbell et al., 2007).
This result leads one to further conclude that collateral IPH is somehow different ﬂom
general IPH. When individuals outside of the primary intimate relationship are involved,
overall dynamics change.

The maximum amount of collateral victims killed in any one story was four,
although the average across all articles was two. The mean age of homicide victim was
15.98, with the sample containing similar levels of male and female collateral homicide
victims (n=157 and n=l 77 respectively). Table 4.3 summarizes the types of collateral
victims included in this dissertation sample, with the most commonly reported victim
being the principal IPV victim’s child.

Exploratory hypotheses supported this research question. Every perpetrator in
this sample was male with an explicitly stated relationship with the female principal IPV
victim. Thus, even though the principal victim was not killed in every case, she played a

clear role in the collateral IPH because each collateral victim was related to her in some

68

way (i.e., child, parent, ﬂiend, etc). There was not a single case of the perpetrator killing
his parent, friend or other family member. Only children he shared with the principal
victim were mentioned. Additionally, as originally hypothesized, the primary collateral
victim was the principal IPV victim’s child (one that the perpetrator had no biological
connection to); a ﬁnding that lends support to the coercive nature of collateral IPH (see
Stark, 2007).

This result also separates collateral IPH ﬂom traditional IPH, as these are
characteristically different homicides that do not always end with the death of the
intimate partner. The current deﬁnition of IPH, or “the most extreme outcome of IPV”
(Garcia et al., 2007, p. 370) focuses exclusively on the death of one intimate partner.
Researchers must expand this deﬁnition if they intend to explain the killing of others,
especially if the principal victim is left alive. This is where coercive control theory
offers insight: If an abuser wants to hurt or punish his victim, killing her child represents

an extreme outcome of IPV.

Research Question #2

The second research question, “Which theoretical ﬂamework better describes the
media narratives utilized to report collateral IPH”, was tested via qualitative and
quantitative analysis. This particular question contains a series of sub-questions and
hypotheses, each of which were directly addressed through quantitative means. While
qualitative analysis was more theory-focused, both procedures utilized Entrnan’s (1993)
generic ﬂaming typology as a starting point. In the quantitative section, variables sought
to delineate conﬂict ﬂom premeditation (e.g., Variable 7), history of violence (Variables

33 and 34), homicide portrayals (Variable 35), the presence of mutual conﬂict or ﬁghting

69

(Variables 36 and 37), symmetry, and causal attribution (Variable 39) in order to
determine appropriate problem deﬁnitions and crime causes set forth by Entrnan.
Findings ﬂom the quantitative portion of analysis are presented prior to qualitative
results. This particular order was selected because the surface trends detected by
quantitative analysis form the foundation of descriptive examples introduced in the
qualitative section. However, relevant statistical ﬁndings will be highlighted as

necessary throughout qualitative results.

Quantitative Results

The ﬁrst sub-question addressed in this section is “why does collateral IPH
occur?” In order to address this question, articles were coded according to the primary
explanation offered for the homicide(s) (see variables 35-42). In order to provide a
complete comprehensive picture of proposed explanations, only results ﬂom the ﬁrst
collateral victim are presented, as they contain complete data (n=208). Due to missing
responses for collateral victims 2-4, they have been excluded ﬂom analysis and are
redundant if provided. Only seventy-four articles reported two victims (35.6%), thirty-
seven reported three victims (17.8%), and ﬁfteen reported four victims (7.2%).
Additionally, the explanation provided for collateral victim #1 was constant across all

subsequent victims, so removing these cases did not affect results.

70

Table 4.3 — Homicide Explanations and Portrayals

 

Proposed Homicide Explanation % Of Articles

 

Collateral IPH was a result of singular conﬂict or 51.9%
ﬁght within the principal relationship (not
intentional)

 

Collateral IPH wagpremeditated or intentional 5.3%

 

No proposed explanation included in article (news 35%
brief, AP articles)

 

 

 

 

 

 

Homicide Portrayals °/o Of Articles
Collateral homicide was expected 9%
Collateral homicide was shocking 31%
Unclear 60%

 

 

In order to address this ﬁrst research question, the following hypothesis was
posited: Journalists more frequently report that the homicide is a result of a singular
relationship conﬂict or altercation. This explanation directly supports the tenets of
farme violence theory — IPV occurs out of relational conﬂict and ﬁghting. Results
support this hypothesis, with 51.9% of articles indicating that the collateral IPH was a
result of a principal relationship conﬂict or ﬁght. While 35% had no apparent
explanation, only eleven articles (5.3%) concluded that the killings were intentional or
premeditated. Additionally, homicide portrayal, or how the murders were interpreted and
explained by witnesses and other involved parties, (i.e., neighbors, ﬁiends, police, etc)
rarely indicated that the homicides were expected (9%); rather shock was the primary
response offered (31%). This particular variable highlights the lack of context or
background information on preexisting abuse. In fact, 60% of articles were “unclear”, as
a concise understanding of the perpetrator and their reasons for committing the
homicide(s) were not included in the article. Please refer to Table 4.4 for a summary of
results.

In further support of this hypothesis, 60% of the articles indicated that the

proposed motivation for collateral IPH was related to the principal relationship, or some

71

issue stemming ﬂom it (i.e., divorce, separation, family ﬁght, etc). 65.4% of stories
contained some sort of causal attribution (i.e. that a speciﬁc problem precipitated the
killings), with 63% attributed to individual factors (e.g., relationship conﬂict, anger,
mental health issues, drug abuse, etc). Please refer to Table 4.4 for a summary of these
results.

Table 4.4 Proposed Motivation & Causes of Collateral Homicides

 

 

 

 

 

Homicide Explanation Percent of Articles
Collateral IPH stems ﬂom issue in principal relationshiL 60%
Causal attribution contained in article 65.4%
Causal attribution attributed to individual factors (i.e., 63%
relationship conﬂict, anger, mental health problems, drug
abuse)

 

 

 

A second research question, “is the collateral IPH reported as unexpected or
premeditated” and its related hypothesis, journalists more frequently report collateral
IPH as an unexpected crime, are also supported by exploratory ﬁndings articulated in the
above paragraph. Although a majority of articles remain unclear (60%), only 9% claimed
that the murders were deliberate or expected while 31% highlighted the shocking and
unexpected nature of the crime. This leads to a preliminary conclusion that if
explanations are offered, collateral homicides are more ﬂequently reported as unexpected
or shocking.

To address the next sub-question and associated hypotheses, a chi-square based
measure of association, phi, was calculated to determine how closely particular nominal
variables were related. Although the researcher learns nothing about the strength and

direction of the association (Norusis, 2002), we can gauge a non-association (<1)=0) and a

perfect association ((I>=1).

72

Table 4.5 Unexpected Nature of The Crime (Variable 35a) and Proposed Conﬂict

 

 

 

 

 

Q’ariable 36a)
36A 36A Total
0 1
35a 0 Count 64 79 143
Expected Count 57.8 85.2 143
% Of Total 30.8% 38.0% 68.8%
35a 1 Count 20 45 65
Expected Count 26.2 38.8 65
% Of Total 9.6% 21.6% 31.2%
Total Count 84 124 208
Expected Count 84 124 208
% Of Total 40.4% 59.6% 100%

 

 

 

 

 

Table 4.6 Unexpected Nature of the Crime (Variable 35a) and Causes of Violence

 

 

 

 

 

 

 

 

 

 

(Variable 39)
Var. 39 Var. 39 Var. 39 Var. 39 Total
0 1 2 3
35a 0 Count 56 84 0 3 143
Expected Count 49.5 90.1 .7 2.8 143
°/o Of Total 26.9% 40.4% .0% 1.4% 68.8%
35a 1 Count 16 47 1 1 65
Expected Count 22.5 40.9 .3 1.2 65
% Of Total 7.7% 22.6% .5% .5% 31.2%
Total Count 72 13 1 1 4 208
Expected Count 72 131 1.0 4.0 208
% Of Total 34.6% 63% .5% 1.0% 100%

 

 

First, the question, “does the proposed explanation for the homicide impact reporting
frames” necessitates a comparison between the “homicide portrayal” variable (Variable
35), “principal relationship” variable (Variable 36), “individual causal attribution”
(Variable 39), the statement of prior harm or abuse against the principal IPV victim
(Variable 40), and explicit mention of IPV or domestic violence (Variable 42). As such,
the ﬁrst hypothesis, if the collateral homicide is reported as unexpected (Variable 35a)
the proposed cause will be relationship conﬂict (Variable 36a), yielded the following
statistic: (I) = 0.132 (see Table 4.5). This ﬁnding demonstrates that there is a modest
association between reporting a homicide as unexpected and subsequent attribution of

relationship conﬂict. The association between 35a and 39 revealed, (D = 0.174, indicating

73

a moderate relationship between the reporting of the collateral [PH as shocking /
unexpected and individual causes of violence (which includes relationship conﬂict). This
relationship is noteworthy because Variable 39 incorporates additional explanations for
the murders, including mental health problems and drug abuse.

Table 4.7 Intentional Nature of the Homicide (Variable 35b) and Mention of Prior
Abuse Against Principal IPV Victim (Variable 40)

 

 

 

 

 

 

 

 

 

Var. 40 Var. 40 Var. 40 Total
0 1 2

35b Count 144 7 38 189
Expected Count 139 6.4 43.6 189

% Of Total 69.2% 3.4% 18.3% 90.9%
35b Count 9 0 10 19
Expected Count 14 .6 4.4 19

% Of Total 4.3% .0% 4.8% 9.1%
Total Count 153 7 48 208
Expected Count 153 7 48 208

% Of Total 73.6% 3.4% 23.1% 100%

 

Table 4.8 Intentional Nature of the Homicide (Variable 35b) and Explicit Mention

 

 

 

 

 

 

 

 

of IPV (Variable 42)
Var. 42 Var. 42 Total
0 1
35b Count 161 28 189
Expected Count 158.1 30.9 189
% Of Total 77.4% 13.5% 90.9%
35b Count 13 6 19
Expected Count 15.9 3.1 19
% Of Total 6.2% 2.9% 9.1%
Total Count 174 34 208
Expected Count 174 34 208
% Of Total 83.7% 16.3% 100%

 

 

74

 

Table 4.9 The Explicit Mention of IPV (Variable 42) and Prior Abuse Against

Principal IPV Victim (Variable 40)

 

 

 

 

 

 

 

 

 

Var. 40 Var. 40 Var. 40 Total
0 1 2

Var. 42 0 Count 147 6 21 174
Expected Count 128 5.9 40.2 174

% Of Total 70.7% 2.9% 10.1% 83.7%
Var. 42 1 Count 6 l 27 34
Expected Count 25 1.1 7.8 34

% Of Total 2.9% .5% 13% 16.3%
Total Count 153 7 48 208
Expected Count 153 7 48 208

°/o Of Total 73.6% 3.4% 23.1% 100%

 

 

The next hypothesis, if the collateral homicide is reported as premeditated or
intentional, the proposed cause will be IPV or domestic violence, stems from coercive
control theory and the connection between the calculated nature of IPV and intimate
partner homicide (i.e., the premeditated nature of the crime) and the connection to
ongoing abuse. Results, which are summarized in Tables 4.7, indicate a moderately
strong association between the deliberate and expected nature of the crime (Variable 35b)
and the mention of prior harm or abuse in the principal relationship (Variable 40), with (I)
= 0.225. This ﬁnding provides convincing evidence that there is an association between
reporting collateral IPH as deliberate and the introduction of prior violence against the
principal victim. Unfortunately, only n=5 articles in the dataset mention prior abuse
against the collateral victim, leading one to conclude that the principal victim, not the
collateral, is the primary impetus for the homicide, an issue that will be discussed at
length later.

The association between Variable 35b and Variable 42 (the explicit mention of
IPV or domestic violence in the article), summarized in Table 4.8, had a relatively weak

association bordering signiﬁcance (CD = 0.131). However, in the process of calculating

75

this statistic, a strong association between the mention of prior abuse (Variable 40) and
an explicit statement IPV or domestic violence (Variable 42) in the article was detected
see Table 4.9). With an (I) = 0.593, it is clear that there is an association between
reporting evidence of abuse in the principal relationship and including a discussion of
IPV. Thus, the above hypotheses are generally supported because there is a noteworthy
connection between the perceived intentional nature of the homicide and introduction of
past violence.

Table 4.10 - Victim Type Impact on Reporting Homicide as Intentional or Conﬂict-
Oriented

 

 

 

 

 

 

 

Victim Type N Percent of Deaths
Intentional Explanations

All Children (Principal Victim’s Child or 3 4.4%
Child of Both Principal Victim and
Perpetrator)
Family Members 2 3.4%

Conflict Explanations
New Boyﬁiend or Love Interest 32 74.4%

 

 

 

 

Findings for the next research question are summarized in Table 4.10. It was
posed in an attempt to understand if and in what ways collateral IPH is reported in
Michigan newspapers (“Does victim type impact reporting”?) In other words, does it
make any difference if children are killed as opposed to other family members or new
boyfriends? This question, although initially intended to be quite complex, ultimately
had a simple answer. . .no. Out of 208 articles, only 11 reported the homicides as
intentional, so it was nearly impossibly to glean speciﬁc results based on victim type.
For instance, if children were identiﬁed as collateral victims, regardless of parent
(victim’s child, perpetrator’s child, or child of both), only three of sixty-eight were

attributed to premeditated, intentional murder, or 4.4%. Likewise, only two family

76

deaths (e. g., principal IPV victim’s mother, father, siblings, etc) were determined as
intentional, or 3.4%. Interestingly, none of the outside family killed were related to the
perpetrator, with the exception of children he shared with his partner / ex-partner.

Lastly, it was originally hypothesized that if new boyfriends or love interests were
killed in the article, their deaths would be attributed to principal relationship conﬂict,
including jealousy, divorce, or other “domestic disputes.” This was obviously supported,
since 95% of the articles alluded to relationship conﬂict as the primary cause of collateral
IPH. Speciﬁcally, 74.4%, or thirty-two of the forty-three new partner deaths were linked
back to principal relationship conﬂict. These descriptive statistics are counterintuitive to
a certain extent, as one might suspect that reporters and witnesses would be more
accusatory in cases of child death, but that was not the case. However, if children were
killed, there were unique characteristics associated with these stories; these themes will
be described in the qualitative results section.

The ﬁnal research question addressed in the quantitative portion of this section
focused on the introduction of key players in the homicide: the principal victim, collateral
victim, and perpetrator. This question was concerned with where collateral victims were
introduced in relation to the principal couple (principal IPV victim and perpetrator) so
that preliminary insights into how they “ﬁt” in the story could be obtained. As such,
coders docmnented their order of introduction in each article in order to detect overall
trends. Interestingly enough, over half (66.5%) of the articles maintained the following
order: perpetrator, principal IPV victim, and collateral IPV victim. 28.1% presented the
perpetrator ﬁrst followed by the collateral IPV victim and principal IPV victim, while

5.4% ﬁrst introduced the collateral followed by the perpetrator and principal IPV victim.

77

Although one can only provide basic reactions to this question, it offers insight into how
a collateral victim was incorporated into reporting of this complex crime. Since nearly
67% of the articles mentioned the members of the principal relationship ﬁrst, one can
cautiously conclude that the connection between these two individuals was ﬁrst
established before mentioning the additional victim, supporting the notion that they are,

in fact, “collateral damage.”

Qualitative Results

The use of open, iterative coding facilitated the qualitative analysis, although the
coding protocol was designed in accordance with dominant media framing characteristics
and IPV theory assumptions. As a result, many themes and trends go beyond the results
gleaned in quantitative analysis alone. Thus, the current section introduces major
ﬁndings, an endeavor primarily driven by research question #2 (“which theoretical
framework better describes the media narratives utilized to report collateral IPH”). As
such, this section begins by introducing the predominate categories of frames set forth by
Entrnan (1993). These include the (1) problem deﬁnition, (2) causal attributions, (3)
moral judgments and (4) potential remedies.

Generic Frame #1 — Determining the Problem

As one might intuitively expect, the main problem presented in each article was
incident-speciﬁc, with the homicide(s) serving as the primary problem of interest. In
fact, 82.7% of the articles were incident-speciﬁc, while only 17.3% (n = 36) included a
connection to past relationship problems or abuse. Although a handful of highly
publicized cases carried over into subsequent follow-up stories (usually when a person

was of high status or the crime was exceedingly gruesome), the focus overwhelmingly

78

remained on the killing(s) and crime commission. For instance, the murder was
introduced to the reader within the ﬁrst two sentences of each article. Stories that
addressed a “startling homicide”, a “double homicide”, or an escalating marital problem
that resulted in “an assault of a former girlfriend and the death of her ﬁiend” were
frequent. The collateral victim was rarely part of the problem deﬁnition, as they were
mentioned last in over half of the articles (53.4%).

Rather, the “problem” was an issue tied to some event within the principal
relationship. This is an important ﬁnding because although the principal IPV victim was
not the focus of the current analysis, they remained a focal point of reporting at all levels,
including problem deﬁnition. This ﬁnding supports quantitative results presented earlier.

For example, problems were frequently presented within the context of the
principal relationship, such as a boyfriend killing his partner’s 8-month old child because
he could not handle the termination of their relationship or killing a ﬁend, acquaintance,
or signiﬁcant other because he suspects he is the wife’s new boyﬁ'iend. Although speciﬁc
causes and motivations for the homicides will be discussed later, it is important to note
that this trend began at the problem deﬁnition stage.

If the collateral IPH was ever tied to problems beyond the principal relationship, it
was typically related to (1) a current wave of community violence (2) other highly
publicized homicides, or rarely, (3) a history of family or domestic violence. Since many
of the interviewed informants included neighbors or other local community dwellers,
community violence issues frequently encompassed the collateral homicide. In a Detroit
Free Press article about the killing of a girlfriend and her mother, the reporter stated that

the deaths represent a “surge of violence”. They go on to claim “there have been 408

79

homicides in the city this year, not including this morning’s, a 14% increase from 2005’s
total of 359.” While describing a homicide that occurred in the workplace, a journalist
noted, “It was the third time in four months that gunﬁre has echoed through a Metro
Detroit auto plant. On Dec. 9, a worker shot a supervisor and wounded a coworker at a
Chrysler Corp plant in Sterling Heights.” In response to the event, a Ford spokesman
said, “While the issue comes up from time to time, it would be impractical to put metal
detectors at plant entrances”, a statement that strips the killings of any intimate
connection. Another illustrative quote demonstrated the blurring of intimate and
community violence: “The killings are fresh reminders of the rampant violence that has
made Detroit one of the nation’s most dangerous cities.” When one reads articles framed
in such a manner, it is difﬁcult to discern if the problem is intimate partner violence,
workplace violence, or community violence; each situation is distinct from the other with
varying origins and implications.

There were a handful of cases in which the crime was tied to other, more highly
publicized crimes. The OJ Simpson homicides are an excellent example. When Worden
Gray killed his wife, Pamela Gray, and her new boyﬁiend, Clinton Leonard, The
Oakland County Legal Review made the connection to the OJ case, claiming that Worden
“has come to be identiﬁed with OJ Simpson...” They go on to note, “Gray was convicted
three days before Simpson’s ex-wife, Nicole Brown Simpson, and her friend, Ronald
Goldman, were found stabbed to death outside her Los Angeles home.” However,
“despite some parallels between the Simpson and Gray cases, no evidence of abuse was
presented during Gray’s trial.” Similarly, in an article written by the Saginaw News, a

well-known local case was tied to the collateral IPH being reported. “The slayings

80

reminded Goyt of a massacre in Clare County that killed seven in 1982. Four adults,
including the convicted killers estranged wife, and three children were killed in that
spree.”

A small percentage of articles, (17.3%), explicitly mentioned a history of
domestic violence within the principal relationship. This issue spills over into causal
diagnoses and other incident-speciﬁc themes, but as expected in qualitative analysis,
many code categories intersect. For example, in the case of Omar Dean, a man who killed
his 9-day old son, the surviving principal IPV victim related an extensive history of
violence and abuse: “Dean had kicked her in the stomach and blackened her eye when
she was 7 ‘/2 months pregnant, and that Omar Dean Jr. was born with a broken
collarbone.” She goes on to say that “I should have left him a long time ago. My baby
would be living now. I just hate that my baby had to come between us for me to open my
eyes.” In a separate case found in the Coldwater Daily Reporter, a man killed his three
children and then himself in a house ﬁre. Information about prior violence and abuse
were provided. “Hicks (the perpetrator), 41, threatened to steal his children, blackened
the eye of his wife on one occasion and subjected her to a ‘history of physical abuse,
mental abuse and violent behavior’”. Lastly, in the Taylor / Fochtman collateral IPH case
reported in the Detroit Free Press, in which Fochtman’s ex-boyﬁiend stabbed her and
killed two of her ﬁiends, a history of abuse was attributed to the homicides (problem).
“Fochtman said her relationship with Taylor was marred by violence during the three
years they lived together. F ochtrnan said she called the police on numerous occasions.”
Interestingly enough, the stories where a history of IPV is introduced are those in which

the principal victim survives the attack and is able to present testimony about the

81

perpetrator’s abusive past. Unfortunately, only 41.8% of principal IPV victims lived
through the attack.
Generic Frame #2 - Causal Diagnosis

Two major causes of collateral IPH causes emerged throughout analysis: (1)
domestic dispute gone awry and (2) the result of perpetrator-speciﬁc problems, both of
which lend support to family violence theory. The ﬁrst category was incredibly complex
and convoluted, as it encompassed many overlapping causes. As a result, Figure 4.1
depicts the themes included in this category. Cases presented for this ﬁrst theme
ﬁ‘equently highlighted the same perpetrator multiple times because examples fell into
various subcategories. Thus, for the sake of concise, clear reporting, each sub-theme will
be described individually. The subcategories associated with each major theme are
presented in Figures 4.1, 4.2, and 4.3

Figure 4.1 Domestic Disputes Category

 
  
 

  

Ambiguous
Relationship
Problems

 
   
 

Domestic He was out

Dispute

of control

 

    

  

If I can’t
have you,
no one
else will

   
   
   
 

82

It was a Domestic Dispute — Child Custody and Visitation Issues

A majority of articles discussed a “domestic dispute” or some form of
symmetrical ﬁght or conﬂict that triggered the collateral homicides. Causes included
marital separation, impending divorce, cheating spouse, or a custody / visitation dispute
with shared children. Each of these events served as an impetus that ultimately led to a
family ﬁght, argument, and overall stress on the perpetrator. For example, in one case
where the principal victim had left the perpetrator for another man, he killed both of them
and then himself. The article reports, “It appears (the woman and the suspect) have kids
in common and visitation was an issue between them.” In a follow-up story of the same
case the journalist wrote, “The suspect is Brownell’s former husband who may have been
angry because his former wife was denying him visitation with his two children”. They
added, “There were issues about the suspect not getting his parenting time.” In another
article where Levi Stuart Palmer Sr. killed his infant son and then himself, experts noted
that the prospect of losing access to the child might have triggered the horriﬁc act.
Witnesses claimed, “his life revolved around his child” and experts offered the following
insight: “They (the male perpetrators) are feeling trapped, trapped because ‘I can’t have

9”

my child, I can’t live without my child, so I have to take him with me. When reﬂecting
on the collateral homicide, Susan Gano-Phillips of UM-Flint said, “Sometimes a person
facing loss of custody of a child is unable to see options.”
It was a Domestic Dispute — Ambiguous Relationship Problems

Although many examples focused on a speciﬁc cause, some stories only

mentioned domestic dispute in the most general of terms. Some articles were so brief

that only an “ongoing dispute between the couple” was mentioned as a potential cause for

83

the homicide. In one situation where the collateral participant, Jason David Freeman,
intervened in a ﬁght between his mother and stepfather, Randy Scott Hall, the event was
referred to as a “domestic dispute” that “escalated to ﬁsticuffs between Freeman and
Hall.” In a case where an 18-year-old man killed two individuals and wounded his
girlfriend, “Investigators say the shootings have been ignited by a dispute between the
suspect and his girlfriend, Courtney Harris.” In another example, a family argument was
attributed to the shooting of a wife and slaying of her father. According to police,
“Sometime before 7:30 pm Sunday, Zhang and his wife got into an intense argument.
Police were not able to piece together Monday exactly what the dispute was about. After
the ﬁght, Zhang rushed upstairs and got a 9mm pistol.” In a further example of general
“domestic disputes”, police claimed that Luther Jenkins, a Detroit man who killed his ex-
girlﬁiend, her aunt, and her two children, murdered them because of an argument,
although Jenkins “hasn’t given police a reason for the domestic dispute.” When William
Watts Jr. killed a coworker because he suspected him of seeing his estranged wife, police
reported that “Watts and his wife were having marital difﬁculties and the shootings were
linked to a domestic dispute.” A fellow colleague of Watts, who sometimes ate lunch
with him, noted “he was suffering stress over marital difﬁculties.” Finally, when Terry
Lee Hitchcock killed his girlfriend’s 4-year-old daughter, Jami Renee White, relationship
troubles triggered a “moment of rage”. According to police, “A souring romance and
failing marriage may have triggered a moment of rage in which a Holland man killed his
girlfriend’s 4-year old daughter last week.” They go on to conclude, “Both were
estranged from their spouses, and police say Hitchcock may have taken the relationship

more seriously than she did.”

84

It was a Domestic Dispute — Relationship Termination

The principal IPV victim’s decision to separate from or divorce her partner was
also cited as a potential cause of collateral IPH. In the case of Gregory Scott Alexander,
he killed his wife’s new boyfriend, Kevin Robert Johnston, because of an impending
divorce. According to police, “We don’t have a history there. . .It appears she had just
recently ﬁled for divorce.” Next, in a highly publicized case in which Anthony Bailey
shot all four of his children, killing three and severely wounding the fourth, the proposed
cause was that he was severely depressed because his marriage was ending. “Bradley
(Anthony’s stepfather) said that Essie (his wife) wanted a divorce, which upset his
stepson.” Anthony’s stepfather went on to conclude, “My son had been depressed about
the marriage. He didn’t want to divorce.” As a ﬁnal example, divorce was the cause for
violence in a situation where a man killed another staying in his estranged wife’s home.
“The couple was involved in an ongoing divorce and had been separated for several
months, court records show.”
It was a Domestic Dispute — Inﬁdelity, New Relationshipsgand Rejection

Just as divorce and separation were offered as potential causes of collateral IPH, a
cheating spouse or their rejections were ﬁequently proposed as stimuli. When discussing
the homicide(s), descriptions of how the perpetrator’s partner or ex-partner was engaging
in sexual / romantic relations with another man were introduced. Interestingly enough,
the articles were not always clear regarding principal relationship status (i.e., if the couple
was still together or broken up). Although it was rare for a reporter or police ofﬁcer to
directly attribute inﬁdelity to a collateral’s death, information about their “indiscretions”

was often presented to provide context for the murder. For example, when Steven

85

Tierney shot Craig Fleck and Sally Paaj anen “in a rage over their apparent romantic
relationship”, the perpetrator claimed that he didn’t want to kill them, just an explanation
from Sally. “Tierney told the jury that he intended to confront Paajanen about her
relationship with Fleck, and then kill himself.” In a similar case where Matthew
Sparagowski killed his wife, Mary Elizabeth Sparagowski and Nader Abdel-Karim
Rammouni, the journalist reported, “Mary Sparagowski and Ramrnouni were involved in
an extra-marital affair.” In relating the events, police stated that “When he came home,
he apparently realized there was somebody else in the house and took the gun with him.”
In a further example, another man caught his wife involved in a lesbian affair, despite the
fact that she had previously promised to end it. “She said Deleon’s wife promised to end
it, but Deleon heard the two talking on the phone and snapped.” Lastly, a man killed his
estranged wife’s new boyfriend after he caught them having sex. “A Holland Township
man accused of killing his estranged wife’s boyfriend admitted shooting the man after
ﬁnding the man engaged in sexual intercourse, police testiﬁed Wednesday.” This
particular case is also tied to the themes of rejection and divorce, as the article noted,
“Phitsarnay Souvarma ﬁled for divorce ﬁom her husband in September, according to
court records. The couple had been married for six years and had two sons, ages 4 and
6.”

Additionally, rejection was offered as a motivator for the homicide of James
Watson. Jeffery MacIntosh killed his ex-girlfriend’s father because, according to
prosecutors, “he was bitter that Watson’s daughter, Michelle Fales, ended their romantic
relationship. He was also attracted to Watson’s wife, Annette, who rejected him too.”

According to a Flint News article, “They are feeling trapped. . .often the men are

86

distraught and can’t handle rejection.” The story concluded, “The cases show how
differently men and women cope with rejection. Women who kill their whole family
usually do so in despair, the outgrowth of mental illness, while men more often are
retaliating against women who left them.”
Out-QCControl Perpetrator

Another causal theme detected in these newspaper articles was that of the “out-of-
control” perpetrator. Because of the horriﬁc and frequently unexplainable nature of the
crimes, police and witnesses frequently attributed his actions to a temporary loss of
control. However, most incidents were still connected to ﬁ'ustration, conﬂict, or anger
within the principal relationship (52% of stories in which an explanation is offered).
Additionally, while being “out-of-control” was the rationale, rarely if ever were there
similar incidents reported when killing co-workers, ﬁiends, or others.

Steven Edward Tierney, whom “intended to take his own life, but instead shot
Craig Fleck and Sally Paajanen, in a rage over their apparent romantic relationship” said,
“I wasn’t myself. . .I didn’t know what I was doing.” He concluded, “I shot her. I can’t
remember how many times. . .I was in a rage. I can’t remember everything.” In another
case, the perpetrator crashed into his ex-girlﬁiend’s car and stabbed her and the presumed
new boyﬁiend to death. Loss of control was presented as the cause for the homicides.
According to police, “I asked him what had happened back there and he told me, ‘I lost
it’”. Likewise, when Shawn Harris shot his girlﬁiend, Tammy Dzurisin, and her friend,
Stacy Leclaire, he claimed he “lost it”. “He said it started out as an argument and it

escalated. Then he just lost it.”

87

If I can’t have you. no one else will

“If Steven Tierney couldn’t have Sally Paajanen in his mind, no one else could
either. The crime was clearly premeditated and deliberated.” The ﬁnal theme in the
causal diagnosis category highlights the intentional, or premeditated nature of the
collateral IPH. This theme was separated from the rest because it went beyond the notion
that the perpetrator was “out of control”, having a dispute with his partner, mentally ill, or
ﬁ'ustrated as a result of social factors. The primary indicators of this theme include
evidence of j ealousy or the explanation that he killed because he “loved her so much.”
As mentioned frequently throughout this section, codes and stories do overlap, but these
examples explicitly highlight the calculated nature of the homicide. Premeditation is
evident, as is the conscious decision to take another’s life. Only eleven articles (n=11)
clearly identiﬁed premeditation as cause.

While ongoing relationship problems were initially attributed to the homicides of
three individuals in a Battle Creek Enquirer story, as the article progressed, evidence of
intense jealousy emerged. “He loved Anna but almost too much. He would rather see
her dead than with someone else.” In another case in which too much love was the
proposed explanation, Oliver Webb IV shot Tommie L. Ford because he was love struck
by 18-year—old Emily Moore. “A love struck Saginaw teen-ager killed a romantic rival,
mutilated the body, then bragged about the slaying and displayed the corpse like a prize,
prosecutors claim.” In a brutal case in which Robert Edward Fletcher killed his two
stepchildren and their mother (his ex-partner), his deep love for her was introduced in
conjunction with his sadness over the failed marriage. Friends claimed, “Here is a guy

who had been alone a long time. . .and then he found the love of his life.”

88

Jealousy was also frequently attributed to the discovery that one’s spouse or
partner was cheating, and thus instigated collateral homicides. “Taylor was angry with
Fochtrnan because he believed she was seeing other men.” When Randy Prater killed his
ex-girlﬁiend and her male friend, police speculated that jealousy was his primary
motivation. “Police alleged in Baton County that Prater was jealous because Mills (his
ex-girlfriend) was dating Baker (the collateral victim)” They commented, “This was a
tragedy, and we felt that our evidence was overwhelming and that it was premeditated
and done during the commission of a serious felony. He is a dangerous, dangerous man.”

Some articles focused extensively on the premeditated nature of collateral IPH.
When Jason Scott Roof killed a two-year-old, “He was convicted of ﬁrst-degree murder
because it was clear it was intentional.” The article went on, “Roof had unsuccessfully
tried to smother the boy, then laid him on a beach towel and stabbed him in the heart.”
The calculated process that Terry Hitchcock engaged in when killing his former
girlfriend’s daughter (Jami) was introduced by the county medical examiner: “Medical
experts testiﬁed Jami likely struggled ﬁercely before passing out after two minutes and
the killer’s hand remained locked over the child’s airway for another minute to two
minutes.”

Perpetrator-Speciﬁc Problems

Although the majority of causal diagnoses emerged from some ongoing problem
in the principal relationship (i.e. divorce, separation, cheating, child custody issues),
many articles also attributed cause to a speciﬁc problem that the perpetrator was coping
with. These problems fell into two major categories: (1) individual factors and (2) social

factors. Figures 4.2 and 4.3 display the codes described in this section.

89

Figure 4.2 Individual Factors

  
   
 

 

    
  

Mental
Problems

   
      

Drugs &
Alcohol

' Childhood
Issues

   

' Individual
Factors

     
 

Figure 4.3 Social Factors

Social ' ~ Financial — .. Unemployment ’

Factors Problems

 

Individual Factors

According to quantitative results, if a causal attribution was provided, it was
primarily individual-oriented (63%). This is not to say that the perpetrator’s motives for
committing the homicide were coercive in nature; rather these attributions stemmed from
preexisting issues that were out of his control and contributed to subsequent family stress
and strife, a ﬁnding that falls in line with family violence theory. One particular
explanation offered for perpetrators’ violent behavior was a violent childhood or
upbringing. For example, Timothy Lee Rumsey “shot and killed a stepdaughter and two
of his children before killing himself. He also shot and critically injured a 2-year-old
daughter.” In discussing the possible reasons for the homicides, his mother said
“although he had a violent history since childhood, and was being treated in Texas with
Prozac for severe depression, Rees (his mother) said it was unfair for Joy Rumsey (the
perpetrator’s wife) to keep calling the police on him.” The story went into further detail

about Rumsey’s turbulent past: “The next thing I know, he was shipped to foster homes

90

because his father couldn’t handle him. . .that made his violent streak worse.” This
particular example highlights the comorbidity of mental problems and a maladaptive
childhood. In the case of Kenneth Sherry, a man who stabbed his wife and mother-in-law
to death, the use of cocaine was linked to the homicides. “He may have been under the
inﬂuence of illegal drugs at the time. In a follow-up article, “Police say Sherry told them
he was high on cocaine when he stabbed 63-year-old Judith L. Dies and 46-year-old
Vickie L. Sherry to death on July 15, 2004.” In yet another follow-up story that provided
more history on Sherry, an individual who knew him claimed, “The stabbings weren’t the
ﬁrst time that the highly addictive drug triggered Sherry’s dark side.” He said “Sherry
once shot a girl in the head while under the inﬂuence of cocaine.” Sherry “claimed he
saw demons. Well, his demons have a name — It’s cocaine.”

Thus far, two highly publicized cases have pointed to childhood, drug, and mental
health issues as potential triggers for collateral IPH. However, it was much more
common for articles to describe the mental “state” of perpetrators, as defense attorneys
frequently requested evaluations of their mental capacity (or incapacity) to avoid ﬁrst
degree murder charges. Particularly in stories that focused exclusively on judicial
proceedings, statements such as “a man who killed his son and wife is undergoing a
mental examination” were not uncommon. Additionally, some perpetrators attempted to
defend their behavior by arguing that mental deﬁciency or incapacitation caused the
homicides. When Jason Scott Roof killed the 2-year-old boy, he pleaded that he was
mentally ill at the time of the homicide. “I just stabbed him. I don’t know why. I just
did it.” Even during events preceding the killings, neighbors and friends depicted a

person who was not “normal.” In the Anthony Bailey case (he killed three of his four

91

children), neighbors and friends reported, “He just stood there like a statue, staring
straight ahead. . .I told him he needed to get help, and he said there was no way to help
him.” Another neighbor said, “I saw him outside and he said, ‘I’m about to go crazy.’ I
asked him if he was alright, and he said ‘I don’t know.’”

When Larry Mitchell killed his wife, Beverly B. Mitchell, and her sister, Barbara
Lownsberry, he was “constantly depressed and was taking several medications to help
deal with his problem.” Terry Lee Hall, “a gunman estranged from his wife” who “went
on a shooting rampage around dinnertime, killing his wife at an Allen deli, then gunning
down her father and stepmother at his wife’s home before ﬁnally shooting himself in the
head” also suffered from mental problems. According to court records, “He was in and
out of mental hospitals at the same time, and even attempted suicide at least once.”
Timothy Lee Rumsey, a man who killed his stepdaughter and two other children was also
being treated for depression.

Other perpetrators reportedly suffered ﬁ'om schizophrenic and related dissociative
disorders. Larry Widdell McNeil Stovall, who killed his ex-girlﬁiend’s roommate,
testiﬁed in court “he heard ‘voices’ telling him to kill Gadziemski (the roommate), whom
he thought was blocking his reconciliation with Hathaway (the ex-girlfriend).” In a
gruesome case where Arthur Hollingsworth Cayce abducted and raped his ex-girlfriend
and then killed her best ﬁiend, Melissa Boyair, setting her home on ﬁre, depression was
proposed as the culprit. According to the perpetrator’s mother, “Her son battled
depression for years, and had attempted suicide more than once. The breakup touched off

another bout of depression.”

92

Social Factors

Although individual explanations dominated newspaper articles (63 %), certain
social issues impacting the perpetrator and his family were offered as possible causes for
the collateral homicides. This particular category supports the family violence
perspective, as “typical” family issues triggered his violence. Major explanations
included ﬁnancial / economic strain and problems with unemployment. It is important to
note that although individual and social factors were separated for simplicity’s sake, a
handful of articles contained elements of both (1 .9%). For instance, a few stories
described how a “domestic dispute” or other relationship problems were magniﬁed by
social factors, such as unemployment or ﬁnancial strain. It appears as if these causal
diagnoses go hand-in-hand when explaining a perpetrator’s motives for committing
collateral IPH.

For example, when Ray, Virginia, and 6-year-old Joshua Dugan were found dead,
neighbors reported, “the Dugans’ were having ﬁnancial troubles and were divorcing (a
theme that overlaps with relationship termination)”. They went on to say, “Ray Dugan
had recently lost his computer company and had been forced to sell his car.” In an article
on the highly publicized Anthony Bailey case, the perpetrator’s unemployment status was
a central topic throughout the story. “Police say Anthony Bailey, broke and unemployed,
' told them he could no longer care for his children.”

In another example in which two causal themes overlap, a “domestic dispute” that
resulted in the deaths of four people was attributed to ﬁnancial strain. “The suspect has
been questioned at length by homicide detectives and has made incriminating statements,

police said, and said the domestic dispute was ﬁnancial in nature.” This particular

93

example came from the Luther Jenkins case, in which he killed his girlﬁiend, her aunt,
and slashed the throats of her two children. The reporter claimed, “A (ﬁnancial)
domestic dispute led to the carnage.”

In describing the possible reasons that led Jimmie Reed Jr. to kill his girlﬁiend
and two-month-old daughter, the casual diagnosis focused on his desire to avoid paying
child support. “A man accused of shooting his girlﬁiend, then setting her and their 2-
month-old daughter on ﬁre with the hope of avoiding paying child support has been
charged with murdering them.” Additionally, when Edward Evans killed his wife,
Rebecca Evans and her mother, Josephine Jones, economic stress was a focal point of the
story. “Edward Evans had been unemployed since his lucrative party store business
failed several years ago.” Witnesses added, “He’d think they were against him.” At the
time of the homicides Rebecca “was carrying all the weight of the bills, and he wanted to
be the man of the house.”

Generic Frames #3 & #4 — Moral Im lications & Remedies “What is this World

Coming to? ”l

The ﬁnal two categories included in Entman’s (1993) generic framing scheme, '

   

moral implications and potential remedies, were infrequently found in this sample of
articles. Due to the substantial amounts of Associated Press (AP) articles and news
briefs, minimal space was allotted for moral reﬂections or how one might alleviate crimes
of this nature. For instance, if a remedy was ever offered, it was only in cases that
explicitly discussed DV or IPV, with shelter and social service resources provided.
Although such stories were a rarity in this sample, their presence illuminated the coercive
nature of collateral IPH by condemning the behavior (instead of exhibiting shock) and

providing helpful information for victims of abuse.

94

Moral judgments generally revolved around the cruel and heartless nature of the
homicides. One theme that emerged quite early in analysis was the shock at how
someone could kill children (31.2% of articles explicitly mentioned shock). For example,
throughout the Bailey case, witnesses exhibited shock and disbelief that he could do such
a thing to his own kids. They reported that Bailey was a “caring father who was
protective of his children.” J immie Reed Jr., who killed his daughter and girlfriend,
received this moral condemnation ﬂom the principal victim’s brother: “Why would he do
that to my sister, but more importantly, why would he do that to the child he just had?”
In a similar vein, family members of the late Melissa Friar and her daughter, Alana,
commented, “I just don’t understand. She was only 8-years-old.”

Another theme that emerged was the cruel, heartless, and “evil” nature of the
collateral homicides. Reverend James Flakes referred to the slaughtering of a woman and
her four children as “the devil’s work.” When Roger Thompson killed four children and
their mother, witnesses claimed that it was “one of the most gruesome and disturbing
crimes in Detroit history” and “This was so evil and coldhearted. . .I can’t explain this. I
just can’t ﬁnd the words.” Clearly, a theme of tragedy was interwoven throughout each
moral judgment.

Incident and Relationsh ' eci tc Themes

In conjunction to generic ﬂames, incident and / or relationship-speciﬁc themes
also emerged throughout data analysis. These particular characteristics set the stories
apart ﬂom other homicide reporting, as they were unique to collateral IPH. Five major
themes were prominent: (l) a focus on the principals, (2) collateral background and

involvement; (3) perpetrator love of family; (4) the preponderance of shock; and (5)

95

implications for child witnesses and survivors. While categories (1) and (2) were more
descriptive in nature, (3) and (4) lend support the family violence perspective because
journalists introduced information about how much the perpetrator loved his children
(despite the killings) and how witnesses could not believe he was capable of such an act
(he must have been out-of-control with anger). The ﬁnal category, like (1) and (2)
represents a unique reporting pattern associated with collateral IPH.

Focus on the Principals

Perpetrator — It was evident early in analysis that articles primarily focused on
the perpetrator and principal IPV victim, as she was the most ﬂequently reported
casualty. Thus, numerous articles tracked the path of the principal relationship across
time. Two types of information were apparent: (1) background on each individual or the
couple as a unit and (2) preexisting relationship troubles. Only (1) will be discussed here
since the ﬁnal section introduces quotes pertaining to principal relationship problems and
their theoretical alignment with either family violence or coercive control theory. It is
important to note that some of the information or context included in articles was
extraneous but incorporated nonetheless. This may have been done to ﬁll space or to
better acquaint the public with the perpetrator and victims.

The type of information provided about perpetrators was diverse, and at times,
unusual. For example, in one article, William Detzler’s video game habit was
introduced: “The suspect didn’t hold a steady job and spent an inordinate amount of time
playing electronic videogarnes at home.” It went on, “Games with names like Resident
Evil and Doom. He’d play those games all day long when he wasn’t working and we’re

talking about a 35-year—old man.” In a follow-up story on the Ray Dugan homicides his

96

political beliefs and frustrations were introduced. “Those who knew Dugan said he was
enraged by the federal government’s ﬁery raid on a complex in Waco, Texas, in 1993 and
blamed the Internal Revenue Service for his failed Troy business, Time Communications,
Inc.” Friend and neighbor reﬂections of the perpetrator were also included. “Neighbors
and other acquaintances recalled his tirades about the government, letters to newspapers
slamming the Troy school board and his antitax views, but ﬁiends and associates also
think he held the potential to improve government.” Levi Palmer, who killed his son and
himself, had worked a part-time job through a temp agency. Said his former boss, “We
were getting ready to bring him on full time. . .he was a good employee, pretty much kept
to himself and never talked about problems.” In separate example, the perpetrator was
depicted as a good citizen and religious individual. The article said, “But relatives of
Terry Lee Hitchcock say he’s a peaceful, churchgoing father of two who’s never been in
trouble.” Terry Hall, who killed his wife and her father and stepfather, was described as
“a kind of quiet-to-himself style person”.

Principal IPV Victim - When describing the principal IPV victim, reporters and
informants ﬂequently focused on their employment status and familial roles (i.e., what
kind of mother they were). For instance, when discussing Reco’s murdered girlﬂiend:
“Bellamy’s focus was her children. She was constantly moving between jobs, at Kmart,
Bob Evans, McDonalds, Toys ‘R’ Us. She had just gotten a position as a telemarketer for
Utilities Analysis, Inc. of Redford Township, where she was well regarded.” When her
ex-husband killed Machekia Franklin and her two daughters, Rockell Johnson and Teria
Johnson, family members had this to say about her: “Family members said Franklin was

a quiet Pontiac native who had one major fault: she always picked the wrong men.”

97

Additionally, “Robinson (she changed her name after the divorce) was a front desk
receptionist and on track for a promotion at the Wingate Inn in Auburn Hills.” Another
principal victim was described as “just one of those people who would do anything for
anyone.” J immie Reed’s principal victim, Markeda Byas, “was a fun-loving and spiritual
woman, who loved to cook for her family and ballroom dance.” Melissa Friar, murdered
by her husband, was “a wonderful morn. . .she was going to school because she wanted to
help people. I think because she had asthma and knew what it was like to be sick.” As a
ﬁnal example, in an article entirely devoted to the victims of Terry Hall, Diane Hall
received this homage: “She will be remembered as an unselﬁsh person, liked by all who
knew her, and a wonderful mother who sometimes worked as many as three jobs so her
ﬁve children would never go without necessities.”

The Principal Couple — “Margaret Wasuikanis, who lived next door to the Bailey
family, said the couple seemed very much in love when they moved in two years ago.”
In order to provide background on the principal couple, it was not unusual to encounter
outside opinions on the principal IPV victim and perpetrator. As such, common quotes
included: “The couple was private and minded their own business”, “She said she never
witnessed any problems between the husband and the wife”, and “She went her way and
he went his.” Often, the employment status of both parties was presented: “DeLeon
worked for General Motors’ Grand Blanc Metal Center. His wife and Jetawyn Lee (the
collateral victim) worked for the Flint Truck Plant”. At other times, a brief family
snapshot was provided: “The Watts’ have preschool age twin girls and a 22-month old
boy, coworkers said.” Additionally, some articles introduced the principal relationship,

but also incorporated a troubled past: “Friends and family say Thursday’s murders were a

98

tragic, blood soaked ending to a relationship that began more than ten years ago,
produced four well-liked children and ultimately ended in Terry Hall making good on
previous threats to his wife.”

Collateral Background & Involvement

Despite the focus being primarily on the perpetrator and principal victim,
collateral background information and their involvement in the primary relationship were
included in many articles. The type of information or description varied greatly
depending on collateral victim, as the article ﬂames were different for children as
opposed to adult men.

Adult Collateral Victims ’ Involvement — If the homicide included an adult
collateral victim, their deaths were ﬂequently attributed to intervening in a domestic
dispute or their attempt to protect the principal IPV victim ﬂom harm. In a story titled,
Man Dead After Family Fight Takes a Violent T urn, William Detzler was killed in an
apparent “family ﬁght” that went too far. “The crime occurred after William Detzler
conﬂonted the suspect, who is married to Detzler’s sister.” Witnesses claimed, “From
what I heard she said (the suspect) has been shoving her around all day, so she called her
brother, Bill (Detzler) to talk man-to-man to (the suspect)”. A related article said, “F 0y
and his wife, who are going through a separation, were having an argument in their
mobile home on the west side of Beaverton on Wednesday night. . .Detzler thought he
could straighten it out.”

When Randy Sanchez murdered Peter Wilkinson, who was reportedly ﬁiends
with his girlﬂiend, the collateral was presumably caught in the crossﬁre of an ongoing

conﬂict. “Wilkinson came to her aid again on Dec. 5, when Coon’s boyfriend, Sanchez,

99

41, came at her with a knife after she’d asked Sanchez to leave, Coon testiﬁed on
Tuesday.” In a separate case, Roger Sanford, a collateral victim who was killed because
he was with the suspect’s ex-girlfriend, Robin Howard, tried to protect her ﬂom harm:
“When the suspect attacked Howard, Sanford tried to help her, police said. He was
stabbed several times as well.” Michael Brattin killed his estranged wife’s new partner in
their workplace parking lot. “Witnesses told police that Michael Brattin shot his wife
ﬁrst, then ﬁred at O’Brien, who apparently tried to intercede.” Even those not
romantically involved with the principal victim, such as parents, were not safe from this
outcome: Haoran Zhang killed his father-in-law when he tried to protect his daughter.
“The wife’s father, who lived at the home with his wife and cared for the two young
children, became part of the argument as well, police said.”

In addition to describing adult collateral victims’ intervention in crime
commission, their “romantic” involvement with the principal IPV victim was also
highlighted. Whether they were friends with or dating her, the article often told a story of
a tragic “love triangle” that resulted in murder. For instance, Michael O’Brien “knew he
was courting big trouble” when he decided to date his coworker, Sandra Brattin. “He’d
been seeing his co-worker, Sandra Brattin, and knew that it angered her estranged
husband, Michael.” One man was killed because he was caught having sex with a
perpetrator’s estranged wife: “A Holland Township man (Souvanna) accused of killing
his estranged wife’s boyfriend admitted shooting the man after ﬁnding the couple
engaged in sexual intercourse, police testiﬁed Wednesday.” Despite the couple’s
impending divorce, he had reportedly threatened the collateral victim in the past. Said a

family member, “He wanted to kill my brother. . .He said he would not let my brother stay

100

alive to the end of the year.” Although these pieces of information did not overtly blame
the collateral victim for their death, they did insinuate a known risk that could have been
avoided.

Child Collateral Victim Description - Interestingly enough, adult collateral
victims were rarely described in detail, but children received signiﬁcant attention and
description. Their hobbies, personalities, or school interests were ﬂequently introduced.
“Several teachers and principal Carol Garland called the girls model students who were
involved in student council, earned high grades, and possessed exemplary reading and
writing skills.” Likewise, reporters provided a signiﬁcant amount of information on
Luther Jenkin’s four collateral victims. “Alicia Jackson was a good little girl, a great
teenager, and a beautiful adult. She made you happy because she was just a happy spirit.
She was a decent, respectful, good girl.” The other collateral victims, ages four and
seven, received attention as well. “J amon, known as Mony, attended Hope of Detroit
Academy in Detroit. A’Janneya, known as Nay Nay, was in preschool. . .Their
grandmother, Rosetta Hamilton, 49, called them ‘beautiful children.”’

The room of Machekia Robinson’s two slain daughters was vividly described in
one story: “Brightly colored cherries adorned the sheets in the little girls’ bunk beds. In
the corner of their room was a small, pink, heart-shaped dressing table. A school poster

9”

on the white wall encouraged: ‘Reading Power. The hobbies and career aspirations of
Joshua and Naomi Ross were also introduced. “They remembered the Ross children as
bright, responsible kids; one was headed to Michigan State to become a police ofﬁcer, the

other was busy with track, basketball and volleyball.” Lastly, Arctavia, the infant victim

101

of Jimmie Reed Jr., “was a contented baby and a ‘sweet sleeper’ who never wanted to be
put down, family said.”

General Collateral IPH Victim Information — Similar to the information reported
about principal victims and perpetrators, the types of descriptions provided on collateral
victims varied greatly. In one article, a collateral victim’s employment, spirituality and
role as father were explored. “Ketchens, 40, was considered a devoted father and
husband, a gregarious salesman at Extreme Dodge and a faithful Christian. He was the
life of Trinity United Methodist Church’s Christmas tree lot in recent weeks.” Roger
Gibson, father of a principal victim, “was a really nice man — he was always puttering
around the garage.” Lastly, Robert Gibson, shot and killed byhis daughter’s ex-
boyfriend, Chadwick Wiersma, “was not a conﬂontational person in any way. He was a
very gentle person. I don’t think I ever heard him raise his voice.” The story concluded:
“They wanted to tell the story of Robert Gibson, a devoted father and husband whom his
wife remembered as a ‘renaissance man’, who was comfortable among books or
watching the Discovery Channel as he was in his workshop.”

Perpetrator Love of F amilv

A third theme that emerged throughout analysis, particularly in those stories
where children were killed, was a discussion of the perpetrator’s love of his family and
children. Once again, the Bailey homicides provide an excellent example. Although he
killed three of his four children, “he was so helpful, so mindful of [them].” One article in
the Detroit Free Press even reported, “He loved his kids” despite the brutality of the

murders. One of the Baileys’ neighbors stated to the Detroit News, “When I heard about

102

it, I thought, no, no, no. Not Tony. He loved those kids so much. If anything, he was
overprotective of them.”

Levi Stuart Palmer, the man who killed his 17-month-old son, was described as an
attentive father, whose “life revolved around his child.” Gary Hicks, who “took two 5-
gallon fuel containers and doused his house with gasoline while his three children slept”,
was described by neighbors “as a private yet friendly man, who loved working on his
house and jogging, but most off spending time with his three daughters, Vanessa, 12,
Ginger, 7, and Erika, 3.” A court afﬁdavit introduced in the story “describes a man
somewhat obsessed with keeping his marriage together and keeping his children in his
life.” To sum up, “For Garry, everything was for the kids. . .everything was. He was a
family man.” Next, Leo Cannona, who killed his girlﬁiend’s 2-year-old son, loved him
like he was his own. “Carmona claims he didn’t kill the blonde toddler he says loved
watching the movie Stuart Little and could already count to 10.” He said in court, “I
couldn’t have loved him more. He was perfect.” In a ﬁnal example, Timothy Lee
Rumsey, convicted of killing a stepdaughter, two of his other children, and himself
“loved his children so much.” “He just loved them so much that he didn’t trust her with
them alone.”

The Preponderance of Shock — “These Incidents are Hard to Explain”

“I just can’t believe it. . .he was always listening to bible tapes and he was rational
when I talked to them yesterday.” Perhaps one of the most apparent themes in this
sample was that of overwhelming shock. Regardless of collateral victim (adult or child)
or type of witness (police, neighbors, family), quotes highlighted extreme shock. Even in

those articles where a history of violence was introduced, including past abuse against the

103

principal IPV victim, witnesses and / or informants just couldn’t believe it. For instance,
in an article on the Terry Hall case, his past abuse was ﬁrst articulated: “In less than one
month in 1990, Diane Hall made three phone calls and four visits to a domestic assault
shelter — deathly aﬂaid her husband would follow through with his oft-repeated threat to
burn her alive.” The story then goes on to say, “But no one expected a hail of bullets”
and “Everyone’s in shock. . .there was no indication that this would happen.”

However, most examples of shock occurred when no family or domestic violence
had previously been identiﬁed (i.e., no context). “No prior calls to the home for domestic
violence by the sheriff’s department were discovered, Richardson said. A check of court
records showed there had been no personal protection orders or divorce papers. ﬁled.” In
the Bailey case, not only were neighbors and friends shocked that Anthony could do such
a thing (“Why would a man who loves his children and was so interested in them do
something like this?”), but the lack of family problems was introduced as well. “If there
were warning signs of problems at home, Bryant said, she didn’t see them.” When
William Watts Jr. killed a coworker because he believed he was seeing his estranged
wife, co-workers exhibited extreme shock: “Co-workers, who huddled across the street
from the factory, said they were aware of William Watts’ marital problems but never
knew him to be violent. He had no known felony convictions.” One coworker even
remarked, “Watts owned a lot of guns and was a deer hunter, but he never talked of
violence.” In a similar vein, Roger Thompson “ had no criminal record, and Wayne
County Prosecutor Kym Worthy said no domestic violence complaints ever had been

ﬁled against him.” Lastly, a coworker of Garry Hicks said, “We are all totally surprised.

104

I’ve known Garry for years and I never would have expected something like this. I don’t
think anybody did.”
Other examples of shock include:

0 “How could this man hurt two innocent children?”

0 “It’s so close to home. How could something like this happen so close to home?”

0 “The brutal deaths of two women and two small children shocked a city torn by

bloodshed of more than 370 murders and 1,300 nonfatal shootings in 2004.”

0 “I don’t understand how anyone could do this.”

Implications for Child Witnesses & Survivors — “My Momma ’s Boyfriend Shot My Dad”
The ﬁnal incident-speciﬁc theme detected in the present analysis was that of

implications for child witnesses and survivors of the collateral IPH. Many homicides
discussed thus far involved children (47.3% of victims mentioned), and in some cases,
not all were killed. As such, a focus on the killings’ impact on survivors and witnesses
(e. g. those that were not a target of attack, but saw the homicides occur) was prevalent.
In one example, child witnesses had to hide with their mother in a bathroom. “The
Alexanders’ two sons, ages 6 and 3, were inside the home when the shooting occurred.
Tammara Alexander locked herself and her sons inside a bathroom fearing that her
husband had entered the house.” When David Gibbard shot his ex-wife’s husband, her
son was also present. It was reported, “He looked like he was very traumatized.” Even a
two-year-old child witnessed the Kenneth Powell homicides. “The 2-year-old told
authorities that her father pulled the trigger. Her exact words were, ‘Daddy shot twice’,
said Sherriff Jeffery V. Goyt.” Lastly, when Diane Hall, along with her father and

stepmother were killed, she left four children behind. “It’s for all those poor kids left

105

behind. They’ll need all the help they can get.” Diane’s nephew, David, is “taking it real
hard because he was really close to his grandpa, and he saw his grandfather get shot and
saw his grandma go down, and when he went to run for help, he saw his Aunt Diane
lying in ﬂont of the store.”

Two particular child survivors received extensive media coverage: Antonia Bailey
and D’anaj eh Robinson. D’anajeh lost her mother and two sisters, and reportedly told
investigators that “Daddy got a knife” and “described how one sister, before she died,
begged to be allowed to say goodbye to her mother.” One article noted, “D’anaj eh
Robinson, now 4, spent the night alone with the bodies of her mother and half-sister,
according to prosecutors.” One article focused on how she was coping in the care of her
maternal grandmother. “Her life today is all about learning to read, riding her two-wheel
bike with the training wheels, playing with her Barbies and visiting the J eepers! Game
center with her aunties and cousins.” Gwendolyn Robinson, her grandmother, said “I call
her my miracle child. . .my daughter is gone, my other grandchildren are gone, but God
has given me this little gir .”

Antonia Bailey claimed, “My daddy killed me. My daddy killed my whole
family” (this quote was repeated ﬂequently across newspapers). Although her mother
was not involved in the attack, she lost all of her siblings in this particular collateral
homicide case. “As the emergency medical technician administered ﬁrst aid to Antonia
Bailey, she mustered the strength to tell him who had shot her and her siblings.” During
her father’s trial she was still in the hospital, so she sent a letter to her deceased siblings
saying, “I love you all, and I miss you very much. . .I know you are all watching down

over me. That’s why I’m still here. Don’t be mad at me.”

106

M11 Violence vs. Coercive Control

A primary goal of the qualitative portion of this study was to gather evidence that
aligned with either family violence or coercive control theory. If reader refer to
Appendix D, they will see that an entire section of the coding protocol was devoted to
identifying examples that support the family violence framework, coercive control, or
both. Each article underwent rigorous, iterative analysis so that themes were accurately
captured and assigned. Please refer back to Table 3.3 for criteria used to determine
theoretical alignment.

For the sake of space, results are presented in table format (see Table 4.11). A
total of ninety-seven (n=97 or 45.1%) articles included evidence of family violence
theory, attributing the collateral IPH to mutual family ﬁghts, “domestic disputes”
between the perpetrator and principal intimate partner, the termination of an intimate
relationship, or unexpected perpetrator anger / rage (which is linked to numerous causes,
such as a cheating girlfriend, jealousy, argument, etc). As such, the key criterion of a
family violence sentence is evidence of a symmetrical cause or contribution, while
coercive control sentences introduced a pattern of abusive and coercive behavior by men.
The “conﬂict” label is similar to that found in quantitative analysis, with approximately
52% of killings identiﬁed as conﬂict-driven. A total of twenty-six (n=26 or 12.1%)
articles included evidence of coercive control theory, indicating prior perpetrator violence
or IPV, coercion, and controlling behaviors. Articles were included in this category if the
perpetrator killed himself in ﬂont of children or the principal victim, or if he killed his
partner’s children but spared his own. Like quantitative results indicating the occurrence

of intentional homicide (5.3 %), this was a small percentage of articles. Next, a total of

107

eight (n=8 or 3.7%) articles contained elements of both family violence and coercive

control, while eighty-three (n=83 or 38.6%) were marked “unclear”. An article was

deemed unclear if it was a news brief or AP article that lacked the background

information necessary to determine theoretical alignment. Please refer to Appendix E to

review the qualitative analysis matrices used in this portion of the study.

Table 4.11 Illustrative Family Violence & Coercive Control Sentences

 

 

 

 

 

 

 

 

 

“Taylor drove most of them
away (her ﬂiends). It was like
he wanted me to himself,
where I didn’t have any
interaction with anybody else.”

 

Source Date Coercive Control Family Violence

Ann Arbor 4/28/98 “Crawford said her boyfriend, who

News lived in the small bungalow with her

Ann Arbor, MI parents and brother, had been
angered because she asked him to
move home with his mother.”

Ann Arbor 8/21/03 “Bailey called his wife about

News 7:30pm Tuesday and told her

Ann Arbor, M1 to watch the news that night.

Investigators believe the
children were shot about an
hour later.”

Battle Creek 10/23/92 “Blood covered the kitchen ﬂoor of a

Enquirer small apartment where an ongoing

Battle Creek, dispute between former lovers ended

MI in a rampage late Thursday.”

“But police records show the
relationship turned violent.”
“They had a long-term, sick
relationship and (Thursday) he just
lost it.”

Battle Creek 10/24/92 “Friends and family said Taylor and

Enquirer Fochtrnan had an ongoing dispute.”

Battle Creek,

MI “They were trying to work things out
and get together, but something
snapped.”

Battle Creek 11/23/92 “She claims he listened

Enquirer through the walls and spied on

Battle Creek, her and her ﬂiends.”

MI

 

108

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence
Battle Creek 12/30/06 “Detective Sgt. Todd Madsen of the
Enquirer Battle Creek Police Department said
Battle Creek, the killings appear to be part of an
MI ongoing dispute between the couple,
although he declined to give further
theories.”
Bay City Times 5/27/99 “A family ﬁght in Beaverton ended
Bay City, MI in the stabbing death of a 36-year-
old Beaverton man on Wednescgy.”
Commercial 2/18/98 “A man suspected of killing
News his three daughters in a
Three Rivers, murder-suicide had previously
MI threatened to steal his children
if his wife left him, court
records show.”
Commercial 4/28/98 “He said that if he couldn’t
News have me and the baby, he
Three Rivers, would kill himself.”
MI
The Daily News 9/21/06 “We don’t have a history there. . . it
Greenville, MI appears she had just recently ﬁled for
divorce.”
The Daily News 10/21/03 “If Steven Tierney couldn’t
Iron Mountain, have Sally Paajanen, in his
MI mind, no one else could either.
The crime was clearly
premeditated and deliberated.”
Daily Tribune 12/19/96 “Elkins said a family member told
Royal Oak, MI her ‘problems’ had recently escalated
between Maria Joseph and her ex-
husband, but could not specify. Nor
did she know how long the two had
been divorced.”
Detroit Free 2/17/94 “Terry was always threatening
Press to blow people up. When she
Detroit, MI was pregnant, he held a gun to
her belly.”

 

 

“Hall told of one incident in
which her husband shot their
dog, twisted the heads off their
goldﬁsh, and put an ax through
their kitten’s head. Another
time, he chased Hall and her
oldest son down the street with
a gas can and a cigarette
lighter.”

 

 

109

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence
Detroit Free 1/4/96 “He said it started out as an
Press argument and it escalated. Then he
Detroit, MI just lost it.”
Detroit Free 6/6/96 “She described the Dugans as
Press ‘private people’, but said the
Detroit, MI neighbors were aware of money and
marriage troubles.”
Detroit Free 6/7/96 “The months leading up to the
Press shootings were ﬁlled with tension for
Detroit, MI Dugan, whose personal and
professional life were crumbling.”
Detroit Free 9/24/97 “In jail, he began calling
Press Bellamy collect. At ﬁrst the
Detroit, MI calls were wistful. Then they
took on a more demanding
tone. ‘Who are you seeing?
Are you cheating on me?”’
Detroit Free 6/23/03 “She hated those letters. In
Press them, her ex-husband would
Detroit, MI threaten to kill her and said it
would be ‘all over’ as soon as
he was released ﬂom prison.”
“He was going to get
her. . .when he got out of
jail. ..he said it would all be
over when he got out, and it’s
all over now.”
Detroit Free 7/25/03 “Investigators say the shootings have
Press been ignited by a dispute between
Detroit, N1] the suspect and his girlﬂiend,
Courtney Harris.”
Detroit Free 8/21/03 Right before Anthony Bailey “There has been some indication that
Press killed his children, he called there has been some abuse in the
Detroit, MI Essie Bailey. “Anthony Bailey family between the husband and
called his estranged wife at wife.”
work Tuesday night with an
order: ‘watch the news
tonight.”’
Detroit Free 12/9/04 “Thompson told police the massacre
Press began with an argument with is on-
Detroit, MI again oﬂ-again girlﬂiend, Shelton,
about a coat he stuffed under his
pillow before going to 516%”
Detroit Free 12/21/04 “Detroit police 2n Deputy Chief
Press James Tate said Pitts and Jenkins
Detroit, MI argued before the killings but no

 

 

 

clear motive has been established.”

“Family members said Pitts had
recently broken up with Jenkins.”

 

110

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence
Detroit Free 4/25/06 “The couple, who had been married
Press about a decade, had been having
Detroit, MI arguments for weeks, Scherlinck
said.”
“There was no sign that alcohol or
drugs were involved or that the
family had ﬁnancial problems.”
Detroit Free 7/27/06 “Police said Reed did not want to
Press pay child support to Byas and he
Detroit, MI wanted to keep the relationship a
secret ﬂom other girlﬁiends.”
Detroit Free 5/23/05 “Rivalry over a woman likely
Press compelled a 19-year-old man to
Detroit, MI ﬁrebomb a Detroit home last week,
killing two children but not the
intended victim, according to court
documents.”
Detroit Free No Date “With just a week or so to go
Press before Daniel Franklin would
Detroit, MI walk out of prison, his ex-wife
and her boss asked corrections
ofﬁcials how she could protect
herself and her children.”
Detroit News & 11/26/92 “A religious man who ﬂequently
Free Press read the bible and listened to gospel
Detroit, MI tapes apparently let his inner demons
consume him Wednesday.”
Detroit News & 7/31/93 “In less than one month in “People who knew the Halls said
Free Press 1990, Diane Hall made three they were aware of their domestic
Detroit, MI phone calls and four visits to a problems. They had reconciled
domestic assault shelter — without marrying after the divorce,
deathly aﬂaid her husband and had several stormy breakups
would follow through with his since then.”
oft-repeated threat to burn her
alive.”
Detroit News & 1/8/95 “There’s a triangle going there. . .he
Free Press warned the other guy. He told him,
Detroit, MI ‘if you wanna see my wife, don’t do
it in front of me.’”
Detroit News 2/10/97 “He told investigators that he and his
Detroit, MI ' wife of 14 years had been having
marital difﬁculties and that he shot
her because he was aﬂaid she was
gpi_ng to leave him.”
Detroit News 3/11/97 “Campbell told police that he and his

Detroit, MI

 

 

 

wife had gotten into an argument
over the stopped-up sink in the
kitchen.”

 

111

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence
Detroit News 12/20/04 “Police went to the house
Detroit, MI responding to domestic
violence calls twice in 2004,
but the suspect wasn’t arrested
on either occasion. In one
instance, he ﬂed before police
arrived; in the other, no one
was injured so police didn’t
make any arrests.”
Detroit News 4/25/06 “It escalated into a heated argument
Detroit, MI and ended after gunshots rang out
down a stairwell, leaving a mother
critically wounded and a grandfather
and father dead.”
Detroit News 7/27/06 “Investigators believe there was
Detroit, MI some kind of argument earlier in the
day, and he shot her and then went to
work.”
Flint Journal 1/96 “Palmer called his wife and
Flint, MI told her he was going to kill
himself. He offered to let her
listen while he pulled the
trigger on a deer riﬂe.”
“It’s the ultimate act of
aggression. ‘If 1 can’t have
you, then you’re not going to
have the thing that is most
important to you.”’
“If you can’t hurt her and you
kill her child, how much more
can you hurt her?”
Grand Rapids 11/15/01 “Police said Watts and his wife were
Press having marital difﬁculties and the
Grand Rapids, shootings were linked to a domestic
MI dispute.”
Grand Rapids 4/25/05 “I am afraid for my children
Press and myself. . .he has pushed
Grand Rapids, me, grabbed me and hit me
MI and threatened me. He has
broken a window in my living
room and has cut most of the
screens in my house to get
inside.”
Grand Rapids 7/21/06 “I pray for the family of Mr.
Press Thompson and my family. In a
Grand Rapids, world of turmoil, this is what
MI happens with bad judgment and

unhealthy relationships.”

 

112

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence

Grand Rapids 9/21/06 “A man shot and killed his

Press estranged wife’s boyfriend

Grand Rapids, early today then killed himself

MI in a ﬁeld nearby as he talked to

his wife on the phone.”

Grand Rapids No Date “A souring romance and a failing

Press marriage may have triggered a

Grand Rapids, moment of rage in which a Holland

MI man killed his girlfriend’s 4-year-old
daughter last week, police say.”
“People handle stress differently. It
may have been a situation of stress
he could not handle.”

The Herald— 7/17/01 “Court documents show a

Palladium relationship often plagued by

St. Joseph / violence and abuse.”

Benton Harbor,

MI “Even before the couple wed in July
1999, the relationship seemed
determined for destruction.”

Hillsdale Daily 7/30/93 “He said investigators believe

News problems in the Halls’ marriage led

Hillsdale, M1 to the deadly rampage. Neighbors
said Diane Hall kicked her husband
out of their home just two weeks
ago.”

“Every time we have a domestic
problem, we have the potential for it
to end up this way.”

Hillsdale Daily 7/31/93 “Many, if not most, battered

News women hear over and over, ‘if

Hillsdale, MI you ever leave me, I’ll kill

you.’ The fear of exactly this
happening keeps women
trapped in abusive
relationships.”

Hillsdale Daily 8/7/93 “That pattern — a pattern

News designed to control and

Hillsdale, MI dominate a woman is prevalent

 

 

in most domestic assault
cases.”

“Physical violence is just one
behavior, but verbal abuse,
isolation, and economic
control (are) other controlling
factors.”

 

 

113

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence
Holland 12/20/97 “Investigators were uncertain if the
Sentinel three were involved in a love triangle
Holland, MI and described the victim as only an

‘acquaintance’ of the wife’s.”
Jackson Citizen 10/2/99 “Clare County court
Patriot documents show Powell had
Jackson, MI twice threatened to kill his
wife if she left him — once in
February 1996 and again in
July.”
“Court records show that
Rebecca Powell alleged that
her husband broke her
collarbone in May 1995. She
also alleged that Powell hit and
slapped her several times over
the past few years.”
Kalamazoo 4/4/96 “Court records show Wiersma
Gazette had a history of violence
Kalamazoo, MI against his ex-girlfriend.”
“In February 1994, he was
convicted of attempted
kidnapping after he broke into
her car, forced her into his car
and drove for an hour and a
half, refusing to let her out and
threatenipg her with a knife.”
Kalamazoo 2/17/98 “Hicks, 41, threatened to steal
Gazette his child if his wife, Theresa,
Kalamazoo, MI left him, blackened her eye on
one occasion and subjected her
to a ‘history of physical abuse,
mental abuse, and violent
behavior’, according to a
personal protection afﬁdavit
ﬁled Nov. 4.”
Lansing State 4/3/04 “These ﬁve horrible deaths
Journal illustrated in the strongest possible
Lansing, MI terms the problem we have with
family violence in this country.”
Lansing State 1/11/93 “Jon Michael Escareno, 18, and
Journal Brenda Martinez, 17, had argued
Lansing, Nﬂ since ending their
relationship. . .McSwain didn’t know
how long ago the coupld had
separated.”

 

114

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence
Livingston Co. 11/28/04 “Word of the slayings shocked the
Press neighborhood though some said they
Brighton, MI had called police on previous

occasions to complain about loud
arguments ﬂom the woman’s
apartment.”
The Macomb 11/18/92 “He was apparently distraught over
Daily his impending divorce.”
Mt. Clemens,
MI
The Macomb 5/5/96 “A 55-year-old Roseville man who
Daily had a stormy relationship with is
Mt. Clemens, live-in girlfriend and her l6-year—old
MI daughter has been charged with their
murders.”
“Warner said Schomaker did not
work, which fueled his rocky
relationship with Lord.”
Muskegon 12/11/03 “There were lots of
Chronicle accusations over time. There
Muskegon, MI were physical conﬂontations,
threats made to me...then he
would apologize and then tell
me he loved me, and it would
start over.”
Muskegon 11/28/04 “The relationship was rife with
Chronicle conflict and abuse.”
Muskegon, MI
“They’d have misunderstandings,
things most people could talk
through, but they’d argue.”
Muskegon 5/20/06 “Stovall murdered his ex-
Chronicle girlﬂiend’s roommate because
Muskegon, MI he believed she stood in the
way of his relationship with
the pregnant woman he had
long beaten, prostituted and
terrorized with threats to harm

 

 

her 3-year-old daughter.”

“It ﬁt with Stovall’s pattern of
cruelty toward his former
girlﬂiend, Misty Hathaway.”

“Stovall had a pattern of using
violence to control women.”

 

 

115

 

Table 4.11 Illustrative Family Violence & Coercive Control Sentences Continued

 

 

 

 

 

 

 

 

Source Date Coercive Control Family Violence
Muskegon 6/7/06 “The shooting was apparently fueled
Chronicle by ‘prior confrontations’ between
Muskegon, MI Gresham and the suspect.”

Oakland Press 7/12/94 “Murder was the logical
Pontiac & offshoot of this continuous
Royal Oak, MI spousal abuse. That kind of
history put the guy in a state of
mind where this was the next
logical step. She wasn’t doing
what he wanted her to do.”
Saginaw News 3/ 1 1/99 “On the day of the slaying, Webb
Saginaw, MI went to the home of his girlfriend,
18-year-old Emily Moore, who was
preparing for church. They argued
about their deteriorating 4-month-old
relationship, the prosecutor said.”
Times Herald 4/18/98 “One day he’s happily married, the
Port Huron, MI next day he’s banned, he’s got a
divorce proceeding and one or two
criminal charges against him. That’s
enough to make anybody upset.”
Sentinel 9/22/06 “While Montcalm County “A man jilted over an impending
Standard deputies were investigating the divorce fatally shot the boyfriend of
Ionia, MI scene, Gregory Alexander his estranged wife.”
made repeated cell phone calls
to Tammara Alexander. He
told her he was going to take
his own life, and that he was
driving around the area.”
No Source No Date She got a PPO because “her “The suspect in two killings this
husband ran her off the road in week in Pleasant Lake and Ingham
September, displayed four or County may have targeted his
ﬁve loaded guns as he moved victims based on their roles in a
ﬂom their house the next difﬁcult divorce ﬂom his wife, the
month, and ‘said he will get county’s top prosecutor said Friday.”
me’ after a June 22 hearing in
their divorce case.”

 

 

 

 

The quotes provided in the above table demonstrate the breadth of information

 

provided in Michigan newspapers across seventeen years. These data also clarify the
dichotomy of conﬂict vs. premeditated explanations covered in quantitative analysis,
capturing the implicit IPV theory that coders were not able to collect in their protocol.

Findings also conﬁrmed that evidence of coercive control and family violence theories

116

can be gleaned out of journalistic accounts, and that these IPV related-ﬂames warrant
further attention, an issue that will be discussed in Chapters 5 and 6.

Summa o ualitative Results

Qualitative ﬁndings clearly compensated for the obstacles encountered in
quantitative analysis — namely, the ambiguous causal nature of collateral IPH. By coding
causal diagnoses in detail, four major categories emerged: (1) domestic disputes
(including the subcategories contained in Figure 4.1), (2) out-of-control perpetrator, (3)
extreme possessiveness (“if I can’t have you, then no one else will”), and perpetrator-
speciﬁc problems (i.e., individual and social factors). These themes inform rigid
dichotomous variables that were limited to conﬂict vs. intentional killings in the
quantitative protocol, providing the researcher with a deeper understanding of the
proposed causes of collateral IPH. This is especially salient when assessing “domestic
disputes” (an extremely ambiguous concept), as four distinct types were evident: (1)
custody / visitation issues, (2) general relationship problems, (3) relationship termination,
and (4) inﬁdelity, new partner, or rejection. These categories are noteworthy not only
because they clarify what a “domestic dispute” is, but they indicate that each dispute
stems ﬂom the principal relationship, a ﬁnding that lends support to family violence
theory. This conclusion supports quantitative results. The powerful role of the principal
IPV victim in the collateral IPH continues to appear over and over again.

Incident-speciﬁc themes generated in accordance with grounded theory also
contribute to and supplement overall results because they directly address research
question #2, which sought to understand the reporting styles associated with collateral

IPH. Despite the applicability of Entman’s (1993) generic framing scheme, ﬁve incident-

117

speciﬁc themes unique to this crime were evident: (1) focus on the principals, both
individually and as a “unit”, (2) the collateral victim’s background and / or involvement
in the crime, (3) focus on the perpetrator’s love of children (even if he killed them), (4)
the preponderance of shock (a theme echoed in quantitative analysis), and (5)
implications for child witnesses and survivors. These themes set collateral IPH apart
ﬂom typical homicide reporting and could only be detected through qualitative means.
In sum, qualitative ﬁndings conﬁrmed and extended quantitative results. The
“domestic dispute” was a prominent theme throughout, with quantitative results
indicating that 51.9% of articles were the outcome or direct result of a principal
relationship conﬂict, ﬁght or dispute. Additionally, 60% of articles attributed the
perpetrator’s motivation to commit the homicide(s) to the principal relationship, which
was also conﬁrmed during qualitative analysis. Most importantly, theoretical alignment
in the sample (family violence vs. coercive control) was constant across quantitative and
qualitative results. Family violence (i.e., conﬂict-oriented explanations) dominated,
while only a small amount of articles contained elements of coercive control (i.e.,
intentional, planned, premeditated). As such, results listed in Table 4.6 triangulate the
general trends uncovered in quantitative content analysis, but also provide textual

evidence of the clear theoretical divide in collateral IPH ﬂaming.

Summary

The results presented in this chapter illustrate the unique ﬂaming characteristics
associated with collateral intimate partner homicide. As a result of mixing qualitative
and quantitative methods, collateral IPH reporting trends between 1990 and 2007 were

captured along with the richer, more nuanced mechanisms involved with ﬂaming this

118

particular crime. Entman’s (1993) generic ﬂaming scheme was successfully applied,
however this typology alone did not cover the unique contexts associated with collateral
IPH. Thus, qualitative incident-speciﬁc ﬂames were generated to compensate for the
lack of description and understanding.

Additionally, the major goal this study to determine if newspaper articles
reporting collateral IPH were more closely aligned with family violence or coercive
control theory was achieved. Despite encountering ambiguous examples typically
associated with crime reporting (e. g., news briefs), it is clear that, whether intentionally
ﬂamed this way or not, journalists adopt a farme violence perspective when reporting
collateral IPH, regardless of collateral victim type. The sample primarily consisted of
homicides that were attributed to mutual relationship conﬂict between the principal IPV
victim and collateral perpetrator, even if the killings were gruesome or involved children.
Collateral victims clearly play an auxiliary role, indicating that this crime ultimately
relates back to preexisting intimate partner violence.

Lastly, there is strong evidence of a gender division in collateral IPH, as 100% of
perpetrators were men and 100% of principal victims were women. In other words, each
collateral victim was connected to the principal. . .whether a child, family member, friend,
or signiﬁcant other. Even if she was not killed in each story, someone that she cared
about was. This clear division between perpetrator and principal victim is not detected in
traditional IPV research (e. g. Garcia et al., 2007), lending support to the notion that
collateral killings capture a distinct typology of IPV, speciﬁcally the nature of coercive
control. This leads to the conclusion that by focusing on collateral killings only, a

different pattern of perpetrator behavior is uncovered. By removing the one-on-one

119

context so ﬂequently investigated in IPV research (i.e., one perpetrator and one victim),
the tactics and behaviors described by Stark (2007) and other feminist researchers
become more evident.

However these gender issues are not incorporated into newspaper reporting. As
quantitative and qualitative results indicate, collateral homicides are ﬂequently deemed
the result of relationship conﬂict and rarely identiﬁed as coercive or premeditated in
nature. Likewise, the female principal victim is commonly portrayed as having a role in
the homicide, contributing to the perpetrator’s anger or loss of control (i.e., divorce,
separation, denying custody or visitation, cheating, etc). Similarly, minimal reporting of
past violence or IPV is introduced, so it is difﬁcult to establish any sort of context to
connect past to present actions. While explanations frequently tie the principal’s actions
to the murders, they do not provide any insight into why they might be denying visitation
or divorcing their partner. These are just a few issues that warrant consideration as

research on this topic continues to expand.

120

Chapter 5: Discussion

General Diseassion

Results presented in chapter 4 provide preliminary ﬁndings that facilitate
numerous points of discussion, thus acting as a springboard for future research and
scholarly inquiry. The most striking ﬁnding to emerge from this dissertation is that
unlike previous research on IPH, men perpetrated 100% of collateral homicides. The
only inclusion criterion for this study was that the collateral IPH was related to the
principal intimate relationship. With that distinction alone a clear gender divide became
apparent. As such, four related conclusions will be addressed in this chapter: (1) the
dominance of symmetrical explanations for collateral IPH; (2) the incongruence of crime
severity and subsequent explanation; (3) the collateral’s role in the crime; and (4) the
ambiguous nature of coercive control in newspaper reporting. Each of these issues
highlights the dominance of family violence in journalistic ﬂaming as the majority of
articles focused on a symmetrical ﬁght, argument, or problem in the principal
relationship, rendering the collateral victim as an auxiliary to the issue. Additionally,
because a substantial amount of articles portrayed the collateral homicides as shocking
and unexpected, with relationship trouble as the root cause (i.e., family violence), the
heinous nature of the killings did not match explanations. As such, evidence of coercive
control was ambiguous at best. After discussing these issues, the chapter will close with

study limitations and recommendations for ﬁrture research.

121

Dominance of Symmetrical Explanations for Collateral IPH

Perhaps one of the most striking ﬁndings is the overwhelming use of relationship
conﬂict as the primary explanation for collateral IPH, with 51.9% of sample articles
providing this type of causal attribution. Phrases such as “domestic dispute”, “escalating
argument”, or “relationship gone bad” offer a glimpse into how stories are constructed
and interpreted by the mass media, with most pointing at common family transitions or
relationship problems as a trigger. For instance, divorce and separation were ﬂequently
cited as reasons why the perpetrator “lost it” and (often) killed his partner and at least one
collateral victim. Although it is well established that divorce, marital separation, and
child custody problems act as stressors in relationships, they rarely end in homicide. This
ﬁnding leads media scholars to question if evidence is omitted ﬂom certain newspaper
articles, speciﬁcally any history of IPV and coercive control. One highly covered
collateral homicide case illustrates this point. Terry Hall had an extensive history of
domestic violence and coercive control, and a handﬁrl of stories introduced this
information. One noted that in addition to threatening Diane Hall regularly, he exhibited
other control tactics such as the killing of animals (he shot the family dog, put an axe
through their kitten’s head, and twisted heads off of goldﬁsh). When she ﬁnally
mustered the courage to leave him for good, he killed her and her parents in a Cracker
Barrel parking lot. “Hall threw him out for good. He began following her and,
confronting her several times a week at the store, Jack’s Cracker Barrel.” In fact, she was

so scared of him that “she would sleep with a butcher knife under her pillow.” This

example illustrates asymmetrical violence and the use of coercive control, as it points to

122

Terry Allen as the sole aggressor. However, a separate article paints a different picture of
the relationship: “People who knew the Halls’ said they were aware of their domestic
problems. They had reconciled without marrying after the divorce and had many stormy

3” This journalist also incorporated neighbors’ reactions of shock.

breakups since then.
“Everyone’s in shock. There was no indication that this would happen.” What changed?
One article introduced extreme violent behavior; while the other pointed to relationship
issues and shock that Terry could do such a thing. By ﬂaming a story like this, the
journalist implicitly shifts blame, as the woman goes ﬂom “victim” to “participant”. The
use of conﬂict and family problems to explain collateral IPH is dangerous because it
inherently blames the victim, attributing symmetrical cause while virtually ignoring
context. For instance, few articles in this sample reported prior perpetrator abuse or
violence (The Terry Hall case is an extreme outlier), and only a ﬂaction explicitly
mentioned “domestic violence” or some variation of it in the story. This lack of context
is tremendously important because journalists and their audience only catch a glimpse of
the problem, thus leading to incorrect conclusions about the nature of IPV and collateral
IPH.

If one refers back to Chapter 2, a lack of context in determining the causes and
consequences of IPV is one of the major ongoing debates between feminist and farme
violence scholars. Feminists ﬂequently cite that a limitation of family violence research
is the exclusion of violence context when explaining IPV (e. g., Anderson, 2005), as

differences between male and female-perpetrated domestic violence are not a focal point

of analysis. For instance, feminist theorists argue that the Conﬂict Tactics Scales (CTS)

 

3 Emphasis is author’s own

123

does not take context into account (speciﬁcally the impact of gender) and fear that what
family violence theorists represent as IPV may actually be a different phenomenon
altogether (i.e., “couple violence”). This concept is different ﬂom “patriarchal terrorism”
and “coercive control”, which are regularly reported by feminist scholars. Ultimately, a
lack of relationship context leads to a general belief that gender symmetry exists in IPV
and that if a homicide occurs, both partners contributed to the problem. This assumption
strips the crime of its male-dominated nature and runs the risk of blaming the victim, an
outcome that must be stopped if societal understanding of violence against women is to
improve. Gender clearly plays a role in collateral IPH as 100% of perpetrators are male
partners or ex-partners and 100% of principal IPV victims are women, a striking ﬁnding
that lends support to coercive control theory.

A clear example of victim blaming can be found in articles where perceived
inﬁdelity was occurring (even if the principal IPV victim had separated from the
perpetrator). Reporters and their informants ﬂequently acknowledged that she (and her
new partner) contributed to the outcome in some way because they “should have known
better”. Additionally, even in stories where police and / or the legal system admit that
services were not adequate in protecting the victim (e. g. the three-day rule, or waiting
three days to ﬁle domestic violence charges, failing to arrest, etc), many law enforcement
representatives provided responses that placed responsibility on the principal IPV victim,
commenting that the woman’s delayed response or failing to offer all necessary
information about her abuse contributed to the problem. Comments such as these place
responsibility back on the principal victim, as they should have “done more”. Therefore,

not only did the principal victim contribute to the collateral IPH by ﬁghting with,

124

divorcing, or separating ﬂom her partner, when his violence was a plausible threat, she
still did not do what she was supposed to.

This is not to say that there weren’t situations in which a perpetrator was mentally
ill, on drugs, upset over a divorce, or out of control at the time of the collateral IPH;
however, by defaulting on relationship conﬂict to explain a majority of homicides has the
potential to stereotype IPV and collateral IPH as a gender symmetrical crime, which fails
to thoroughly explain the phenomenon. Since the present sample was a secondary
source, there is no way to know what else was occurring in each relationship (one would
need access to court or police records, child protective services reports, etc); rather
researchers had to rely on journalistic accounts and interpret text based on the
information provided. This is an issue that mass media researchers must take into
consideration, regardless of their personal alignment with feminist or family violence

theories.

Does the Explanation fit the Crime?

Quantitative results indicate that when an explanation is provided it tends to
highlight principal relationship conﬂict or ﬁghts. Qualitative ﬁndings support this,
determining that “domestic disputes”, including general relationship problems, make up a
majority of causal explanations. Additionally, the notion that the perpetrator was “out of
control” or upset / jealous at something going on in his primary relationship were
ﬂequently posited as potential causes. In addition to this, processes and issues revolving
around the perpetrator’s cognitive state and social life, including mental health problems,

drug abuse, and ﬁnancial woes were introduced. Although these explanations are logical

125

contributors to relationship strife and conﬂict, a tremendous disparity exists between
these explanations and the gruesomeness of many killings. For example, ﬁndings
indicate that regardless of collateral victim type (e. g., a child, elderly woman, new
boyﬁiend, friend) there was no dz'ﬂerence between the types of explanation offered.
Conﬂict-related results dominated newspaper articles. This creates cognitive dissonance
on behalf of the reader, as how can someone justify killing an infant or 63-year-old
woman because of a marital or ﬁnancial problem? This possibly explains why a third of
all articles ﬂamed the incident as shocking or unexpected, even if a history of abuse and
violence was present in the principal relationship. Journalists and their informants are
obviously attempting to make sense out of an upsetting, implausible situation, which is
why simple explanations were ﬂequently offered. Regardless, a gap exists between the
simplicity of explanation and crime that requires further consideration. A few examples
illustrate this point.
0 Reco Jones killed his girlﬁiend and four children because he wanted to
keep the relationship a secret from his other girlﬁ'iends.
0 Anthony Bailey shot and killed three of his children, severely injuring the
fourth, Antonia, because he was upset over an impending divorce.
0 Roger Thompson killed his four children and their mother because of an
argument he was having with her.
0 After Gloria Pitts broke up with Luther Jenkins, he killed her, Alicia
Jackson, and her two children, J amon and A’Janneya.
0 J immie Reed Jr. killed his daughter and girlﬁiend because he did not want

to pay child support and wanted to keep the relationship a secret.

126

The above examples illustrate how gruesome homicides are often attributed to
simple causes, including the desire to avoid paying child support, settle an ongoing
argument, or deal with the prospect of divorce. These deaths were not quick and painless
either; they ﬂequently involved stabbing, shooting at close range, or setting children on
ﬁre. One article even described the death of 9-day—old Omar Dean Jr.: “I heard a
gunshot. I opened my eyes and my baby’s head had just exploded.”

So why does this discrepancy exist? Does the explanation actually ﬁt the crime?
The simple answer is no. A person does not set children on ﬁre, stab babies in the heart,
or shoot elderly people in the head because of a domestic spat gone awry. The disparity
in brutality of crime -) explanation reinforces the conclusion that collateral IPH taps into
a distinct typology of IPV, namely coercive control. Although tenets of coercive control
have yet to be incorporated into news reports, this study clearly illustrates that other
forces are at work in collateral IPH, so media scholars must look further into the sources
of information that contribute to story construction as well as what journalists do with
available data when writing an IPV-related article. Additionally one must consider the
internal processes of the journalist and why they offer such explanations for heinous

crimes.

Where do Collateral Victims ﬁt?

The overarching goal of this dissertation was to uncover the ﬂaming patterns
associated with collateral intimate partner homicide. Approaching this exploratory study
with a combination of traditional mass media and IPV theories, a preliminary

understanding of the primary perpetrators and victims was discovered. However, even

127

though collateral victims were killed in every article, the primary cause and / or
motivation for the homicide came back to the principal IPV victim and perpetrator in
every case. The only factor tying collateral individuals into this complex web of causes
and consequences was their relation to the primary couple, most notably the principal
IPV victim. Since the most ﬂequently occurring type of collateral in the dataset was the
principal victim’s child, researchers can begin to see how closely their deaths tie into
general IPV and child witnessing literatures. Additionally, in no story did the perpetrator
kill members of his family; with the only exception being children he shared with the
principal victim. There were eighty-eight cases of principal victim family deaths,
including mothers, fathers, sisters, and aunts. All of the ﬁiends killed (n=28) were the
principal victim’s as well. So this leaves scholars with a question, “where do collateral
victims ﬁt?” This study, although exploratory, offers some clues to this inquiry. Clearly,
the reporting ﬂame highlights the principal female IPV victim and male perpetrator ﬁrst
and then ties the collateral in to the problem. Whether they were in the “wrong place at
the wrong time” or perceived as an impediment to a successful reconciliation, their
involvement was auxiliary to the overarching issue. However, this conclusion is
incomplete, as there is still much more to learn about this crime and the collateral’s
speciﬁc role. Although some background information was provided on collateral
victims, results are still inconclusive. However, at this stage in the research process the
term “collateral victim” is obviously appropriate, as they are truly portrayed as “collateral

damage” in newspaper reports.

128

The Ambiguous Nature of Coercive Control in Collateral IPH

It is clear ﬂom this study that assumptions associated with family violence theory
were more ﬂequently incorporated into collateral IPH ﬂames than coercive control. In
fact, only n=11 (5.3 %) articles on collateral IPH cited premeditation, while 23.1% and
2.4% respectively mentioned prior harm or abuse against the principal IPV and collateral
victim. Through qualitative theoretical content analysis some examples of coercive
control tactics were identiﬁed, but they were still in the minority. However, even when
sentences were identiﬁed as “coercive control”, they were ambiguous and often followed
by a statement about a family-speciﬁc issue (i.e. divorce, a ﬁght, cheating, etc.) For
example, when reporting on the collateral homicide of a principal victim’s (Fochtrnan)
friend, the journalist ﬁrst stated, “Taylor (the perpetrator) and F ochtrnan had an ongoing
dispute.” That same story introduced prior abuse against Fochtrnan: “Police say they
have several reports of alleged abuse by Taylor on Fochtrnan.” A witness even testiﬁed
that a relationship dispute could not instigate such a crime: “I don’t believe he went
crazy. I believe he was pissed off about something. I get pissed off sometimes but I
don’t try to kill people.” In another example, Reco Jones’ relationship with his girlﬂiend
was described as “hot” and “cold”: “When it got hot, they broke up. When it got cooler
they got back together. They just needed space to breathe sometimes.” The story then
goes on to provide evidence of coercive control, “In jail he began calling Bellamy collect.
At ﬁrst his calls were wistful. Then they took on a more demanding tone. ‘Who are you

seeing? Are you cheating on me?’”

129

This type of reporting is sending mixed messages to the public. So what really
caused the killing? Was it a pattern of abuse that ultimately ended in homicide? Or was
it the result of conﬂict, ﬁght, or stress that just got out of hand? If one refers the tenets of
coercive control theory, IPH and collateral IPH are the ﬁnal step in a complex web of
power and control tactics, with the abuser engaging in various behaviors to maintain
control over his partner (e. g., monitoring car mileage, making her keep a log of daily
activities, physical aggression, threatening to take her children away, psychological
torment, such as sleep disruption, etc). Killing his woman’s children or parents
represents the most extreme control tactic. Unfortunately we still can’t be sure by
analyzing newspaper articles alone. One must draw cautious conclusions at this point.
Yet, this type of ambiguity is not going to enhance public understanding of IPV and its
consequences; rather it is going to lead to uncertain and often false conclusions, which is
why journalists and policymakers must pay attention to the sensitive nature of reporting

intimate crimes, an issue explored in Chapter 6.

Limitations

The data for the current study are comprehensive, containing all available articles
ﬂom Michigan newspapers across seventeen years. It is important to note, however, that
individuals operating the clipping service were not trained in the dynamics of intimate
partner violence, and many included stories were ambiguous, requiring a certain amount
of subjective interpretation. As such, assumptions had to be made regarding the nature of
the principal relationship, as the connections were not always clearly deﬁned.

Additionally, as a researcher/advocate adopting a feminist perspective, an implicit bias

130

that most intimate partner homicides are rooted in gendered violence and male
domination played a role throughout the dissertation process. However, as with any
social research endeavor, personal and professional biases were acknowledged and
controlled throughout the study (see “trustworthiness of ﬁndings”). Still, primary
research is necessary to supplement ﬁndings ﬂom this study. Newspaper articles should
be analyzed in conjunction with other secondary sources (e.g., death reports, police
reports) to ﬁll in potential gaps and increase objectivity.

Second, with this speciﬁc secondary source comes a certain amount of bias in
reporting. Analysis was based on only the information and textual accounts provided to
journalists and selected informants. As mentioned earlier, there are many factors that
inﬂuence news content, so potentially useful data may have been omitted due to space
constraints and editorial requirements. By doing so, slants or skews may occur. For
example, objective police reports and court records contributed a majority of the
information on the crime(s), with sensitive or graphic neighbor and family testimony
sprinkled in for shock value. However, researchers can’t be sure if journalists even had
access to abuse history. Unfortunately, this type of background information is necessary

if a full, contextual picture of collateral IPH is to be achieved in the mass media.

Future Research

The present study was the ﬁrst to investigate collateral IPH and how traditional
theories of intimate partner violence informed newspaper ﬂaming. Since this dissertation
was primarily exploratory and descriptive, the next logical step is to utilize generic and
incident-speciﬁc collateral IPH media ﬂames to conduct research studies in three areas:

(1) primary data collection with members of the media industry (i.e., journalists and

131

editors) regarding personal and professional motivations for story construction, (2)
experimental research with media consumers to document reactions to collateral IPH
ﬂaming styles (audience ﬂaming); particularly ﬂames’ impact on opinions and attitudes
towards IPV; and (3) assess cases in which females murder their children, a partner’s
family, or new love interests to determine if differences truly exist between the sexes.
This particular line of research will illuminate if the nature of coercive control and
domination in collateral IPH is gender symmetrical or truly a male-perpetrated act.

Additionally, four further areas of research warrant future consideration:

0 New vs. Old Media — As electronic sources of news become more and more
popular, mass media researchers must investigate ﬂaming differences between old
and new media. One could compare a variety of information sources including
online newspapers and traditional newspapers, television reports and specialized
blogs.

0 The Tendency to Report IPH as a Gender Symmetrical Act — Media
researchers should continue to explore the issue of gender symmetry in reporting,
moving ﬂom exploration towards explanation. This type of research requires
many data sources, including police reports, newspaper articles, and primary
interviews with media, criminal justice personnel, and survivors.

0 The Health and Social Consequences of IPV - The investigation of intimate
partner violence is truly an interdisciplinary effort that requires scholars ﬂom
mass media, women’s studies, public health / epidemiology, and criminal justice
to not only collaborate, but explore the distinct dimensions of IPV that fall within

and across disciplinary boundaries. Although a large literature exists on child

132

witnessing and the health consequences of abuse, there is no literature on
collateral IPH. An especially salient situation to explore is one in which the
principal partner survives. Researchers must determine the impact of collateral
deaths, including health and social consequences and implications for future
intimate relationships.

Coping with losing a collateral — As an extension of the health and social
consequences of IPV, researchers need to investigate the effects on the principal
IPV victim after a loved one has been killed by her abusive partner. Just as social
scientists investigate the coping strategies (both adaptive and maladaptive) and
recovery processes necessary for children who witness abuse or lose a mother to
IPV, this phenomenon also warrants investigation. How does a mother move on
after her children are killed? How does the family restructure itself if her mother
or father is murdered? These questions are salient across many contexts,
including the social service, criminal justice, and public health realm.
Comparison to other crimes — Media researchers need to examine collateral IPH
in the broader context of general crime reporting in order to illuminate similarities

and differences in reporting different crimes and victims.

133

Chapter 6: Translational Science

An exploratory study such as this dissertation provides numerous avenues for
exploration and application. However, perhaps one of the most important outcomes of
this study is its applicability to professional realms outside of academia. Not only will
social scientists and media scholars beneﬁt ﬂom ﬁndings, but implications extend to
other professionals as well. For instance, policymakers, journalists, advocates, and
service providers can utilize information about collateral intimate partner homicide and
apply it in their own work. As such, this chapter is dedicated to delineating the unique
applications for those outside the academic realm, and how results are translated into
tools for professionals that come in contact with survivors of intimate partner violence
and homicide.

The notion of translational science is a prominent goal of feminist research, as
scientiﬁc ﬁndings are expected to be useﬁrl for study participants and the general
population. Since it is the only theory to emerge ﬂom grassroots advocacy campaigns
(Rarnazanoglu & Holland, 2005; Stark, 2007), feminist scholarship is dedicated to
empowerment and giving voice to those who cannot speak for themselves. Therefore,
pertinent results are extrapolated and packaged in such a way that is understandable and
useful to a diversity of individuals. A secondary goal of feminist research, then, is to
elicit change at a societal level. Violence against women is rooted in patriarchy and the
social structures that support and perpetuate male domination. For this reason alone,
macro-level themes generated in this study are of extreme importance if one wants to

understand the societal-level changes necessary to address collateral IPH as a male-

134

perpetrated coercive control tactic. Therefore, societal-level implications uncovered in
this dissertation, along with general reporting trends will serve as preliminary guidelines

for journalists, policymakers, and service providers.

Mass Media Implications of Collateral IPH

One of the most prominent problems detected in this dissertation was the general
lack of understanding of IPV at a societal level. Since the media have the ability to reach
a large amount of people simultaneously, they remain one of the most important
information outlets for the US population. Therefore, if collateral IPH is poorly ﬂamed,
omitting important contextual information (e.g., past domestic violence) and attributing
murders to symmetrical disputes within the principal relationship, there is the potential to
perpetuate myths of domestic violence, encourage victim blaming, ad inﬂuence public
perceptions and attitudes of IPV. Throughout the century, scholars have demonstrated
the power of the media, as numerous studies have found that news media (via editors /
journalists) have the potential to impact people’s perceptions of an issue, including IPV,
rape, and other public threats (Bems, 2004; Franiuk, Seefelt, & Vandello, 2008; Ridout,
Grosse, & Appleton, 2008). As Zillmann (2002) said, “. . .The capacity to reach large
audiences carries with it the risk of misleading the public in case the disseminated
information proves to be distorted and inaccurate or simply in error” (p. 21 ). Therefore, it
the media’s responsibility to provide accurate accounts of a phenomenon while also being
prepared to deal with consequences if distortion occurs (Zillmann, 2002). Clearly the

ﬁrst step is to disentangle the dispute between family violence and coercive control

135

theories. However, recommendations on how journalists, policymakers, and advocates

can begin to accomplish this will be discussed below.

Recommendations and Guidelines for Journalists

Since the media have the responsibility to report coherent, accurate, and
comprehensive news to the public, journalists should be among the ﬁrst to utilize ﬁndings
ﬂom this study. This group stands to beneﬁt ﬂom the present ﬁndings, as tangible
guidelines and recommendations can be explicated and put to into immediate use. As
such, four major recommendations, or “best practices” are listed below. These topics are
dynamic and can be incorporated into nearly every stage of the reporting process
including pre-investigation and training (i.e., before a crime is even reported), homicide
investigation, story construction, and public presentation. In order to remain concise,
recommendations are bulleted.

1. Responsible Reporting -— A clear gender division is evident in the perpetration of
collateral IPH. As such, journalists should keep this phenomenon in mind when
investigating IPV-related incidents. Reporters must take a step back, incorporate a
critical lens, and reﬂect on the nature of the crime, including what factors beyond
a “dispute” might contribute to the killing of children, parents, and other loved
ones. Therefore, as much relationship context as possible should be collected so
that a complete account is provided to the public. This includes interviewing key
witnesses and experts that go beyond default respondents (i.e., neighbors or
bystanders), speciﬁcally those that know the players or are familiar with the

dynamics of IPV. Assessing crime records would also help construct stories.

136

2.

4.

Continuing Education — Since results ﬂom this study are novel and media-
oriented, journalists can immediately expand and update their current knowledge
base of intimate partner violence, intimate partner homicide, and collateral
intimate partner homicide. It is recommended that all major media outlets keep a
resource manual that contains relevant statistics, policies, and necessary
references available to their journalists, especially those assigned to crime
investigation. Regular training and information sessions are also highly
recommended because the better their understanding of these phenomena, the
more accurate their news stories.

Comprehensive Inter-Organizational Communication — Gathering information
and reporting crimes are not a solitary efforts. They require insight ﬂom
numerous agencies and systems, including police departments, legal
representatives, social service pro grams, and advocacy organizations. Therefore,
in conjunction with improving contextual reporting of collateral IPH, inter-
organizational communicative efforts should be launched, as regular interaction
between these entities will improve reporting. Most importantly, it is vital that
media representatives and IPV experts maintain constant communication and
collaborate on resource / training manuals regularly. Speciﬁc examples will be
presented later.

Social Activism — As journalists become more informed about IPV and its
consequences, they gain the power to end the perpetuation of domestic violence

myths (see Zillmann, 2002). Potential positive outcomes of adopting this social

137

activist role are numerous. Perhaps the most salient, however, is fostering a
societal—level support network for victims of violence. This can be achieved by
accurate reporting, providing necessary resources and hotline information. By
making this a regular practice, victims will feel empowered, supported, and may

decrease risk of future harm.

Public Policy Implications

A notable amount of aftermath stories pointed to gaps in police policy and how
changes to these policies might prevent collateral IPH or protect victims of intimate
partner violence ﬂom ﬁrrther harm. Therefore, policy change recommendations for
collateral IPH are analogous to those of traditional IPV or domestic violence policy, as
they are inherently connected. If one wants to protect collaterals and prevent their death,
laws that protect principal IPV victims must ﬁrst be enhanced. As highlighted in Chapter
4, nearly 58% of principal IPV victims were killed along with collaterals, so these groups
cannot and should not be delineated.

As evidenced in this sample, IPV policy improvements must start at the local
level. Although the Violence Against Women Act (V AWA) of 1994 was a notable
societal-level achievement, sustainable policy change, especially regarding crimes that
receive minimal national attention, must progress ﬂom local to state to national levels.
For example, many victims of IPH made numerous attempts at police protection but were
unsuccessful because of outdated laws or failing to meet an arrest requirement, such as
physical evidence of abuse. In discussing the danger associated with the “three-day

rule”, an informant noted, “It’s the kind of policy that departments have used historically

138

to discourage domestic violence victims ﬂom using the criminal justice system. . .they

have these hoops for women to jump through and when the women get discouraged, they

blame the women.”

Therefore change must begin with local police departments and communities and

move out ﬂom there. Concerned citizens, advocacy groups, and social service agencies

should contact their local police departments and representatives in order to begin the

process. This dataset contains an excellent collection of issues that should serve as a

starting point for policymakers, so the following policy concerns should be considered

ﬁrst.

The Three-Day Rule — In many jurisdictions, IPV victims are told to wait three
days to ﬁle charges. Unfortunately, violence may escalate in this time period,
increasing risk of IPH and collateral IPH.

Mandatory Arrests in Domestic Violence — Lawmakers should consider
adjusting the evidence needed to arrest an abuser. Many communities still require
signs of physical abuse before making an arrest. Macomb County is an exemplar
of policy change in this area, as they now require their ofﬁcers to make arrests
based on information and belief that an assault has occurred.

Stalking Laws — Stalking laws should be revised so that they better protect IPV
victims. Stalking laws are unique in that they can serve as an additional “layer”

of protection if executed properly.

' Uniform Law Enforcement Policy - Consideration should be given to

establishing a standardized protocol for protecting IPV victims and their families /

friends ﬂom abuse. There is potential for creating research-informed policy at

139

both state and national levels, sending the message that as a society, the United

States will not tolerate IPV, IPH, or collateral IPH.

According to a domestic violence expert, “Law is a lot better than it used to be in
Michigan, especially with the stalking and domestic violence laws, but there are still
improvements that need to be made to take every advantage of those laws.” This quote is
a call clear call for tailored policy change that takes the unique challenges and
demographics of battered women into account. However, in order for this change to
occur, societal attitudes must shift. This begins with a more contextual understanding of
intimate partner homicide. As one article concludes, “But all of society needs to own up
to this.” The public needs to view IPV as a serious social and public health problem and
work together to reform laws so that they better protect victims and their loved ones ﬂom

harm.

Service Provider Implications

Professionals and advocates working in the violence against women movement
(V AW) can also utilize ﬁndings ﬂom this dissertation, as it is vital that those working
directly with victims and their families keep their ﬁnger on the “pulse” of IPV reporting
practices.

In addition to regularly updating and collecting newspaper articles, domestic
violence programs and media professionals need to forge a stronger partnership. If these
two groups improve inter-organizational communication, a mutually beneﬁcial
relationship will emerge. The production of collaborative training manuals with media

outlets is one worthwhile effort. For example, the Michigan Coalition Against Domestic

140

and Sexual Violence (MCADSV) currently provides media toolkits for journalists and
service providers (Michigan Coalition Against Domestic and Sexual Violence, 2008).
Additionally, the Rhode Island Coalition Against Domestic Violence (RICADV)
developed a handbook of best practices for reporting IPV that acknowledges the
numerous constraints journalists work under (Ryan, Anastario, & DaCunha, 2006). Their
particular model has become the standard for other states that have created their own
guidebooks as well (e.g., Arizona Coalition Against Domestic Violence, 2003;
Washington State Coalition Against Domestic Violence, 2002).

Lastly, advocates and service providers should incorporate information on
collateral IPV and IPH into their training protocols and manuals. Most agencies
primarily focus their training efforts on the principal IPV victim. However crisis
intervention should be extended to family members and friends as well. For example, the
National Coalition Against Domestic Violence (N CADV) currently publishes training
manuals on numerous topics such as teen dating violence, the provision of services to
disabled populations, and criminal justice issues (i.e., child custody issues, child
witnessing of domestic violence, failure to protect, etc). Each of these resources is ﬂeely
accessible to advocacy programs across the country (National Coalition Against
Domestic Violence, 2008) and should be updated as new research is published. By doing
so, agencies will be better equipped to respond to principal IPV victims that have lost
children, parents, or ﬂiends. This is not to say crisis support and counseling are not
currently offered to a range of individuals, but as researchers learn more about the

intricacies of collateral IPH, training should be adjusted as well.

141

Conclusion

Findings ﬂom this study have numerous implications for those within and outside
academia. Not only will these exploratory results facilitate more sophisticated inquiries
into the nature and reporting of collateral IPH, they can be used in training curricula for
social service providers, journalists, legislators, police, lawyers etc. The general public
will also beneﬁt ﬂom more comprehensive, contextual reporting on IPH. If people are
only provided with minimal background or misinformation on intimate relationships that
end in violence (e.g., they engaged in a “dispute”), the public will never get a
comprehensive picture of the methods and tactics behind male-perpetrated IPV. Lastly,
as the theory of coercive control continues to expand, further insight into the controlling
nature of IPH can be diffused to the public, dispelling long-held myths of domestic

violence and its primary causes.

142

APPENDICIES

143

APPENDIX A

Newspapers Included in Dissertation Sample

144

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

# Newgpaﬂr # Of Articles % Of Sample
1 Alpena News l 0.47%
2 Ann Arbor News 6 2.79%
3 The Argus Press 2 0.93%
4 Battle Creek Enquirer 1 1 5.12%
5 Bay City Times 9 4.19%
6 Cheboygan Daily Tribune 1 0.47%
7 The Chesterﬁeld Review 1 0.47%
8 Coldwater Daily Reporter 1 0.47%
9 Commercial News 3 1.40%
10 The Daily Globe 2 0.93%
l 1 Daily News — Midland 3 1.40%
12 The Daily News — Greenville l 0.47%
13 The Daily News — Iron Mountain 2 0.93%
14 Daily Telegram 2 0.93%
15 Daily Tribune 1 0.47%
16 Detroit Free Press 30 13.95%
17 Detroit News and Free Press 7 3.26%
18 Detroit News 12 5.58%
19 Flint Journal 8 3.72%
20 Gaylord Herald Times 3 1.40%
21 Grand Haven Tribune 2 0.93%
22 Grand Rapids Press 10 4.65%
23 Greenville Daily News l 0.47%
24 The Herald Palladium 2 0.93%
25 Hillsdale Daily News 4 1.86%
26 Holland Sentinel 3 1.40%
27 Jackson Citizen Patriot 8 3.72%
28 Kalamazoo Gazette 12 5.58%
29 Lansing State Journal 19 8.84%
30 Livingston County Press 3 1.40%
31 The Macomb Daily 2 0.93%
32 Maple Valley News l 0.47%
33 Mining Journal 1 0.47%
34 Morning Sun 1 0.47%
35 Muskegon Chronicle 16 7.44%
36 News Advocate l 0.47%
37 Niles Daily Star I 0.47%
38 Oakland County Legal Review 1 0.47%
39 Oakland Press 6 2.79%
40 Ogemaw County Herald 1 0.47%
41 Oscoda County Herald 1 0.47%
42 The Pioneer 2 0.93%
43 Saginaw News 3 1.40%
44 Sentinel Standard 1 0.47%
45 Sturgis Journal 2 0.93%
46 Times Herald 1 0.47%
47 Traverse City Record Eagle 1 0.47%
48 Wyandotte News Herald 1 0.47%
49 Unknown Newspaper/ Publication Date 2 0.93%
TOTAL = 215 100%

 

145

 

APPENDIX B

Collateral IPH Coding Sheet

146

Collateral IPH Coding Sheet

Coder ID (Bryce=1; Brendan=2): __
Stag Identiﬁcation
(A) SIDV1: Story ID number: __
(B) SIDV2: Story Date (mm/dd/yy) _
Newsgmr Information
(C) NPV1: Newspaper ID number: __
(D) NPV2: Sex of Reporter: _

1=male

2=fema|e

3=unknown

(1) NPV3: Victim] Perpetrator Story Position

Please document when the following individuals were introduced in the story: Mark 1

(ﬁrst), 2 (second), or 3 (third)
Principal IPV Victim

Collateral IPV Victim(s)
Perpetrator

General Stag Characteristics
(2) SCVt: Crime Stage:

0=cantteﬂ
1=crime commission through arrest

2=judicial proceedings (events that occur from trial to sentencing and legal appeals)
3=afterm ath (an aftermath story is one that discusses the implications, outcomes of
the homicide(s), including family reactions to sentencing, funerals, etc.)

(3) SCV2: Crime Victims (for the present collateral IPH ONLY):

O=not mentioned
1 =mentioned

SCV2a Principal IPV victim (current or ex-partner of perpetrator)
SCV2b Children of principal victim, perpetrator, or both

SCV2e Family of IPV victim

SCV2d Friends of IPV victim

SCV2e New boyfriend, girlfriend

SCV2f Neighbors/bystanders

SCV29 Other (write in)

147

(4) SCV3: Suicide: Did the perpetrator commit suicide in this article?

0=no
1=yes

Princiggl Victim Portrayals
(5) PV1: Principal IPV Victim Sex _
1=male 2=female 0=not provided
(6) PV2: Was the principal IPV victim killed in this story? __

O=no
1 =yes

Collateral Victim Portrayals
(7) CV1: Total Number of collateral victims:

Please complete the following for each collateral victim in article (Note: please default on
the order in which collaterals are presented in article for organizational purposes) -

Collateral Victim #1 (0=no; 1=yes)

(8) CV2: Collateral Victim Name:

 

(9) CV3: Collateral Victim Age:

 

(10) CV4: Collateral victim #1 sex

Male
Female

(11) CV5: Collateral victim’s relationship to principal IPV victim and / or perpetrator:

Victim’s child

Perpetrators child

Child of both principal IPV victim and perpetrator
Victim’s mother

Victim’s father

Perpetrator’s mother

Perpetrator’s father _

IPV Victim’s new boyfriend, love interest
Other Family

Fnend

Other (write-in)

(12) CV6: Was any explanation provided for homicide?

YES
NO

148

(13) CV7: If yes, what were they?

CV7a Accident (i.e. individual caught in the crossﬁre)

CV7b Result of principal relationship conﬂict or ﬁght (i.e. ongoing argument; marital
dispute; custody or visitation issues; relationship problems including divorce,
separation, or being “estranged”)

CV7c Intentional or Premeditated (i.e. the perpetrator intended to kill them as the
homicide(s) was/were planned in advance)

CV7e Story did not provide information _

CV7f Other explanation __

Collateral Victim #2 (0=no; 1=yes)
(14) CV8: Collateral Victim #2 Name:
(15) CV9: Collateral Victim #2 Age:
(16) CV10: Collateral victim #2 sex

Male
Female

(17) CV11: Collateral victim’s relationship to principal IPV victim and l or perpetrator:

Victim’s child

Perpetrator’s child

Child of both principal IPV victim and perpetrator
Victim’s mother

Victim’s father

Perpetrator’s mother
Perpetrator’s father

New boyfriend, love interest
Other family

Fnend

Other (write-in)

(18) CV12: Was any explanation provided for homicide?

YES
NO

(19) CV13: If yes, what were they?
CV13a Accident (i. e. individual caught in the crossﬁre)
CV13b Result of principal relationship conﬂict or ﬁght (i.e. ongoing argument; marital
dispute; relationship problems including divorce, separation, or being
“estranged”)
CV13c Intentional or Premeditated (i.e. the perpetrator intended to kill them as the
homicide(s) was/were planned in advance)
CV13d Custody or visitation dispute
CV13e Story did not provide information _
CV13f Other explanation _

 

149

Collateral Victim #3 (0=no; 1=yes)
(20) CV14: Collateral victim #3 name:
(21) CV15: Collateral victim #3 age:
(22) CV16: Collateral victim #3 sex

Male
Female

(23) CV17: Collateral victim’s relationship to principal IPV victim and l or perpetrator:

Victim’s child

Perpetrator's child

Child of both principal IPV victim and perpetrator
Victim’s mother

Victim’s father

Perpetrator’s mother
Perpetrator’s father

New boyfriend, love interest
Other family

Fnend

Other (write-in)

(24) CV18: Was any explanation provided for homicide?

YES
NO

(25) CV19: If yes, what were they?

CV19a Accident (i.e. individual caught in the crossﬁre)

CV19b Result of principal relationship conﬂict or ﬁght (i.e. ongoing argument; marital
dispute; relationship problems including divorce, separation, or being
“estranged”)

CV19c Intentional or Premeditated (i.e. the perpetrator intended to kill them as the
homicide(s) was/were planned in advance)

CV19d Custody or visitation dispute

CV19e Story did not provide information

CV19f Other explanation _

Collateral Victim #4 (0=no; 1=yes)
(26) CV20: Collateral victim #4 name:
(27) CV21: Collateral victim #4 age:
(28) CV22: Collateral victim #4 sex

Male
Female

150

(29) CV23: Collateral victim’s relationship to principal IPV victim and I or perpetrator:

Victim’s child

Perpetrator's child

Child of both principal IPV victim and perpetrator _
Victim’s mother

Victim’s father

Perpetrator’s mother
Perpetrator’s father

New boyfriend, love interest
Other family

Fnend

Other (write-in)

(30) CV24: Was any explanation provided for homicide?

YES
NO

(31) CV25: If yes, what were they?

CV25a Accident (i.e. individual caught in the crossﬁre)

CV25b Result of principal relationship conﬂict or ﬁght (i.e. ongoing argument; marital
dispute; relationship problems including divorce, separation, or being
“estranged”)

CV25c Intentional or Premeditated (i.e. the perpetrator intended to kill them as the
homicide(s) was/were planned in advance)

CV25d Custody or visitation dispute

CV25e Story did not provide information _

CV25f Other explanation _

Perpetrator Information
(32) PP1: Perpetrator Sex _
1=man 2=woman
(33) PP2: Perpetrator history of violence: _

0=No information included
1=lnformation of prior violence included

(34) PP3: Perpetrator’s prior violence (actual physical violence or threats of physical
violence): Code all that apply (0=N/A or not mentioned, 1=mentioned).

PP3a Prior violence against IPV victim

PP3b Prior violence against IPV victim’s children __

PP3c Prior violence against own children _

PP3d Prior violence against other family members __

PP3e Prior violence against IPV victim’s new intimate partner _
PP3f Prior violence against non-family members _

PP3g Personal Protection Orders issued against perpetrator

PP3h Other (write in; assign unique number) _

151

(35) PP4: Homicide(s) Portrayal: Based on witness testimony regarding the current homicide,
how is the incident portrayed (mark 0 or 1)?

PP4a It was shocking, unexpected (look for quotes describing shock, surprise, and
information on how he was a good father, husband, etc.)

PP4b It was deliberate, expected (i.e., the homicide was premeditated or
calculated)

PP4c The story is not clear

(36) PP5: Was the proposed motivation for the collateral homicide related to the principal intimate
relationship (i.e. impending divorce, marital dispute, cheating, jealousy, etc.)?

PP5a YES
PP5b NO
PP5c N/A or not clear

 

Violence Symmetg

(37) VSV1: Was mutual ﬁghting or conﬂict discussed in the article between the perpetrator and
other involved parties (i.e. the principal or collateral victims)? Note: This need not be limited
to the current incident.

O=no
1=yes

Violence Cause

(38) VCV1 a: Story contains causal attribution (if you answer “1 ”, please complete VCV1b)

0=N/A or no causal attribution in story
1=story contains causal attribution

(39) VCV1 b: Causes of Violence
1=violence attributed to individual factors (e.g. relationship conﬂict, anger, mental
health issues, drug abuse, etc.)
2=violence attributed to social factors (e.g. unemployment, ﬁnancial problems)

3=violence attributed to both (e.g. “He was angry about marital troubles and just lost
his job”)

Crime Context Information

(40) CCV1: Story mentions prior harm / abuse against principal IPV victim
O=no story mention
1=story states that there was no prior harm / abuse
2=story mentions prior harm / abuse

(41) CCV2: Story mentions prior harm / abuse against collateral IPV victim
O=no story mention

1=story states that there was no prior harm / abuse
2=story mentions prior harm / abuse

152

(42) CCV3: Was IPV or domestic violence explicitly mentioned in the article?

0=NO
1= YES

(43) CCV5: If yes, please write the relevant quote here:

 

 

 

(44) CCV6: Overall, was the current homicide(s) presented as an isolated crime (i.e., the “frame”
focuses only on the present incident)?

0=NO
1 =YES

153

APPENDIX C

Coder Reliability Results

154

Reliability Results

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Variable Description Percent Scott’s
Label Agreement Pi
D Sex of Reporter 0.933 0.905
1 Victim / Perpetrator Story Position 1.0 1.0
3a Crime Victims: Principal IPV Victim 1.0 1.0
3b Crime Victims: Children of Principal Victim, Perpetrator, 0.977 0.872
or both
3c Crime Victims: Family of IPV Victim 1.0 1.0
3d Crime Victims: Friends of IPV Victim 1.0 1.0
3e Crime Victims: New Boyﬁ‘iend / Girlfriend 1.0 1.0
3f Crime Victims: Neighbors / Bystanders 1.0 1.0
__3_g Crime Victims: Other 1.0 1.0
5 Principal Victim Sex 1.0 1.0
6 Was Principal Victim Killed in Story? 0.933 0.867
7 Total # of Collateral Victims 1.0 1.0
9; 15; 21; 27 Collateral Victim Age 0.963 0.954
10; 16; 22; 28 Collateral Victim Sex 0.975 0.951
11; 17; 23; 29 Collateral Victim’s Relationship to IPV Victim and/or 0.942 0.92
Perpetrator
12; 18; 24; 30 Any Explanation Provided for Collateral IPH? (Yes or No) 0.957 0.907
13a; 19a; 25a; Explanation: Accident 1.0 1.0
3 1 a
13b; 19b; 25b; Explanation: Principal Relationship Conﬂict or Fight 0.917 0.833*
31b
13c; 19c; 25c; Explanation: Intentional or Premeditated 0.967 0.874
310
13d; 19d; 25d; Explanation: Story did not provide Information 0.983 0.967
31d
13e; 19c; 25c; Explanation: Other Explanation 0.983 0.97
31c
32 Perpetrator Sex 1.0 1.0
33 Perpetrator History of Violence 0.933 0841*
PP3a Perpetrator’s Prior Violence: Against Principal IPV Victim 0.967 0.902
PP3b Perpetrator’s Prior Violence: Against Principal IPV 1.0 1.0
Victim’s Children
34c Perpetrator’s Prior Violence: Against Own Children 1.0 1.0
34d Perpetrator’s Prior Violence: Against Other Family 1.0 1.0
Members
34c Perpetrator’s Prior Violence: Against Principal IPV 1.0 1.0
Victim’s New Intimate Partner
34f Perpetrator’s Prior Violence: Against Non-Family 1.0 1.0
Members
34g Perpetrator’s Prior Violence: PPOs Issued against 1.0 1.0
Perpetrator
34h Perpetrator’s Prior Violence: Other 1.0 1.0
353 Perpetrator Portrayal: Homicide Portrayed as Shocking, 0.933 0.814*
Unexpected
35b Perpetrator Portrayal: Homicide Portrayed as Deliberate, 1.0 1.0
Expected
35c Perpetrator Portrayal: Story is not Clear 0.933 0.864
36a Motivation for CIPH Related to Principal Relationship: 0.933 0.864
Yes

 

 

 

 

 

 

155

 

 

 

 

 

 

 

 

 

 

36b Motivation for CIPH Related to Principal Relationship: No 0.933 063*

36c Motivation for CIPH Related to Principal Relationship: 0.933 0.864
Not Clear

37 Violence Symmetry: Mutual Conﬂict or Fighting 0.933 0.814*
Discussed in Article

38 Story Contains Causal Attribution: Yes or No 0.933 0.814*

39 Causes of Violence: Individual Factors, Social Factors, or 0.933 0.852
Both

40 Crime Context: Story Mentions Prior Abuse Against 1.0 1.0
Principal IPV Victim (Yes or No)

41 Crime Context: Story Mentions Prior Abuse Against 1.0 1.0
Collateral IPV Victim

42 Crime Context: DV or IPV Explicitly Mentioned in Article 1.0 N/A
(Yes or No)

44 Crime Context: Overall, Collateral Homicide(s) Presented 1.0 1.0

 

as Isolated Crime

 

 

 

156

 

APPENDIX D

Collateral IPH Qualitative Coding Sheet

157

Collateral IPH Qualitative Coding Sheet

Stag Identification
SIDV1: Story ID number:

SIDV2: Story Date (mmddyy)
SIDV3: Page where story found

1=Page One 2=Section Front Page 3=Inside Page 4=Can’t Tell

Newsgagr Information

NPV1: Newspaper ID number:

Generic Framing Characteristics

Deﬁnition of Problem:

Cause of Crime (i.e., diagnosis):

Moral Judgments regarding the collateral IPH:

Any remedies suggested:

158

Article Themes / Points of Interest:

Family Violence Sentences:

Coercive Control Sentences:

159

Final Assessment/ Interpretive Package:
Does this article align more with family violence or coercive control theory?

FAMILY VIOLENCE

COERCIVE CONTROL

Notes:

160

APPENDIX E

Example Qualitative Analysis Matrices

161

Example Qualitative Analysis Matrices

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Article ID Problem Cause Moral / Remedies Emerging
Themes
1-1
1-2
Article ID Family Violence Sentences Coercive Control
Sentences
1-1
1-2
Article ID FV CC Unclear Conclusions
1-1 X
1-2 X

 

 

 

 

 

162

 

 

 

BIBLIOGRAPHY

163

BIBLIOGRAPHY

Adams, B. N. (1965). Coercion and consensus theories: Some unresolved issues.
American Journal of Sociology, 71, 714-716.

Adams, D. (1988). Counseling men who batter: A profeminist analysis of ﬁve treatment
models. In M. Bograd & K. Yllo (Eds), Feminist perspectives on wife abuse (pp.
176-199). Beverly Hills: Sage.

Agadjanian, V., & Menjivar, C. (2008). Talking about the "epidemic of the millennium":
Religion, informal communication, and HIV/AIDS in Sub-Saharan Africa. Social
Problems, 55(3), 301-322.

Ahmad, F., Hogg-Johnson, S., Stewart, D. E., & Levinson, W. (2007). Violence
involving intimate partners - Prevalence in Canadian family practice. Canadian
Family Physician, 53, 461-486.

Allen, B., O'Loughlin, P., J asperson, A., & Sullivan, J. L. ( 1994). The media and the gulf
war: Framing, priming, and the spiral of silence. Polity, 27(2), 255-284.

Allen, C.M., & Straus, M.A. (1980). Resources, power, and husband-wife violence. In
M.A. Straus and GT. Hotaling (Eds). The social causes of husband-wife violence
(pp. 188-208). Minneapolis: University of Minnesota Press.

Altheide, D. L. (1987). Ethnographic Content Analysis. Qualitative Sociology, 10(1), 65-
77.

Altheide, D. L. (1996). Qualitative Media Analysis (Vol. 38). Thousand Oaks: Sage.

Anderson, KL. (1997). Gender, status, and domestic violence: An integration of feminist
and family violence approaches. Journal of Marriage and the Family, 59(3),
655-669.

Anderson, K. L. (2005). Theorizing gender in intimate partner violence research. Sex
Roles, 52, 853-865.

Archer, J. (2000). Sex differences in aggression between heterosexual partners: A meta-
analytic review. Psychological Bulletin, 126, 651-680.

Arizona Coalition Against Domestic Violence (2003). Arizona media handbook.
Accessed 27 June ﬁom http://www.azcadv.org/docs/MediaHandbook.pd£

Associated Press (2008). Body of 1 of 4 kids thrown from bridge found.
http://www.foxnews.com/story/0,2933,322366,00.htm1. Accessed 26 February.

164

Bachman, R., & Saltzman, L. E. (1995). Estimates from the redesigned survey (No. NCJ
154348): US. Department of Justice.

Bailey, J. E., Kellermann, A. L., Somes, G. W., Banton, J. G., Rivara, F. P., & Rushforth,
N. P. (1997). Risk factors for violent death of women in the home. Archives of
Internal Medicine, 15 7(7), 777-782.

Bancroft, L. (2002). Why does he do that? Inside the minds of angry and controlling
men. New York: Putnam's Sons.

Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology.
Cambridge: Cambridge University Press.

Bassis, M. S., Gelles, R. J ., & Levine, A. (1982). Social Problems. New York: Harcourt,
Brace J ovanovich.

Bateson, G. (1972). Steps to an ecology of mind: Collected essays in anthropology,
psychiatry, evolution, and epistemology. New York: Ballantine.

Becker, HS (1963). Outsiders: Studies in the sociology of deviance. New York: The
Free Press of Glencoe.

Benson, R., & Saguy, A. C. (2005). Constructing social problems in an age of
globalization: A French-American comparison. American Sociological Review,
70, 233-259.

Berelson, B. (1966). Content analysis in communication research. In B. Berelson & M.
J anowitz (Eds.), Reader in public opinion and communication (pp. 260-266). New
York: Free Press.

Bems, N. (1999). "My problem and how I solved it": Domestic violence in women's
magazines. The Sociological Quarterly, 40, 85-108.

Bems, N. (2004). Framing the victim: Domestic violence media and social problems.
New York: Walter de Gruyter, Inc.

Best, J. (Ed.). (1995). Images of Issues: T ypifving contemporary social problems. New
York: Walter de Gruyter, Inc.

Biroscak, B.J., & Smith P.K. (2005). Intimate Partner Homicide in Michigan,
1999—2003 . Lansing, MI: Michigan Department of Community Health.

Biroscak, B.J., Smith, P.K., & Post, LA. (2006). A practical approach to public health
surveillance of violent deaths related to intimate partner relationships. Public
Health Reports, 121, 393-399.

165

Black, M. C., & Breiding, M. J. (2008). Adverse health conditions and health risk
behaviors associated with intimate partner violence - United States, 2005.
Morbidity and Mortality Weekly Report, 5 7(05), 1 13-1 17.

Block, C. R., & Christakos, A. (1995). Intimate partner homicide in Chicago over 29
years. Crime & Delinquency, 41(4), 496-526.

Blumer, H. (1971). Social problems as collective behavior. Social Problems, 18, 298-
306.

Bogat, G. A., Levendosky, A.A., Theran, 8., von Eye, A., and Davidson, W.S. (2003).
Predicting the psychosocial effects of interpersonal partner violence (IPV): How
much does a woman's history of IPV matter? Journal of Interpersonal Violence,
18(11), 1271-1291.

Bogat, G.A., Leahy, K., von Eye, A., Maxwell, C., Levendosky, AA, and Davidson 11,
W.S. (2005). The inﬂuence of community violence on the functioning of women

experiencing domestic violence. American Journal of Community Psychology,
36(1-2), 123-132.

Bograd, M. (1988). Feminist perspectives on wife abuse: An introduction. In K. Yllo &
M. Bograd (Eds.). Feminist perspectives on wife abuse (pp. 11-25). Newbury
Park: Sage.

Bonami, A. E., Thompson, R.S., Anderson, M., Reid, R.J., Carrell, D., Dimer, J .A., and
Rivara, F.P. (2006). Intimate partner violence and women's physical, mental, and
social functioning. American Journal of Preventive Medicine, 30(6), 458-466.

Bowman, C. G., & Altman, B. (2003). Wife murder in Chicago: 1910-1930. The Journal
of Criminal Law & Criminology, 92(3-4), 739-790.

Brady, D., & Kall, D. (2008). Nearly universal, but somewhat distinct: The feminization
of poverty in afﬂuent Western democracies, 1969-2000. Social Science Research,
3 7(3), 976-1007.

Brownmiller, S. (1978). Against our will: Men, women, and rape. New York: Bantam
Books.

Campbell, J .C., Glass, N., Sharps, P.W., Laughon, K., & Bloom, T. (2007). Intimate
partner homicide - Review and implications of research and policy. Trauma

Violence & Abuse, 8(3), 246-269.

Campbell, J. C., and Soeken. (1999). Women's response to battering over time: An
analysis of change. Journal of Interpersonal Violence, 14, 21-40.

166

Campbell, J. C., Webster, D., Koziol-McLain, J ., Block, C., Campbell, D., & M.A. Curry,
e. a. (2003). Risk factors for femicide in abusive relationships: Results from a
multisite case control study. American Journal of Public Health, 93(7), 1089-
1097.

Cascardi, M., & O'Leary, K. D. (1992). Depressive symptomology, self-esteem, self-
blarne in battered women. Journal of Family Violence, 7(4), 249-259.

Cohen, D., Llorente, M., & Eisdorfer, C. (1998). Homicide-suicide in older persons.
American Journal of Psychiatry, 155(3), 390-396.

Consalvo, M. (1998). "3 shot dead in courthouse": Examining news coverage of domestic
violence and mail-order brides. Women '3 studies in communication, 21(188-211).

Coser, L. A. (1956). The function of social conﬂict. New York: The Free Press.

D'Angelo, P. (2002). News ﬁraming as a multiparadigmatic research programme: A
response to Entrnan. Journal of Communication, 52(4), 870-888.

Dahrendorf, R. (1959). Class and class conﬂict in industrial society. London: Routledge
and Kegan Paul.

Das Dasgupta, S. (2002). A framework for understanding women's use of nonlethal
violence in intimate heterosexual relationships. Violence Against Women, 8(11),
1364-1389.

Demaris, A., and Kaukinen, C. (2005). Violent victimization and women's mental and
physical health: Evidence from a national sample. Journal of Research in Crime
and Delinquency, 42(4), 384-411.

DeVoe, E. R., & Smith, E. L. (2002). The impact of domestic violence on urban
preschool children: Battered mothers' perspectives. Journal of Interpersonal
Violence, 1 7, 1075-1101.

Dobash, R.E., & Dobash, RP (1979). Violence against wives: A case against the
patriarchy. New York: Free Press.

Dobash, R.P., Dobash, R.E., Wilson, M., and Daly, M. (1992). The myth of sexual
symmetry in marital violence. Social Problems, 39, 71-91.

Dutton, D.G. (1994). Patriarchy and wife assault: The ecological fallacy. Violence and
Victims, 9, 167-182.

Dutton, D.G. (2006). Rethinking domestic violence. Vancouver: UBC Press.

167

Dutton, D.G. (2007). The abusive personality: Violence and control in intimate
relationships. New York: Guilford Press.

Edelson, J. (1999). Children's witnessing of adult domestic violence. Journal of
Interpersonal Violence, 14, 839-870.

Entrnan, R. M. (1991). Framing US coverage of the international news: Contrasts in
narratives of the KAL and Iran air incidents. Journal of Communication, 41(4), 6-
27.

Entrnan, R. M. (1993). Framing: Towards clariﬁcation of a fractured paradigm. Journal
of Communication, 43(4), 51-58.

Farley, J. E. (1987). American Social Problems. Englewood Cliffs: Prentice-Hall.

F elson, RB. (2002). Violence and gender reexamined. Washington, DC: American
Psychological Association

Felson, R.B., & Messner, SE. (2000). The control motive in intimate partner violence.
Social Psychology Quarterly, 63(1), 86-94.

F errand-Bullock, C., & Cubert, J. (2002). Coverage of domestic violence fatalities by
newspapers in Washington state. Journal of Interpersonal Violence, 1 7(5), 475-
499.

Ferrand,-Bullock, C. (2008). Ofﬁcial sources dominate domestic violence reporting.
Newspaper Research Journal, 29(2), 6-23.

F raniuk, R., Seefelt, J .L., & Vandello, J .A. (2008). Prevalence of rape myths in headlines
and their effects on attitudes toward rape. Sex Roles, 58(11-12), 790-801.

Friedland, L. A., & Zhong, M. (1996). International television coverage of Beijing Spring
1989: A comparative approach. Journalism & Mass Communication Monographs,
1 5 6.

Gamson, W. (1989). News as framing. American Behavioral Scientist, 33(2), 157—161.

Gamson, W. A. (1992). Talking politics. New York: Cambridge University Press.

Gamson, W. A., & Modigliani, A. (1987). The changing culture of afﬁrmative action. In
R. G. Braungart & M. M. Braungart (Eds), Research in political sociology (V 01.
3, pp. 137-177). Greenwich: JAI Press.

Gamson, W. A., & Mogdiliani, A. (1989). Media discourse and public opinion on nuclear
power: A constructionist approach. American Journal of Sociology, 95, 1-3 7.

168

Ganley, A., & Schechter, S. (1996). Domestic violence: A national curriculum for
children 's protective services. San Francisco: Family Violence Prevention Fund.

Gans, H. (1979). Deciding what’s news. New York: Patheon.

Garofalo, J. (1981). Crime and the mass media: A selective review of research. Journal
of Research in Crime and Delinquency, 18, 319-350.

Gauthier, D. K., & Brankston, W. B. (2004). "Who kills whom" revisited: A sociological
study of variation in the sex ratio of sex killings. Homicide Studies, 8(2), 96-122.

Garcia, L., Soria, C., & Hurwitz, EL. (2007). Homicides and intimate partner violence:
A literature review. Trauma, Violence, & Abuse, 8(4), 370-383.

Gelles, R.J. (1985). Family violence. Annual Review of Sociology, 11, 347-367.

Gelles, R.J. (1993). Through a sociological lens: Social structure and family violence. In
R.J. Gelles & D.R. Loseke (Eds.). Current controversies on family violence (pp.
31-46). Newbury Park: Sage.

Gelles, R.J., & Straus, M.A. (1988). Intimate violence: The causes and consequences of
abuse in the American family. New York: Simon & Schuster.

Gitlin, T. (1980). The whole world is watching: Mass media in the making & unmaking of
the new left. Berkeley: University of California Press.

Glaser, B., & Strauss, A. (1967). The discovery of grounded theory. Chicago: Aldine.

Gofﬁnan, E. (1974). Frame Analysis: An Essay on the Organization of Experience. New
York: Harper and Row.

Graham-Hermann, S. (2001). Designing intervention evaluations for children exposed to
domestic violence. In S. Graham-Bermann (Ed.), Domestic violence in the lives of
children: The future of research, intervention, and social policy (pp. 23 7-267).
Washington, DC: American Psychological Association.

Gusﬁeld, J .R. (1981). The culture of public problems. Chicago: The University of
Chicago Press.

Hallahan, K. (1999). Seven models of framing: Implications for public relations. Journal
of Public Relations Research, I I , 205-242.

Hartman, P., & Husband, C. (1974). Racism and Mass Media. London: Davis Poynter.

Horsti, K. (2003). Global mobility and the media: Presenting asylum seekers as a threat.
Nordicom Review, 24(1), 41-54.

169

Houskamp, B. M., & Foy, D. W. (1991). The assessment of posttraumatic stress disorder
in battered women. Journal of Interpersonal Violence, 6, 367-3 75.

Homing, C.A., McCullough, B.C., & Sugimoto, T. (1981). Status relations in marriage:
Risk factors in spouse abuse. Journal of Marriage and the Family, 43, 675-692.

Huth-Bocks, A. C., Levendosky, AA, and Bogat, GA. (2002). The effects of domestic
violence during pregnancy on maternal and infant health. Violence and Victims, 8,
1-1 7.

Iyengar, S. (1991). Is Anyone Responsible? How television frames political issues.
Chicago: University of Chicago Press.

J argowsky, P.A., & Yang, R. (2006). The "underclass" revisited: A social problem in
decline. Journal of Urban Aﬂairs, 28(1), 55-70.

Johnson, MP. (1995). Patriarchal terrorism and common couple violence: Two forms of
violence against women. Journal of Marriage and the Family, 5 7(2), 283-294.

Johnson, MP, & Leone, J .M. (2005). The differential effects of intimate terrorism and
situational couple violence. Journal of Family Issues, 26(3), 322-349.

Johnson, S.P., & Sullivan, CM. (2008). How child protection workers support or further
victimize battered mothers. Aﬂilia, 23(3), 242-258.

Jones, A. (2000). Next Time She 'll be Dead: Battering & How to Stop it. Boston: Beacon
Press.

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under
risk. Econometrica, 47, 263-291.

Kellermann, A., & Heron, S. (1999). Firearms and family violence. Emergency medicine
clinics of North America, 1 7(3), 699-716.

Kessler, R. C., Molnar, B.E., Feurer, I.D., and Appelbaum, M. (2001). Patterns and
mental health predictors of domestic violence in the United States: Results ﬁom
the National Comorbidity Survey. International Journal of Law and Psychiatry,
24, 487-508.

Kimmel, MS. (2002). "Gender symmetry" in domestic violence: A substantive and
methodological research review. Violence Against Women, 8, 1332-1363.

Kinder, D. R., & Sanders, L. M. (1990). Mimicking political debate with survey
questions: The case of white opinion on afﬁrmative action for blacks. Social
Cognition, 8, 73-103.

170

Koch, T. (1990). The news as myth: Fact and context in journalism. Westport:
Greenwood Press.

Kramer, A., Lorenzon, D., and Mueller, G. (2004). Prevalence of intimate partner
violence and health implications for women using emergency departments and
primary care clinics. Women 's Health Issues, 14(1), 19-29.

Martin, D. (1976). Battered wives. New York: Pocket Books.

Maxwell, K. A., Huxford, J ., Borum, C., & Hornik, R. (2000). Covering domestic
violence: How the OJ. Simpson case shaped reporting of domestic violence in the
news media. Journalism & Mass Communication Quarterly, 77, 258-272.

McCombs, M., & Reynolds, A. (2002). News inﬂuence on our pictures of the world. In
J. Bryant and D. Zillmann (Eds), Media eﬂects: Advances in theory and research.
Mahwah: Lawrence Erlbaurn Associates.

McCoy, Kelley Campos (2007). Framing race: J oumalists and news coverage of a fatal
shooting in Seattle. PhD Dissertation, University of Washington, United States -
Washington. Retrieved February 24, 2008, from ProQuest Digital Dissertations
Database. (Publication No. AAT 0818456.

McLeod, D. M., & Detenber, B. H. (1999). Framing effects of television news coverage
of social protest. Journal of Communication, 49(3), 3-23.

McManus, J ., & Dorfrnan, L. (2003). Distracted by drama: How California newspapers
portray intimate partner violence. Issue 13, January. Accessed 2/13/08 from:
http://www.bmsg.org/pub-issues.php

McNair, B. (1988). Images of the Enemy. London: Routledge.

McQuail, D. (2005). McQuail's Mass Communication Theory. London: Sage.

Meyers, M. (1994). News of battering. Journal of Communication, 44(2), 47-63.

Michigan Coalition Against Domestic and Sexual Violence (2008). Working with the
media: A toolkit for service providers. Accessed 27 June 2008 at

http://www.mcadsv.org/products/media.html

Misra, J ., Moller, S., & Karides, M. (2003). Envisioning dependency: Changing media
depictions of welfare in the 20th century. Social Problems, 50(4), 482-504.

National Coalition Against Domestic Violence (2008). Accessed 27 July 2008 at
http://www.ncadv.org/resources/Publications_1 23 .html

171

 

Nelson, T. E., & Oxley, Z. M. (1999). Issue framing effects on belief importance and
opinion. The Journal of Politics, 61(4), 1040-1067.

Norusis, M.J. (2002). SPSS 11.0 guide to data analysis. Upper Saddle River: Prentice
Hall.

Neuman, W. R., Just, M. R., & Crigler, A. A. (1992). Common knowledge. News and the
construction of political meaning. Chicago: The University of Chicago Press.

O'Brien, J .E. (1975). Violence in divorce-prone families. In S.K. Steinmetz & M.A.
Straus (Eds.). Violence in the family (pp. 66-74). New York: Dodd, Mead.

Pence, E., & Paymar, M. (1993). Education groups for men who batter. The Duluth
Model. New York: Springer.

Pokhariyal, GP. (2005). Models for understanding social problems in slums.
International Journal on World Peace, 22(2), 59-76.

Post L.A., Klevens J ., & Maxwell C., An Examination of Whether Coordinated
Community Responses Impact Intimate Partner Violence. Journal of
Interpersonal Violence. (In Press).

Price, V., Tewksbury, D., & Powers, E. (1997). Switching trains of thought: The impact
of news ﬂames on readers' cognitive responses. Communication Research, 24,
481-506.

Ramazanoglu, C., & Holland, J. (2005). Feminist Methodology: Challenges and Choices.
London: Sage Publications.

Reese, S. D. (2001). Introduction. In S. D. Reese, O. H. Gandy & A. E. Grant (Eds),
Framing public life: perspectives on media and our understanding of the social
world (pp. 1-31). Mahwah: Erlbaum.

Rennison, C. M., & Welchans, S. (2000). Intimate partner violence. Crime Data Brief
(No. NCJ 197838): Bureau of Justice Statistics.

Ridout, T.N., Grosse, A.C., & Appleton, AM. (2008). News media use and Americans'
perceptions of global threat. British Journal of Political Science, 38, 575-593.

Rothman, A. J ., & Salovey, P. (1997). Shaping perceptions to motivate healthy behavior:
The role of message framing. Psychological Bulletin, 121(1), 3-19.

Ryan, C., Anastario, M., DaCunha, A. (2006) Changing coverage of domestic violence

murders: A longitudinal experiment in participatory communication. Journal of
Interpersonal Violence, 21(2), 209—228.

172

Sacco, V.F. (1995). Media constructions of crime. Annals of the American Academy of
Political and Social Science, 539, 141-154.

Scanzoni, J. (1972). Marital Conﬂict as a positive force. In J. Scanzoni (Ed.), Sexual
Bargaining (pp. 61-102). New Jersey: Prentice-Hall.

Scheufele, D. A. (1999). Framing as a theory of media effects. Journal of
Communication, 49(1), 103—122.

Scheufele, D. A. (2000). Agenda-setting, priming, and ﬂaming revisited: Another look at
cognitive effects of political communication. Mass Communication & Society,
3(2&3), 297-316.

Schiff, F. and Reiter, G. (2004, Aug) "Measuring Intercoder Reliability: Scott's Pi
Reconsidered" Paper presented at the annual meeting of the American
Sociological Association, Hilton San Francisco & Renaissance Parc 55 Hotel, San
Francisco, CA, Online <.PDF> Retrieved 2008-07-19 ﬂom
http://www.allacadernic.com/meta/pl 10405_index.html

Schon, D. A., & Rein, M. (1994). Frame reﬂection: Towards the resolution of intractable
policy contraversies. New York: Basic Books.

Schneider, J .W. (1985). Social problems theory: The constructionist view. Annual
Review of Sociology, 11, 209-229.

Schwartz, H., & Jacobs, J. (1979). Qualitative Sociology. New York: Free Press.

Shoemaker, P. J ., & Reese, S. D. (1996). Mediating the message: Theories of inﬂuences
on mass media content. New York: Longman Publishers.

Simmel, G. (1955). Conﬂict and the web of group affiliation. Glencoe: The Free Press.

Snow, D., & Benford, R. (1988). Ideology, frame resonance, and participant
mobilization. International Social Movement Research, 1, 197-217.

Spector, M., & Kitsuse, J .I. (1977). Constructing social problems. Menlo Park:
Cummings.

Spector, M., & Kitsuse, J .I. (2000). Constructing social problems. Edison: Transaction
Books.

Sprey, J. (1969). The family as a system in conﬂict. Journal of Marriage and the Family,
31 , 699-706.

Stark, E., & Flitcraft, AH. (1991). Spouse abuse. In M.L. Rosenberg & M.A. Finley

(Eds.). Violence in America: A public health approach (pp. 123-157). New
York: Oxford University Press.

173

Stark, E., & Flitcraft, AH. (1996). Women at risk: Domestic violence and women's
health. Thousand Oaks: Sage.

Stark, E. (2007). Coercive Control: The entrapment of women in personal life. New
York: Oxford University Press.

Starosta, W. J. (1984). Qualitative content analysis: A Burkean perspective. In W. B.
Gudykunst & Y. Y. Kim (Eds), Methods for intercultural communication (pp.
185-194). Beverly Hills: Sage.

Stover, C. S. (2005). Domestic violence research - What have we learned and where do
we go ﬂom here? Journal of Interpersonal Violence, 20(4), 448-454.

Straus, M. (1997). Physical assaults by women partners: A major social problem. In M.
R. Walsh (Ed), Gender: Ongoing debates. New Haven: Yale University Press.

Straus, M. (1999). The controversy over domestic violence by women: A
methodological, theoretical and sociology of science analysis. In X. B. Arriaga &
S. Oskamp (Eds), Violence in intimate relationships. Thousand Oaks: Sage.

Straus, M., & Smith, C. (1990). Violence in Hispanic families in the United States:
Incidence rates and structural interpretations. In M. A. Straus, R. J. Gelles & C.
Smith (Eds), Physical violence in American families: Risk factors and
adaptations to violence in 8,1 45 families (pp. 341-367). New Brunswick:
Transaction.

Straus, M. A. (1973). A general systems theory approach to a theory of violence between
family members. Social Science Information, 12(3), 105-125.

Straus, M. A. (1979). Measuring intrafamily conﬂict and violence: The conﬂict tactics
(CT) scales. Journal of Marriage and the Family, 41(1), 75-88.

Straus, M.A. (1987). The conﬂict tactics scales and its critics: An evaluation and new
data on validity and reliability. Technical Report.

Straus, M. A. (1992). Sociological research and social policy: The case of family
violence. Sociological Forum, 7(2), 211-237.

Straus, M. A. (1993). Physical assaults by wives: A major social problem. In R. J. Gelles
& D. R. Loseke (Eds), Current controversies on family violence. Newbury Park:
Sage.

Straus, M. A. (2006). Future research on gender symmetry in physical assaults on
partners. Violence Against Women, 12(1 1), 1086-1097.

174

 

Straus, M. A., & Gelles, R. J. (1986). Societal change and change in family violence ﬂom
1975 to 1985 as revealed by two national surveys. Journal of Marriage and the
Family, 48(3), 465-479.

Straus, M.A., Gelles, R.J., & Steinmetz, SK. (1980). Behind closed doors: Violence in
the American Family. Garden City: Anchor.

Straus, M. A., Gelles, R. J ., & Steinmetz, S. K. (2007). Behind closed doors: Violence in
the American family. New Brunswick: Transaction Publishers.

Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarrnan, D. B. (1996). The revised
conﬂict tactics scales (CTS2): Development and preliminary psychometric data.
Journal of Family Issues, I 7, 283-316.

Tannen, D. (1979). What's in a ﬂame? Surface evidence for underlying expectations. In
R. O. Freedle (Ed.), New directions in discourse processing (pp. 137-181).
Norwood: Ablex.

Taylor, C. A., & Sorenson, S. B. (2002). The nature of newspaper coverage of homicide.
Injury Prevention, 8, 121-127.

Taylor, S.J., & Bogdan, R. (1998). Introduction to qualitative research methods: A
guidebook and resource. John Wiley & Sons.

Terkildsen, N., & Schnell, F. (1997). How media frames move public opinion: An
analysis of the women's movement. Political Research Quarterly, 5 0(4), 879-900.

Thompson, R. S., Bonami, A.E., Anderson, M., Reid, R.J., Dimer, J .A., Carrell, D., and
Rivara, F.P. (2006). Intimate partner violence: Prevalence, types, and chronicity
in adult women. American Journal of Preventive Medicine, 3 0(6), 447-457.

Tjaden, P., and Thoennes, N. (2000a). Full report of the prevalence, incidence, and
consequences of violence against women: Research report. Washington, DC:
National Institute of Justice.

Tjaden, P., & Thoennes, N. (2000b). Prevalence and consequences of male-to-female
and female-to-male intimate partner violence as measured by the National
Violence Against Women Survey. Violence Against Women, 6, 142-161.

Tuchman, G. (1978). Making news: A study in the construction of reality. New York:
Free Press.

Van. Wormer, K., & Bednar, SC. (2002). Working with male batterers: A restorative-
strengths perspective. Families in Society, 83(5-6), 557-566.

175

 

Vitanza, S., Vogel, L. C. M., & Marshall, L. L. (1995). Distress and symptoms of
posttraumatic stress disorder in abused women. Violence and Victims, 10(1), 23-
34.

Washington State Coalition Against Domestic Violence. (2002, revised 2006) Covering
domestic violence: A guide for journalists and other media professionals. Accessed
10 February 2008 from: http://www.wscadv.org

White, J. (2002). Understanding rape: A meta-theoretical ﬂamework. In C.B. Travis
(Ed.), Evolution, violence, and gender. Boston: MIT Publishers.

Wozniak, Jessica (2007). Fact or ﬁction? Gender issues related to newspaper reports of
intimate partner homicide. Psy.D. dissertation, University of Hartford, United
States -- Connecticut. Retrieved February 6, 2008, ﬂom ProQuest Digital
Dissertations database. (Publication No. AAT 3268725).

Yllo, K. (1993). Through a feminist lens: Gender, power, and violence. In R.J. Gelles &
D.R. Loeske (Eds.). Current controversies on family violence (pp. 47-64).
Newbury Park: Sage.

Yllo, K., & Bograd, M. (Eds) (1988). Feminist perspectives on wife abuse. Newbury
Park: Sage.

Yodanis, CL. (2004). Gender inequality, violence against women, and fear. Journal of
Interpersonal Violence, 19(6), 655-675.

Zillman, D. (2002). Exempliﬁcation theory of media inﬂuence. In J. Bryant and D.
Zillmann (Eds), Media eﬁ’ects: Advances in theory and research. Mahwah:

Lawrence Erlbaurn Associates.

Zwick, R. (1986). Another look at inter-rater agreement: Research report. Educational
Testing Services.

176

      

MCHIGAN STATE U RSITY LIBRARIES
I

lllll/lllllllllllIllIlllzllllllllllllllIll/Ill!

3 1293 O 956 7975