Media representations of violence against women and their children:
State of knowledge

Landscapes

State of knowledge | November 2015
Issue 15 | 2015

ANROWS/Our Watch acknowledgement

This material was produced with funding from the Australian Government and the Australian state and territory governments. Australia’s National Research Organisation for Women’s Safety (ANROWS) gratefully acknowledges the financial and other support it has received from these governments, without which this work would not have been possible. The findings and views reported in this paper are those of the authors and cannot be attributed to the Australian Government, or any Australian state or territory government.

Acknowledgement of Country

ANROWS and Our Watch acknowledges the traditional owners of the land across Australia on which we work and live. We pay our respects to Aboriginal and Torres Strait Islander elders past, present and future; and we value Aboriginal and Torres Strait Islander history, culture and knowledge.

© ANROWS & Our Watch 2015

Published by

Australia’s National Research Organisation for Women’s Safety Limited (ANROWS)
PO Box 6322, Alexandria NSW 2015 | www.anrows.org.au | Phone +61 2 8374 4000
ABN 67 162 349 171

Our Watch
GPO Box 24229, Melbourne VIC 3000 | www.ourwatch.org.au | Phone +61 3 8692 9500
ABN 60 164 123 844

Media representations of violence against women and their children: State of knowledge paper / Georgina Sutherland, et al.
Sydney : ANROWS, c2015.
Pages ; 30 cm. (Landscapes : State of knowledge: 15/2015)
I. Violence against women. II. Media coverage. III. Women – Violence against – Press coverage.
I. Sutherland, Georgina. II. McCormack, Angus. III. Pirkis, Jane. IV. Easteal, Patricia. V. Holland, Kate. VI. Vaughan, Cathy.

ISSN: 2204-9657 (print) 2204-9665 (online)

ISBN: 978-1-925372-19-9 (print) 978-1-925372-18-2 (online)

Creative Commons Licence

Attribution-Non Commercial

CC BY-NC

This licence lets others distribute, remix and build upon the work, but only if it is for non-commercial purposes and they credit the original creator/s (and any other nominated parties). They do not have to license their Derivative Works on the same terms.

Version 3.0 (CC Australia ported licence): View CC BY-NC Australia Licence Deed | View CC BY-NC 3.0 Australia Legal Code

Version 4.0 (international licence): View CC BY-NC 4.0 Licence Deed | View CC BY-NC 4.0 Legal Code

University of Melbourne
University of Melbourne, Parkville VIC 3010

Media representations of violence against women and their children: State of knowledge paper

Prepared by

Dr Georgina Sutherland, Centre for Mental Health, Melbourne School of Population and Global Health, University of Melbourne

Mr Angus McCormack, Centre for Mental Health, Melbourne School of Population and Global Health, University of Melbournes

Professor Jane Pirkis, Centre for Mental Health, Melbourne School of Population and Global Health, University of Melbourne

Professor Patricia Easteal, Faculty of Business, Government and Law, University of Canberra

Dr Kate Holland, Faulty of Arts and Design, University of Canberra

Dr Cathy Vaughan, Centre for Health Equity, Melbourne School of Population and Global Health, University of Melbourne

Author’s acknowledgement

The authors wish to acknowledge the financial and other support provided by ANROWS and Our Watch. We extend our thanks to Dr Trishima Mitra-Kahn and the project advisory group for their invaluable advice and support.

This work is part of the ANROWS Landscapes series. ANROWS Landscapes (State of knowledge papers) are medium length papers that scope current knowledge on an issue related to violence against women and their children. Papers will draw on empirical research, including research produced under ANROWS’s research program, and/or practice knowledge.

This paper addresses work covered in the ANROWS/Our Watch research project 2.2 “Media representations of violence against women and their children”, part of the ANROWS Research Program 2014-16 (Part 1). Please consult the ANROWS website for more information on this project, and the research program more broadly.

The research is jointly funded by Our Watch as part of its National Media Engagement Project. The National Media Engagement (NME) Project is engaging media to increase quality reporting of violence against women and their children and building awareness of the impacts of gender stereotyping and inequality.

Violence against women and their children generates daily media coverage in Australia. What is missing is an understanding of the links between sexism, gender inequality, community attitudes and this violence. In fact some reporting perpetuates attitudes and myths that give rise to the violence in the first place. The NME Project has been designed to address these issues.

Contents

Executive summary

Background

Aim and framework

Method

Results

Conclusions

Background

Definitions and forms of violence against women

Prevalence and impact of violence against women

Primary prevention

The role of media

Aim and framework

Aim

Framework

Method

Inclusion and exclusion criteria

Database searches

Selection process

Quality appraisal

Results: Studies of media representation

Overview

Outcomes

Summary

Results: Studies of audience reception

Overview

Outcomes

Summary

Results: Studies of news production

Overview

Outcomes

Media guidelines

Summary

Discussion and conclusion

Summary of key findings

Implications

Limitations of the review

Conclusion

References

Appendix A

Primary studies of media representations of violence against women

Appendix B

Unpublished reports of media representations of violence against women

Appendix C

Primary studies of audience reception related to violence against women

Appendix D

Unpublished reports of audience reception related to violence against women

Appendix E

Primary studies of media production related to violence against women

Appendix F

International media guidelines

Executive summary

Background

Violence against women is one of Australia’s most significant public health problems. It refers to acts of gender-based violence that cause or have the potential to cause physical, sexual, emotional, mental and/or economic harm. It is one of the most widespread violations of human rights. Children and young people are harmed by being exposed to violence perpetrated against women.

Evidence-based primary prevention of violence against women is a new and emerging field but there are promising strategies for prevention. This evidence highlights the importance of challenging cultural and social norms that condone, tolerate or excuse violence against women. It is within this context that the news and information media – a dominant force in shaping the discourse on matters of public importance – is seen to play a crucial role.

Aim and framework

The aim of this state of knowledge paper is to provide an overview of the best available contemporary evidence on the way news and information media portray violence against women. We aimed to include a broad sample of literature including peer-reviewed scientific papers, reports and other relevant sources of information, including guidelines for journalists.

As a framework for the paper, we grouped studies into three broad areas of inquiry:

• Studies of representation: understanding the nature of the content and discourse in news items on violence against women.
• Studies of audience reception: understanding how audiences interpret news on violence against women and how risk is perceived and managed.
• Studies of news production: understanding the practices of gathering and reporting on violence against women and their children.

Method

We used the following criteria to identify literature for inclusion in this paper:

• definitions and behaviours most relevant to the Australian context: domestic, family and intimate partner violence and sexual violence/assault;
• studies that explored media representation, reception, production and violence against women, as a key theme;
• published in the English language; and
• published on or after January 1, 2000.

Databases that were systematically searched included Scopus, ProQuest and PubMed. Google was used to search the grey literature for non-peer reviewed reports and media guidelines. To assess the quality of included studies, we used tools developed by the UK’s national Critical Appraisal Skills Programme (CASP) collaboration, which primarily assesses key elements of the quality of scientific inquiry according to its rigour, credibility and relevance.

Results

In the area of media representation, we identified 48 primary studies; over half of which were published in the last five years (2010-15). The majority of studies were conducted in the United States (US); six studies were from Australia. Newspapers were the most common type of media analysed. Of the 44 peer-reviewed primary studies, 34 were assessed as being high quality, seven were moderate quality and three were of a low quality.

Findings from the studies illustrated a number of key themes in the way news and information media portray violence against women including:

• not reporting the social context in which male perpetrated violence against women occurs;
• sensationalising stories through language or by disproportionately focusing on stories that fit key news values;
• perpetuating myths and misrepresentations;
• directly and indirectly shifting blame from male perpetrators of violence and assigning responsibility for violence to women; and
• relying on law enforcement as the expert “voices”.

In the area of audience reception, we identified ten primary studies; seven were from the US; none were from Australia.

The studies broadly covered audience effects in terms of emotional responses and attributions of responsibility. Only one study examined the relationship between televised media items on intimate partner violence and incidence. Of the nine peer-reviewed primary studies, three studies were assessed as being high quality, five as moderate and one as low quality.

Despite there being a relatively small number of primary studies, findings illustrated some common themes regarding the way audiences take in and interpret media portrayals of violence against women including how:

• audience emotional responses and attributions of responsibility are affected, and can be manipulated, by how media frames news stories;
• “rape myths” affect attitudes about perpetrators of sexual violence;
• images and language moderate audience responses; and
• media effects intersect with audience gender and prior knowledge and understandings of violence against women.

In the area of news production, we identified only four primary studies, all conducted by US researchers. These broadly covered two main areas of inquiry: the practice of news-making from the perspective of journalists, editors and producers, and the way media guidelines impact on reporting practices for media professionals. The two studies evaluating media guidelines were assessed as being high quality and the two studies exploring journalist practices were assessed as being of moderate quality.

We identified 23 media guidelines that generally reflect national and international activity in the area. The content of these guidelines were similar with a number of re-occurring themes including recommendations to:

• report the social context in which male perpetrated violence against women occurs;
• use correct language and terminology;
• avoid blaming the victim;
• avoid offering excuses for men’s violence;
• consider how source selection shapes the story; and
• provide women with information on where to seek help.

Conclusions

While many factors are known to influence public perceptions about violence against women, in this state of knowledge paper we examined the role of news and information media. We found that media representation studies dominate research in the area. Collectively, these studies illustrate that the media frequently mirrors society’s confusion and ambivalence about violence against women. Although the link between media reporting and behaviour is not well-established, studies of audience reception show that media can play a role in dispelling myths and reinforcing information about the true nature and extent of the problem. Unfortunately, reporting on violence against women that challenges rather than reinforces cultural and social norms about gender is still the minority. There remain gaps in our knowledge with more research needed as to the impact of news coverage on public understandings, attitudes and behaviours and how to best craft mediated messages to more effectively promote positive social change.

Background

Violence against women and their children is a global public health problem. It affects women of all ages, racial, cultural and economic backgrounds and is one of the most pervasive human rights violations. In Australia, violence against women is widespread. Women are most at risk in their own homes and at the hands of a current or former male partner. There are long-term detrimental harms for children and young people who are exposed to violence perpetrated against women. It adversely impacts on the health, welfare and development of individuals and families, the communities in which they live and society as a whole.

Definitions and forms of violence against women

Definitions of what constitutes violence against women tend to differ according to the context in which it is being investigated. For example, the legal sector, researchers, advocates and service providers often define the problem in different ways. These range from broad-based definitions that include “structural violence”, such as restricted access to health care and education (UNICEF, 2000), to more narrow, legalistic definitions that focus on specific behaviours consistent with criminal offending (Chung, 2013). Legislative Acts of Parliament in each state and territory in Australia (e.g. Victoria’s Family Violence Protection Act 2008; Tasmania’s Family Violence Act 2004; Western Australia’s Restraining Orders Act 1997) define both behaviours and the relationships in which those behaviours occur to enable protections under law (Department of Families Housing Community Services and Indigenous Affairs, 2009). These definitions do not always align with community knowledge and understandings of what constitutes violence against women and their children.

Although there is no single nationally or internationally agreed definition, the one adopted by the United Nations (UN) over 20 years ago is among the most commonly used. The UN’s Declaration on the Elimination of Violence against Women defines violence against women as:

“…any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or private life” (United Nations, 1993).

In Australia, the terms domestic violence, intimate partner violence and family violence are often used interchangeably. Collectively, they refer to a range of behaviours that are violent, threatening, coercive or controlling that occur within current or past family or intimate relationships. These behaviours are overwhelmingly perpetrated by men against women and include direct and indirect threats of physical assault, sexual assault, emotional and psychological torment, economic control, social isolation and any behaviour that results in women living in fear.

The term family violence is often used to describe violence perpetrated against women by family members in addition to the use of violence by intimate partners. Use of the term family violence also reflects Indigenous communities’ preference because it signifies the broader impacts of violence on extended families, kinship and social networks and community relationships (Cripps & Davis, 2012).

Non-partner violence against women generally refers to women’s experience of physical and/or sexual assault by a person that is not a former or current intimate partner since the age of 15.

Gender equivalence

The UN’s definition of violence against women appropriately recognises that violence is “gendered”. Taking this approach is not designed to ignore or diminish men’s experiences of violence. Men experience substantial amounts of interpersonal violence in Australia, some within the context of intimate partner relationships. Despite recent campaigns in Australia and overseas, such as One in Three (www.oneinthree.com.au) that seek to highlight men as victims, the evidence suggests that men who report being a victim of domestic violence are also perpetrators of violence (Domestic Violence Death Review Team, 2015).

Prevalence and impact of violence against women

In 2013, the World Health Organization (WHO), together with the London School of Hygiene and Tropical Medicine and the South African Medical Research Council, released a landmark report bringing together all available global data to provide worldwide prevalence estimates of violence against women. Focusing on two main forms of violence – physical and sexual violence – the study found that 35 percent of women worldwide have experienced either physical and/ or sexual intimate partner violence or non-partner sexual violence. Most women experienced violence within the context of intimate partner relationships; the primary perpetrator being a male partner (WHO, 2013).

These global figures reflect what is currently known about lifetime prevalence rates in Australia. One of the most widely cited sources of data on violence against women is the Australian Bureau of Statistics (ABS) Personal Safety Survey (PSS). The PSS is a cross-sectional, population-based survey that measures both men’s and women’s experiences of interpersonal violence and uses a similar definition as the global WHO report that focuses on physical and sexual violence only. Data from their most recent survey (2012) showed over a third of women in Australia have experienced physical and/or sexual violence since the age of 15. One in three women (34%) experienced physical violence; one in five (19%) experienced sexual violence (ABS, 2012).

Despite high lifetime prevalence estimates in Australia and overseas, these data are acknowledged as being a conservative or under-estimate of the actual extent of the problem. The most pressing and universal concern is that women do not report or disclose physical, sexual or other forms of violence, particularly by an intimate partner. In Australia, one quarter of all women who experienced violence by a current partner never told anyone about it (ABS, 2012).

The social, health and economic consequences of violence against women are enormous and affect individuals, families and communities. The 2013 WHO report’s systematic review found strong evidence to support links between women’s exposure to violence and direct health outcomes, including mental health problems, substance use disorder, suicide and self-harm and fatal and non-fatal injury. The health, welfare and development of children and young people is also impacted by exposure to violence in the home. A recent comprehensive review of the evidence noted that children’s exposure to domestic violence frequently co-occurs with direct child maltreatment and abuse and that children living in violent homes are at substantial risk of emotional, developmental and learning difficulties (Holt, Buckley, & Whelan, 2008). Violence against women and their children also comes at a very high economic cost. In 2009, the cost to the Australian economy of violence against women and their children was estimated to be $13.6 billion per annum. Without a significant reduction in rates of violence against women and their children, by 2022 it would be an estimated $15.6 billion per annum (National Council to Reduce Violence against Women and their Children, 2009).

Primary prevention

In Australia and overseas, public policy addressing violence against women and their children is increasingly focused on primary prevention: stopping violence before it occurs (VicHealth, 2007). The concept of primary prevention borrows from public health models of disease prevention that target the underlying causes of illness and disease by modifying risk and protective factors. For example, the most effective primary preventive strategy for lung cancer is preventing initiation of cigarette smoking by young people or helping current smokers to quit.

In the context of addressing violence against women the goal of primary prevention is to reduce incidence by targeting factors that give rise to, or create, conditions in which violence against women occurs (Quadara & Wall, 2012). Unlike the clear cause and effect relationship between smoking and lung cancer, the causes of violence against women are complex. Nevertheless, gender inequality is recognised as being one of the key features underpinning explanations for why men use violence against women (WHO, 2013).

Among the many theoretical perspectives that seek to describe the connections between gender inequality and violence, socio-ecological models have gained traction recently (see Figure 1; Brofenbrenner, 1977). Socio-ecological perspectives build on feminist theorists from the 1970’s that situate gender, power and patriarchy as the key explanatory factors, but place a stronger emphasis on the interactions among key factors. For example, power inequalities between women and men, social constructions of masculinity and social norms and attitudes are risk factors for violence against women that primarily happen at the societal level of the model, but filter down and are enacted at the community, relationship and individual level.

Figure 1. A socio-ecological model for understanding violence against women

The primary prevention literature on violence against women is relatively new and mostly focused on building the evidence to intervene at the community and society level (Wathen & MacMillan, 2003). This is largely because intervening here aims to change not just individual behaviours and attitudes, but also wider community perceptions that support, tolerate or condone violence against women. It is within this context that the news and information media – a dominant force in shaping the discourse on matters of public importance – is seen to play a crucial role.

The role of media

The media has been identified in numerous Australian state and national policy documents as a priority area for action on preventing violence against women, including in the Council of Australian Government’s National Plan to Reduce Violence Against Women and their Children 2010-2022 (Council of Australian Governments, 2011). Improving the way media engage on the issue is one of the key objectives under the Second Action Plan 2013–2016 of the national plan to reduce violence against women and their children 2010-2022 (Australia. Department of Social Services, 2014).

The media features as a priority area in primary prevention because of its potential influence on public understanding of violence against women (Carll, 2003). News and information media are key factors in shaping the public discourse because they report on current events and provide a framework for their interpretation. Although news media audiences are not simply passive recipients of information, who or what is selected to appear in the news and how those individuals and events are portrayed can have a profound influence on people’s attitudes, beliefs and behaviours (Flood & Pease, 2009).

Despite a vast repository of research investigating media influence, there are no simple mechanisms to explain how or why audiences are influenced by what they see, read and hear in the news. Rather media influence is conceptualised as a complex process involving multiple players: sources, journalists, editors, and audiences. But unquestionably, the way information is structured can increase public understanding of violence against women and, more importantly, challenge its place in society (Easteal, Holland, & Judd, 2015).

A note on terminology used in this review

Where possible we have elected to use the phrase “women who have experienced violence” in preference to the word “victim”. In some instances, however, we have used the word because it most accurately reflects research findings. For example, the concept of “victim-blaming” is frequently discussed in studies exploring media and violence against women.

We have also adopted phrases such as “male-perpetrated violence” and “men’s use of violence” in preference to terms that portray a sense of mutual responsibility such as “violence within relationships” or “abusive relationships”. For further information on language and recommendations from local and international media reporting guidelines, see Studies on news production in the Results section.

Aim and framework

Aim

The aim of this state of knowledge paper is to provide an overview of the best available contemporary evidence on the way news and information media portray violence against women and their children.

Framework

As a framework for the paper, we have grouped studies into three broad areas of inquiry: representation, reception and production.

Studies of representations

These studies involve analysis of media messages themselves. This body of work seeks to understand the way media portray violence against women by deriving from the text the discursive practices used and the dominance of particular themes or constructions. Studies may investigate the formal properties of media messages as linguistic, narrative or semiotic systems. These studies are descriptive: observing the practice without changing the environment.

Studies of audience reception

These studies involved investigations of media audiences with an interest in how audiences take in media messages about violence against women, and the context within which mediated information is received and how it can be manipulated. These studies largely position audiences as active, as opposed to passive recipients of information.

Studies of news production

These studies concern the manner in which the practice of gathering information and reporting on violence against women happens, including source selection, journalists’ perceptions and structural, organisational or procedural barriers to getting a story published (e.g. the process involved in determining whether or not a story is newsworthy and/or profitable).

Included in this broad area of inquiry are media guidelines and the extent to which they have been developed and evaluated.

Method

Inclusion and exclusion criteria

For the purpose of this state of knowledge paper, we focused on definitions and behaviours most relevant to the Australian context: domestic, family and intimate partner violence and sexual violence/assault (including fatal and non-fatal cases). To be eligible for inclusion, studies also had to explore media representation, reception or production (related to violence against women) as a key theme. All reviewed pieces of literature were primary, peer-reviewed studies or grey literature; books or book chapters, dissertations and theses were excluded (except if the study was subsequently published as a primary, peer-reviewed article).

Studies that did not explore (at least one of) “media representation”, “audience reception” (or “media influence”) and/or “media production” and “violence against women” were excluded. We did not include search terms related more specifically to violence experienced by women in other cultural contexts, such as forced and child marriage, trafficking and forced prostitution, female genital cutting/ mutilation, or “honor” or dowry-related crimes. However, if articles exploring these themes were captured using our search criteria, they were included. Similarly, we didn’t explicitly search for studies exploring sexual harassment; however such literature were not excluded if they met our search criteria. All literature reviewed were written in English and published on or after January 1, 2000.

Database searches

Databases that were systematically searched included Scopus, a bibliographic database of scientific, technical, medical and social sciences, including arts and humanities, and ProQuest, a general database for peer-reviewed literature. These searches were cross-checked with PubMed, a search engine accessing the Medline database of references and abstracts on life sciences and biomedical topics, and Google Scholar, a general-faculty web-based search engine of scholarly literature.

In order to identify studies that related to media representations, production or reception/influence, keyword and subject headings were searched using the following terms:

In title: (journal* OR “mass media” OR media OR newspaper OR news OR TV OR televis* OR radio OR magazine OR press) [within 5 words of] (guideline* OR handbook OR protocol OR manual OR ethic* OR moral* cover* OR represent* OR broadcast* OR portray* OR fram* OR report* OR production OR influen* OR recept*)1^[1].

In order to identify studies that explored “violence against women” we used the following search strings, in addition to those above:

In title and/or abstract: (viol* OR crim* OR rape OR assault OR femicide OR batter* OR IPV) [within 10 words of] (women OR female OR woman OR girl OR prostitut* OR “sex worker” OR partner OR girlfriend OR wife OR wive* OR man OR male OR husband OR boyfriend).

Or, in title, abstract or keyword: (“intimate partner violence” OR IPV OR “intimate partner terrorism” OR “sexual assault” OR “sexual violence” OR homicide OR “murder suicide” OR “homicide suicide” OR “femicide suicide” OR “family violence” OR “domestic abuse” OR “domestic assault” OR “violence against women” OR “gendered violence”).

The literature was supplemented by scrutinising the reference lists of all included studies to identify any relevant studies that had not been considered, and also by using the web-based search engine Google. Grey literature was primarily sourced using Google and examining reference lists of primary sources. All searches were undertaken between March and May 2015.

Selection process

Studies that met the inclusion criteria were disaggregated into three categories depending on their primary focus: media representation, reception and production. Studies categorised under “representation” primarily explored how violence against women is portrayed in the media. Studies categorised as “audience reception” primarily explored how audiences respond to media stories surrounding violence against women. Studies categorised under “production” primarily explored journalistic practice, ethics and protocol behind development and publishing of stories. In this section we also included media reporting guidelines from Australia and overseas. Some articles were categorised in more than one category when there were multiple parts to the study.

Quality appraisal

Critical Appraisal Skills Programme (CASP) collaboration

To assess the quality of primary peer-reviewed studies, we employed assessment tools developed by the UK’s national Critical Appraisal Skills Programme (CASP) collaboration (www.casp-uk.net). The overarching aim of this set of assessment tools is to critically appraise research according to its rigour, credibility and relevance. In accordance with each study’s research design and data collection methods we used either an amended version of the CASP Qualitative Checklist or the CASP Cohort Study Checklist to assess the quality of included studies.

CASP Qualitative Checklist

The CASP Qualitative Checklist includes two screening questions and then a further eight questions to assess overall quality. Each question includes a number of issues to consider. To more accurately reflect the methods used in primary studies of representation, we removed one question on the way the relationship between researchers and participants was considered and replaced it with a question about whether the development and use of coding frameworks was presented and adequately described. Although the tool is not designed to provide a final score, we assigned one point if the study met criteria for each question and then categorised those scores according to low (scores between 1-3), moderate (scores between 4-6) and high quality (scores between 7-10). Unpublished reports were not critically appraised for quality, but are only included in this review if a case could be made that they enriched the knowledge base.

CASP Cohort Study Checklist

As is the case with the qualitative checklist (above), the CASP Cohort Study Checklist includes two screening questions and then a further ten questions; each question outlines a number of issues to consider. As noted above, the tool is not designed to provide a final score, however we assigned one point if the study met criteria for each question and then categorised those scores according to low (scores between 1-4), moderate (scores between 5-8) and high quality (scores between 9-12).

Even though the critical appraisal tools we employed were adaptations of the original, the changes we made were minor. The original CASP tools were pilot tested in workshops, including feedback and review of materials using successively broader audiences, so we are confident that studies included for review meet our standards for scientific rigour, credibility and relevance.

Results: Studies of media representation

Overview

We identified 48 primary studies published on or after the year 2000 that analysed the way news and information media represents violence against women and their children, including 44 peer-reviewed research articles and 4 unpublished reports. Over half the studies were published in the last five years (n=27; 2010-15). The studies originated and analysed media from a broad range of countries, with the majority conducted in the United States (US; n=19), followed by Australia (n=6) and the United Kingdom (UK; n=6), Canada (n=3) and South Africa (n=2). Three studies purposively monitored media reporting in more than one country: one in the US and Canada, one in the US and the UK and one in Saudi Arabia, Bahrain and Kuwait. One study was conducted in each of the following countries: Norway, Israel, Botswana, Spain, Pakistan, Turkey and New Zealand. One study used a cross-cultural comparison; comparing media reports in Canada and Romania, although the selected countries appeared more a circumstance of convenience than design. One study searched English language news sources worldwide.

Study characterisations including country of origin, medium, forms of violence, study design and quality appraisal ratings are included in Appendices A and B.

Medium

The vast majority of research on violence against women in the news and information media concerned traditional media, newspapers in particular. Studies of newer forms of digital and social media were rarer. Forty of 47 studies identified in this review analysed media portrayals of violence against women as they appeared in newspapers, with more recent studies including both online and print news. One study examined newspaper headlines only.

There are a number of possible reasons why studies exploring representations of violence against women are dominated by analyses of newspapers. In Australia and in many other parts of the world newspapers are considered a highly trusted and credible source of information (Morgan & Politoff, 2012). Newspapers have also traditionally served as forum for society to engage, discuss and debate issues of public importance. On a more practical level, newspapers are an accessible and inexpensive way to monitor media on any subject matter, including violence. Electronic databases such as Dialog, LexisNexis, Newsbank, Newstext, Access World News and Factiva offer historical collections of newspapers. These databases facilitate accurate and comprehensive monitoring by allowing researchers to search, retrospectively, particular newspapers, timeframes, jurisdictions, search terms or a combination of all these elements. Most studies used either one of these databases or similar country-specific databases, such as the Norwegian Atekst. Other studies described alternative methods for sourcing media including “hand” searching archival copies or using pre-existing databases.

Four studies analysed television news coverage only. Three studies analysed multiple media platforms including a combination of newspapers, radio, television, blogs and webcasts. We included one study that examined user-driven content to an online news site, whom the authors described as “citizen journalists”.

Forms of violence

Studies of representation investigated many forms of violence against women including sexual assault and rape, physical and emotional abuse and intimate partner homicide. Some studies specifically selected and analysed media coverage of interpersonal violence generally, from which violence against women was either a subset of the data or provided a means to compare “styles” of reporting. These studies included media reports of violence perpetrated by men and women. In a number of studies, the frequency of media reports of violence against women were compared with “true” crime rates in the community from locally derived data.

In studies that elected to focus on one form of violence only, the most frequent was the murder of an intimate partner; referred to in the literature as intimate partner homicide or in the case of the murder of a female intimate partner; femicide. Fourteen of the 48 primary studies identified in this review focused on media depictions of intimate partner homicide and/or femicide. Some studies explored media portrayals of both male- and female-perpetrated intimate partner homicide, despite the rarity of the latter event. One study purposively sampled newspaper articles in which women had killed a male intimate partner.

Ten studies specifically explored media portrayals of sexual violence against women. The studies investigated media reporting of both men and women as perpetrators of sexual violence. Two studies focused on media reporting of cases in which the allegation of rape was called into question – described in the studies as unfounded or false rape.

Study designs and data collection

Data were collected for studies of representation in one of three main ways. The first can be broadly described as cross-sectional, whereby data were collected from one or more news media over a pre-defined study period. There were many variations on this study design ranging from scrutinising a large number of media over a short period of time (e.g. 2 weeks), to searching for items in a single type of media or on a single form of violence over a longer period of time (e.g. 25 years). Studies seeking to draw inferences about rarer events, such as portrayals of women who kill in the context of intimate partner violence required longer data collection periods than studies investigating all forms of violence. There was a large variation in the range of media items captured and analysed (range n=16 to n=2452). Some studies initially collected large amounts of media items and then analysed a purposive or random subset of that data; we refer only to their analytic sample.

The other main approach to data collection was in the context of case studies. There were 12 studies in which data were collected in connection to one or more “cases”. Typically, these studies investigated media representations of a particular incident, event or act of violence. The incident or act was generally a highly publicised case, so the volume of media generated provided researchers with a wealth of information to study. Data were mostly time and geographically limited and collected in accordance with when and where the incident occurred. Examples include media reporting of: a rape in India; sexual assaults perpetrated by sportsmen including footballers in Australia, lacrosse players at Duke University in the US and US basketballer Kobe Bryant; a triple murder in Norway; the rape of two women in Israel; the murder of two women in Australia; the rape and murder of a British woman in India and the rape of, and subsequent recantation (withdrawal of complaint) by, a woman in the US. The major limitation of case study approaches, frequently noted by researchers, is that highly publicised cases are often “unusual” – garnering ongoing public interest because they are “different”. One study from South Africa used an awareness campaign as their “case” and monitored media for the 16 Days of Activism campaign.

Five studies were specifically designed to examine changes in media reporting of violence against women over time. One study collected data from newspapers in one state of Australia (Victoria) over three separate time periods. One study examined changes in media reporting before and after the OJ Simpson trial in the US. One study examined television coverage in the Catalonian region of Spain in the years 2002, 2004, 2007 and 2010. One study analysed all male-perpetrated intimate partner homicide reported in daily newspapers in Toronto, Canada over two separate time periods. The final study that explored change over time was a series of unpublished audit reports from New Zealand that examined changes in media reporting in response to a public campaign raising awareness of family violence in that country.

Even though all studies of representation were published in the last 15 years, the period of time the researchers monitored and extracted data ranged from 1975-2013.

Analytic approach

Although approaches to studies of media representations were informed by a broad range of theoretical perspectives, all employed a content, discourse and/or frame analytical approach.

Content analysis refers to a general set of techniques used for organising and understanding collections of text (Stemler, 2001). It is not a single method but a collection of quantitative approaches to counting and comparing text-based data. Content analysis is commonly used in media and communication studies because it is applicable to a broad range of “texts”.

Discourse analysis refers to qualitative approaches where aspects of the text are examined, not just for their presence or absence, but for meanings that underpin their inclusion or omission (Mogashoa, 2014). Discourse analysis is a collection of qualitative techniques that can be used independently from, or in parallel with, content analysis to examine the meaning- making process of media discourse.

Frame analysis was used less frequently than either content or discourse analysis and refers to analysing the way news “frames” particular events. Frame analysis refers to an approach whereby the entire structure of the story is examined for meaning including the selection of words, phrases and images used and whether these selections invite the audience to see issues, action and events in a particular way (Pan & Kosicki, 1993). One study included in this review described frame analysis as bridging the gap between quantitative and qualitative research designs (McManus & Dorfman, 2005).

Quality appraisal

Most primary peer-reviewed studies were categorised as high quality (n=34). The characteristics of high quality studies included a clear statement of aims, an appropriate and well-articulated methodological approach, well defined and replicable data collection protocols combined with rigorous analyses and a clear statement of findings.

Seven studies of representation were categorised as moderate quality and included some but not all of the above criteria. Many of the studies were categorised as being of moderate quality because of a lack of detail surrounding their methodological approach.

Only four peer-reviewed primary studies were categorised as low quality. These studies were not included in the synthesis of outcomes below as the aim of this review was to provide only the best available evidence (Harries & Bird, 2005; Korn & Efrat, 2004; Moffitt, Fikowski, Mauricio, & Mackenzie, 2013; Waterhouse-Watson, 2012). Unpublished reports were not subject to quality appraisal and are included below, but are referred to as unpublished reports.

Outcomes

Overall, studies included in this review reported strikingly similar results indicating predominant patterns in the way news and information media report on violence against women. Many of these themes were previously identified by two recent reviews of the literature (one unpublished), both of which were conducted in Australia. We used these reviews to underpin our understanding of the key themes, paying particular attention to the recent work of Easteal, Holland & Judd (2015) – Enduring themes and silences in media portrayals of violence against women.

Below, we present the common themes as separate categories. However, many share common elements and, thus, should not be considered as mutually exclusive. For example, overly- sensational news frames tend to minimise the social context while drawing attention to isolated incidents. Similarly, event- based reporting frequently intersects with perceptions of blame and responsibility: together they enable the focus to be shifted from the perpetrator of the violence. An over-reliance by journalists on law enforcement as sources of information may result in details about women that inadvertently place their behaviour, demeanor or conduct as contributing to violence perpetrated against them.

Social context

One of the most common study findings was that media reports of violence against women were highly skewed towards event-based reporting or “episodic framing”. That is news reports that focus primarily on discrete incidents or events located at specific places and times while largely ignoring the overarching social context in which violence against women occurs. The social context generally refers to the way social factors of the crime are reported, for example, that violence perpetrated against women is gender-based and rooted in the social phenomena of gender inequality. The way stories are framed can impact on how news audiences assign responsibility. Episodic framing tends to elicit individualistic rather than societal attributions of responsibility. Thematic framing – those that include elements of the social factors of the crime – has the opposite effect (Scheufele, 1999). While there are a number of ways to frame news stories of violence against women thematically, such as providing statistics on the extent of the problem, expert analysis or other information about the social phenomena of gendered violence, few media reports do so.

Homicide in the context of intimate partner violence

Failure to present murder within the context of intimate partner violence was noted frequently among studies investigating media representations of intimate partner homicide (Bullock, 2010; Bullock & Cubert, 2002; Morgan & Politoff, 2012; Richards, Gillespie, & Smith, 2011; Taylor, 2009; Wozniak & McCloskey, 2010). These researchers noted that newspapers rarely documented a history of violence by perpetrators, with a “hyper focus” on method over cause as if it were more important for readers to know how but not why men kill their partners (Sweeney, 2012).

Richards et al. (2011) analysed 995 newspaper articles in the US state of North Carolina from 2002 to 2007 and found clear differences in reports that framed homicides within the context of intimate partner violence compared to those that did not. Newspaper articles framing the murder as intimate partner violence were more likely to include information for women on where to seek help, although this occurred in less than 13 percent of their sample. These results were replicated by Gillespie et al. (2013). They noted that reports framing femicide within the broader social context of violence against women used advocates as sources of information about the crime and included information for readers about safe places for women to go if they needed help. An unpublished report examining press coverage in one state of Australia (Victoria) found very similar results, albeit across all forms of violence. Only 37 of 1739 newspaper articles included in their sample provided readers with information about services for women experiencing violence: most of these were contained within reports that framed the news thematically (Morgan & Politoff, 2012).

One study noted that episodic framing tended to be more evident in breaking news and first reports of homicide than in follow up stories (Wozniak & McCloskey, 2010). While this study from the UK explored traditional news media (newspapers), the finding seems particularly pertinent given the proliferation of online news platforms with a clear commercial interest in “breaking news”. It may be that the growing demand for shorter, bite sized pieces of information (common in online news media reports) is less compatible with reporting styles that place violence against women within a broader social context. Marhia (2008) analysed news reports of rape and sexual violence across different press media in the UK and found that online news stories offered less in-depth analysis of sexual violence than traditional printed formats.

Violence reported within a broader social context

Researchers exploring media representations of other types of violence against women also found a predominance of event-based reporting. Maxwell et al.’s (2000) study investigated how the OJ Simpson trial in the US influenced coverage of violence against women. They reported individual explanations appeared nearly twice as frequently as social explanations. Their hypothesis that the OJ Simpson case would result in more social reporting was not supported.

Bou-French (2013) found that contributions to an online news site by readers most often situated violence against women as an individual or private problem. Although this study included contributions from only 45 “citizen journalists,” it supported Easteal et al.’s (2015) assertion that online and social media platforms may simply mirror myths and stereotypes that appear in traditional media. A recent study of press portrayals of violence against women in the Arabian Gulf States (Saudi Arabia, Bahrain and Kuwait) noted very little difference from news frames used by media in Western cultures, whereby most reports ignored the systematic nature of violence against women (Halim & Meyers, 2010).

Sims (2008), in the only study identified in this review that examined media reporting of emotional abuse, suggested episodic framing positioned violence against women as an anomaly in society. They described that readers were typically left with the impression that abuse, in any form, was not a part of everyday life for most women. Media reporting of violence against women as atypical anomalous events will be explored further in the following sections (see Sensationalism and Misrepresentations and “rape myths”).

Only one study, exploring television news broadcasts in the Catalonian region of Spain, showed that most news items on violence against women included contextual information (Comas-d’Argemir, 2014). The research findings were in contrast to other studies and the author attributed the difference to political, legislative, judicial and social actions in Spain that have substantially changed the landscape in how that country approaches gendered violence. This study, along with a cross cultural comparison between Canadian and Romanian newspapers (Oxman-Martinez, Marinescu, & Bohard, 2009) were the only primary studies that referenced the way media represent children in reports of violence against women.

Comparison to news media reporting of other issues

Episodic framing in news reporting is not unique to media coverage of violence against women. Previous research on media representations of mental illness and suicide show that news media typically highlight particular aspects of an event while ignoring the overarching patterns of risk that may help explain that event (Blood, Putnis, & Pirkis, 2002). Studies exploring public understandings of other issues, such as war and terrorism, have also noted a lack of context in media reports (Iyengar & Simon, 1993). Nevertheless, a study by McManus and Dorfman (2005) in the US found that, in comparison to newspaper coverage of other violent crimes, news stories about intimate partner violence were more likely to appear “as a simple description of a particular violent episode” (p. 53).

Sensationalism

One of the key themes that emerged from this review was the frequency in which news stories of violence against women were sensationalised. Sensationalism in the news is often defined in terms of its capacity to provoke attention or arouse certain emotional responses in viewers (Grabe, Zhou, & Barnett, 2001). It shares much in common with “newsworthiness”. Sensationalist news reporting, although open to a certain degree of subjectivity, can manifest in what stories are chosen to feature in the news, what facts are included and emphasised, what is omitted and the language used. Media reporting of the most salacious aspects of violence against women provides the public with a perspective that is provocative but not representative.

Female perpetrators of violence

A number of studies found that female perpetrators of violence, including sexual and physical forms of violence, as well as female perpetrated homicide of a male partner accounted for a high proportion of the total volume of media reports on violence, despite the rarity of these events (Anastasio & Costa, 2004; Naylor, 2001; Skilbrei, 2013). While the over- representation in the news of female perpetrators of violence may reflect the news value of “novelty,” Naylor (2009) argued that violence by women is considered “more deviant, more anxiety-provoking, more transgressive than men’s violence” (p. 188) and, therefore, more in need of explanation.

Focusing specifically on women who killed their abusive partners, Noh et al. (2010) examined 25 years (1978-2002) of newspaper articles in the US and Canada. Their study was premised on the hypothesis that “a woman who kills provides extant sensationalism” (p. 27) and this was indeed borne out by their analysis of 250 newspaper articles. Women who killed were presented simplistically as being either “mad or bad”. Morgan and Politoff (2012) also found that press stories of female-perpetrated crimes, particularly murder, were more likely to include sensationalistic elements than male-perpetrated crime. Skilbrei (2012), who analysed newspaper coverage of a high profile triple murder on a farm in the Norwegian countryside in 1999, proposed that the incident and subsequent court proceedings gained and sustained media attention because women featured both as “victims” and perpetrators of the crime, alongside men (for a detailed overview of media representations of female- perpetrated homicide – partner and non-partner, see Easteal, Bartels, Nelson, & Holland, 2015).

Although Skilbrei (2012) was highly critical that media coverage of both the case and court proceedings were imbued with entertainment value, she also acknowledged that commercial and economic interests for media owners and producers create a bias towards sensationalism. This sentiment was echoed by other researchers. For example, Jackson (2012) examined national television news coverage of the kidnapping and rape of a woman (Megan Williams) in West Virginia in 2007 and her recantation in 2009. Jackson noted that the presence of several female perpetrators presented a story that was clearly “good copy” and was particularly critical of the coverage by Fox News. Jackson described that constructions of both race and gender in Fox News’ television news coverage were entrenched in dominant white patriarchal ideologies, but also acknowledged that economic and professional demands of newsrooms reward journalists for focusing on atypical cases. Other researchers have noticed differences in the degree to which press stories use sensationalism and noted that tabloids tend to emphasise sensational news topics more than broadsheets (Marhia, 2008; Morgan & Politoff, 2012).

Some types of violence are more “newsworthy” than others

Sims (2008) noted that the imperative to fill the daily news with stories that capture public attention is one of the reasons why emotional abuse receives such scant attention. Sims, along with a number of other researchers suggested the media has a “murder-centric” focus, whereby reports of homicide between intimate partners takes precedence over reporting other forms of violence against women, particularly emotional abuse (Carlyle, Slater, & Chakroff, 2008). The “murder-centric” focus of news and information media is problematic because it may result in erroneous assumptions in the community that violence against women often ends in death. It is interesting to note that while researchers criticised media for having a “murder-centric” focus, this same criticism could be levelled at academic scholars. A disproportionate number of the primary studies identified in this review – over a third of all studies – focused exclusively on media coverage of intimate partner homicide: stories that are newsworthy but statistically unlikely.

In the same way that female perpetrators of violence garner attention in the media, a number of researchers noted that the high volume of news coverage dedicated to apparently random acts of violence against women illustrates that such stories fulfil key news values (Marhia, 2008; Mason & Monckton-Smith, 2008; Morgan & Politoff, 2012; Naylor, 2001). The over-representation of stranger perpetrated violence cultivates and sustains misplaced fear and community misunderstandings about who is at risk. Nevertheless, this is a complex area to untangle – does the media dedicate more time to random incidents of violence against women because they are sensational or provocative or does the media report incidents in such a way as to appear as random acts of violence? This will be explored further in the following section on Misrepresentations and “rape myths”.

Language

A number of researchers focused on the way language is used to sensationalise stories with a general consensus that sensational and exceptional language is commonly used in media reports of violence against women (Ahmed, 2014; Alat, 2006; Comas-d’Argemir, 2014; End Violence Against Women Coalition, 2012; Evans, 2001; Jackson, 2013; Mason & Monckton-Smith, 2008; O’Hara, 2012). Sensationalist language is used to titillate, shock, fascinate, amuse and entertain. Jackson (2012), in their case study of the kidnapping and rape of Megan Williams, found the media consistently used sensational language to describe the crime including “out of a horror movie,” “unreal,” “unspeakable,” “bizarre,” “shocking” and “hard to comprehend”. O’Hara (2012) analysed 124 newspaper articles relating to specific cases of sexual violence and noted that some reports used animal metaphors to describe perpetrators including words like “predator,” “beast”, “nocturnal” and “pounce,” “stalk” or “prowl” to describe behaviours.

Language can also be used to obfuscate the true nature of the crime. Sampert (2010) who analysed all stories on sexual assault that appeared in six English-Canadian newspapers in one year (2002), found that “sex assault” or “sex attack” was used more frequently than the term “sexual assault” and that it appeared as a deliberate lexical choice made by journalists and/or editors. This deliberate choice of language perpetuates the myth (perhaps unintentionally) that a sexual assault is, at least partly, analogous to sex by using language that ignores the violence inherent in the act.

Morgan and Politoff (2012) found that, while 40 percent of the newspaper articles incorporated some elements of sensationalism, there was a general decreasing trend over time in the Australian press. They also noted less provocative or salacious news coverage in their sample of news stories than had been reported in other countries, such as the UK.

Also of note is a series of reports from New Zealand that examined stories from all metropolitan and provincial newspapers and major television and radio networks about family violence over four time periods (McGuiness, 2007- 2008). The aim of the quarterly reports was to assess the impact on media of a social marketing campaign – Campaign for Action on Family Violence – designed to reduce tolerance for the country’s high rate of family violence. The reports showed a number of positive results that were sustained over time including the uptake by media of the campaign’s main message “It’s not OK”. The reports noted the phrase was used in newspaper headlines and quoted by radio programs, and proposed that it provided audiences with the language to speak out about family violence.

Misrepresentations and “rape myths”

Whether as a by-product of episodic framing or sensationalism in news media, one of the most common criticisms is that media depictions of violence against women bear little resemblance to the reality of gender-based violence. Although a number of studies found media misrepresentations of violence against women were common, it seemed particularly problematic among media portrayals of sexual violence –sometimes described as “rape myths”. The term “rape myth” was first coined in the 1970s (Brownmiller, 1975) and defined in a research paper more than 35 years ago as being “prejudicial, stereotyped or false beliefs about rape, rape victims and rapists” (p. 217; Burt, 1980). According to Mason and Monckton-Smith (2008), in the public consciousness “a real rape happens in public places, is perpetrated by a stranger and involves aggravated violence” (p. 691). While it is acknowledged that a multitude of sources contribute to stereotypical constructions of sexual violence in the community, the media play an active role. Understanding how the media perpetuate or challenge such stereotypes was the focus of a number of studies identified in this review(Bonnes, 2013; Dwyer, Easteal, & Hopkins, 2012; Franiuk, Seefelt, Cepress, & Vandello, 2008; Franiuk, Seefelt, & Vandello, 2008; Jackson, 2013; O’Hara, 2012; Sampert, 2010; Toffoletti, 2007; Tranchese & Zollo, 2013).

Stranger danger

One falsehood commonly described by researchers examining news coverage of sexual violence was that the greatest risk for women comes from a stranger. A number of researchers showed that media focus a disproportionate amount of news coverage on cases involving stranger rape. Using police records as a source of comparison, Marhia (2008) examined a random sample of 136 newspaper articles about rape and sexual assault appearing in mainstream UK newspapers and on one online news site. They found that while around 90 percent of rapes reported to police in the UK in 2006 were by a known man, the majority of rapes reported in the press were by unknown male perpetrators. They noted that rape by current or former male intimate partners was almost invisible in their sample of press coverage, accounting for only two percent of all stories about rape. Morgan and Politoff (2012) compared crime data from the Australian Bureau of Statistics (ABS) with press reports of sexual violence perpetrated by men and found similar results to Marhia (2008). They also noted that while stranger rape was commonly reported, a relatively high proportion of newspaper articles failed to specify if there was a current or pre-existing relationship, thereby reinforcing the myth of “stranger danger” in the context of rape. Their analyses of newspaper reporting showed little change over time.

A study examining articles from a local newspaper in a small town in South Africa was the only one that found that the relationship between men and women in the context of sexual violence was more often reported than not (Bonnes, 2013). The very small number of newspaper articles included in their study (n=16) may explain why their findings are in contrast to the bulk of the studies in this review.

Victims and perpetrators of sexual violence as “other”

A number of studies highlighted a group of interrelated myths that commonly appear in media portrayals of sexual violence and relate to misrepresentations about both male perpetrators of, and women who experience, sexual violence, including where rape happens. Burt (1980) suggested that “rape myths” are taken up by the community because they provide a way for people to view and understand violence as something that happens to people who are different to themselves. A number of researchers noted that media reports on rape frequently portray both women and men as “other” – their “otherness” deriving from the notion that most women don’t get raped and that most men don’t rape (Barnett, 2012; Evans, 2001; Jackson, 2013; O’Hara, 2012; Sampert, 2010; Saroca, 2013; Toffoletti, 2007).

For women, a number of studies showed that depictions of “otherness” were often used by media when women deviated from their stereotypical female role. Studies included in this review noted many variations on this theme, including descriptions of female “victims” as promiscuous, welfare dependent, with underworld dealings, as mail order brides, dangerous, manipulative, antisocial, eccentric, deranged or mentally ill. Barnett (2012) examined how four newspapers in the US covered the rape, arrest and dismissal of charges against members of the men’s Lacrosse team at Duke University in North Carolina. The study eloquently summed up the way press portray women,

Ultimately, media accounts presented Mangum as a woman who was something women are not supposed to be – out of control. She was sexually promiscuous, mentally ill, a drug and alcohol abuser – the antithesis of the traditional feminine role (p. 24).

There were an equal number of variations of “otherness” for male perpetrators of sexual violence. A number of studies note that media describe men as monsters, sick, insane, disturbed, evil or deviant in a way that reinforces the notion that sexual violence can be attributed to wayward individuals rather than cultural and social phenomena (Bullock & Cubert, 2002; O’Hara, 2012; Toffoletti, 2007). Sampert (2010) analysed 1532 newspaper articles about sexual assault that appeared in six English Canadian newspapers in 2002 and found that the press often identified male perpetrators by way of their beliefs, race or nationality creating the impression and misunderstanding “that women can recognise their rapists” (p. 312). The reports framed rape as an anomaly that only happens outside of normative white society. Sampert noted that rape perpetrated by a male current, former or dating partner remained largely invisible in the media.

Jackson (2012), in a case study of the kidnapping and rape of Megan Williams in the US state of West Virginia in 2007 highlighted that news stories repeatedly used words that reinforced dominant ideologies about sexual violence and where it happens. Many of the news stories constructed the scene in terms of its “rural” location and added words like “trailer,” “mobile home” and “shack” to describe the scene of the crime. Jackson noted that these words were actually used more commonly than the word rape. Other studies confirmed that reports of sexual violence frequently draw on stereotypical ideas of disadvantage and underclass to describe where women may be most at risk of sexual violence (Evans, 2001).

Reporting allegations

Two studies identified in this review specifically examined media reports in which an incident of sexual violence against a woman was called into question – described in these studies as false or unfounded rape (Barnett, 2012; Franiuk, Seefelt, Cepress, et al., 2008). A case study of media coverage of rape by the US basketballer Kobe Bryant found that news media perpetuated a contemporary rape myth: that women can and do lie about sexual violence perpetrated against them (Franiuk, Seefelt, Cepress, et al., 2008). Even though the newspaper articles analysed in this study were all published prior to when the assault charges were officially retracted, 42 percent of all media reports questioned whether the woman might be lying. Barnett (2012), in the study on media reporting of false rape allegations involving Duke University lacrosse players, found that in the 633 newspaper articles about the incident, arrest and dismissal of charges, the focus never strayed from narratives of injustices for men. Despite ample opportunity, the media did not explore why a woman may not pursue a complaint of sexual assault or retract it.

Sexual violence and sport

One of the misrepresentations surrounding sexual assault that emerged in this review of media representations of violence against women is the intersecting stories of sportsmen and violence. Three studies explored sexual violence and sportsmen, two of which were in the context of “unfounded” rape allegations, as described above. One further study, conducted in Australia, examined how the sports sections of two major newspapers reported sexual violence in the wake of highly publicised cases involving Australian Rules footballers (Toffoletti, 2007). One of the key themes that emerged was the enduring myth of predatory women. In each of these studies, media portrayals of women as “other” created and sustained assumptions that incidents of sexual assault occur when women deviate from socially constructed conventions of heterosexual relations. The men, on the other hand, were frequently described in the context of their sporting prowess. Franuik et al.’s (2008) study on media representations of a rape case involving US basketballer Kobe Bryant found that among the 156 newspaper articles, 25 percent included at least one positive comment about Kobe Bryant as an athlete. In response to the Leveson Inquiry in the UK, the End Violence Against Women Coalition examined press coverage in 11 British national newspapers and found numerous examples of coverage in which the loss of a man’s sporting career and/or reputation were the focus of reports on violence against women. In each of these studies it was clear that the media privileged sporting men’s reputations over the interests of women.

Blame and responsibility

Many studies identified in this review highlighted the way news and information media construct stories that either directly or indirectly attribute blame and/or assign responsibility for violence against women. It is most often referred to in the literature as “victim blaming” and refers to circumstances in which victims of crime are portrayed so that it appears they are partially or entirely responsible for the transgressions committed against them. Overall, studies showed that attributing blame and assigning responsibility was not generally overt but mediated through subtle gendered messages (perhaps unintentionally). Such messages were promulgated by how the stories were structured, the language used, the order in which incidents were described and whether coverage included information on precipitating factors or other reasons that could shift the blame from male perpetrators to women. Taylor (2009) noted that one of the primary barriers to substantial social change is the way media sustains the myth that women are responsible for men’s use of violence.

Behaviour of women

Morgan and Politiff (2012), along with a number of other researchers, noted that media do not necessarily engage in direct “victim blaming” but include subtle details that suggest to the reader that women may have played some role in the violence perpetrated against them. News portrayals of violence may ascribe responsibility to women by focusing on characteristics that emphasise their inherent difference, commonly by way of race or class (as previously described -see Misrepresentations and “rape myths”).

Sexual deviancy or promiscuity was commonly identified in studies that examined media coverage of sexual violence and rape, but researchers also found that it featured as a background detail in media reports on all forms of violence against women (Barnett, 2012; Noh, Lee, & Feltey, 2010; Saroca, 2013; Skilbrei, 2013; Toffoletti, 2007). For example, Saroca (2013) found that Australian news coverage of the murder of two women positioned them as being complicit in their deaths because they were threatening – physically and sexually – to men and women. This was enabled in media reports by sustaining racist stereotypes of what it means to be a Filipino woman living in Australia.

A number of studies also noted that the use of drugs and alcohol – being drunk, passed out, out of control – were frequently used by media to convey evidence of recklessness by women (Barnett, 2012; Dwyer et al., 2012; Evans, 2001; Richards et al., 2011). In this way media shifted blame from perpetrators to women because their behaviour placed them at risk. Focusing specifically on the issue of sexual assault involving alcohol, Dwyer et al. (2012) examined reports in four newspapers in one state of Australia (New South Wales) from 2009 to 2011. The study found that print media frequently convey both implicit and explicit messages that women can avoid rape by limiting alcohol consumption. However, the study also found evidence of media reports that challenged that view, mostly appearing in opinion pieces rather than news items.

Mutuality and precipitating factors

In the context of domestic violence and intimate partner homicide, studies commonly found that media confer a mutuality of responsibility for violence against women by including information about arguments that preceded the violence (Evans, 2001; Exner & Thurston, 2009; Roberto, McCann, & Brossoie, 2013). Researchers found arguments between intimate partners were described in media reports as “squabbles”, “disputes” or “marital difficulties” and that describing relationships as “volatile” conveyed a sense of mutual combat. In the case of homicide, Sims (2008) suggested that the pairing of arguments and death gives media audiences the impression that murder is a consequence of a fight between intimate partners that goes awry (Sims, 2008). Easteal et al. (2015) also noted that a sense of mutuality situates family violence in the context of a relationship problem and not as a crime.

Another precipitating factor found among studies of media representation was provocation. The idea women can and do provoke or trigger the violence that is used against them was sustained in media coverage. The most commonly described “triggers” were infidelity, terminating the relationship or the age-old stereotype of the overbearing or nagging wife/ girlfriend. Mason and Monckton (2008), in their case study analysis of newspaper reporting of violence against women in the UK, found that press coverage frequently prioritised sexual motives as the etiology for the violence. In one case study of femicide, the researchers found that the woman’s “affair” was cited regularly by the media as the cause of her murder and was subsequently used by defense lawyers in the trial to construct a mutuality of responsibility between the dead woman and her partner.

In a study investigating newspaper reports of intimate partner violence among older adults in the US, Roberto et al. (2013) found most cases reported were murder-suicides where a man killed his female partner before killing himself. The poor health of women, along with caregiver stress was frequently reported as contributing to men’s use of violence later in life. Many cases of intimate partner homicide were framed as “mercy” killings conflating the issue of violence and euthanasia.

Responsibility for prevention

A number of studies in this review demonstrated that media, often subtly, create and sustain a public discourse that the prevention of men’s violence is women’s responsibility (Alat, 2006; Dwyer et al., 2012; Gillespie, Richards, Givens, & Smith, 2013; McManus & Dorfman, 2005; Morgan & Politoff, 2012; Richards et al., 2011; Saroca, 2013; Toffoletti, 2007). A number of studies noted that reports or accounts that warn women about sexual assaults or predators in the area highlight women’s potential role in violence against them. The reports implicitly, and sometimes explicitly, imply that women should modify their behaviour in the face of such warnings including where to go, when to go, what to wear, how long to stay out, how much to drink and who to go with. An unintended consequence of such reporting is that women who are subject to violence are somehow responsible because they did not heed the warnings (Richards et al., 2011).

One of the most enduring myths sustained by media is that women living with abusive and violent partners are responsible for the violence because they did not leave the relationship. This is particularly evident in media reporting of intimate partner homicide. Studies showed this myth can be communicated in many ways and was described by a number of researchers as an example of direct “victim blaming”. For example, Richards et al. (2011), in their study of over 900 newspaper articles in the US state of North Carolina, found that ten percent of press coverage directly blamed women who were killed in the context of intimate partner violence because they did not take the necessary action to protect themselves: by not reporting the violence, not filing charges or for not leaving the relationship. They noted that although these represented a minority of newspaper items, the message was a powerful one about women’s responsibility for prevention, particularly from current partners. Blaming women who are killed by a partner reflects and sustains deeply embedded cultural skepticism about why women don’t leave. In Australia, a recent survey found that 78 percent of respondents reported that they cannot understand why women stay with an abusive partner (Webster et al., 2014).

Excuses for perpetrators

For men, reasons proffered by the media to explain or excuse why men use violence against women or kill an intimate partner are common. One study that analysed newspaper articles on violence against women in Botswana found that even in articles that explicitly blamed the perpetrator for the violence, press coverage also spent considerable time discussing mitigating circumstances for men; most commonly anger management, impulse control and emotional stress (Exner & Thurston, 2009). These researchers also highlighted the frequency with which “love” was used to excuse perpetrators, with newspaper articles variously describing intimate partner homicide as “passion killings,” “love-inspired” or “love turned sour”. Studies from a number of other countries, including Canada, Australia and Turkey have also noted that media can romanticise intimate partner homicide (Alat, 2006; Sampert, 2010; Saroca, 2013). Media representations that men were “acting out of love” ensure everyone is a victim of the violence including the perpetrator. Fairbairn and Dawson (2013) in their analysis of newspaper coverage of male perpetrated intimate partner homicide in Toronto Canada found a significant decline in perpetrator- excusing news frames in the years 1998 to 2002 compared to media coverage in the mid 1970’s.

Other researchers described more subtle ways that media reduce male perpetrator responsibility including empathising with the perpetrator or by making men invisible in portrayals of violence against women. A report from the End Violence Against Women Coalition in the UK found that press coverage of “celebrities” commonly expressed empathy towards the perpetrator (End Violence Against Women Coalition, 2012). This was particularly evident in reporting sportsmen’s use of violence against women. Bonnes (2011) provided examples whereby newspapers linguistically separated the perpetrator from the crime including in headlines like “Raped for a cellphone” (p. 216). De-emphasising or excusing perpetrators can also happen when media describe crimes in ways that make them comical, humorous, frivolous or even mundane. Morgan and Politoff (2012) noted that although sensationalistic elements appeared to have reduced over time in Australian press reporting, the inclusion of humor was particularly troubling because it substantially diminished the seriousness of the issue.

It is interesting to note that some of the factors most commonly used by media to excuse men who use violence are the same factors that media use to blame women for the violence. Drugs, alcohol, mental illness, jealousy and provocation were commonly described in studies as mitigating factors used by journalists to reduce men’s ownership of the violence. Researchers describe this as “gendered” reporting: women are treated in one way because they are women, men are treated in another because they are men.

One study, comparing newspaper reporting of violence against women in the US with data from the Centre for Disease Control (CDC), found the most significant discrepancy was in the area of substance use (Carlyle et al., 2008). In the media, the perpetrator was described as being under the influence of alcohol in about six percent of stories, yet in crime statistics, perpetrators of intimate partner violence are reported as using alcohol around the time of the incident in three out of four cases. The researchers noted the dilemma between accurate reporting in media and providing contexts that may be misconstrued as excuses.

Voices of authority and opinion

Many of the studies in this review of representation explored who broadcasts news on violence against women and the sources of information they draw on to provide expert opinion (Barnett, 2012; Bullock, 2010; Bullock & Cubert, 2002; Fairbairn & Dawson, 2013; Lindsay-Brisbin, DePrince, & Welton-Mitchell, 2014; Marhia, 2008; Morgan & Politoff, 2012; Oxman-Martinez et al., 2009; Richards, Gillespie, & Smith, 2013; Taylor, 2009; Wozniak & McCloskey, 2010). The people, professions or associations cited by media indicate to audiences who are regarded as the authority on the situation. The way the incident is portrayed, including situating violence within a broader social context or providing sources of help for women, can be influenced by who media choose to consult with and cite in their coverage.

Sources of information

Researchers have paid much attention to what sources of information media use when reporting on violence against women. The reasons it appears so frequently in research as an outcome may be because these data are relatively easy to extract and code; unlike other elements of news reporting, such as perpetrator excusing frames, which are more difficult to identify from the text and may be subject to bias. Yet, regardless of the reason, studies have provided good evidence that media rely heavily, and sometimes exclusively, on those connected to the criminal justice system as their main source of information.

In intimate partner homicides, for example, Bullock and Cubert (2002) found that press accounts in the US state of Washington relied significantly on information from criminal justice sources such as police, legal professionals and court documents. Bullock, in a follow-up study in 2010 on newspaper reports on intimate partner homicide in the US state of Utah, referred to this as a “police frame” – coverage that relies almost exclusively on factual, unattributed information and/or information from law enforcement professionals (Bullock, 2010). Bullock noted that a “police frame” resulted in a “no frills, fact-oriented approach that tended to focus on the who, what, where, when and how of the crime” (p. 46) and provided readers with a vastly distorted view of women’s experiences of violence. Comments from family, friends, neighbours and circumstantial witnesses may have the same effect – inadvertently contributing to news frames that blame women for men’s use of violence by, for example, questioning why someone chose to stay with a man who was known to use violence.

A range of studies noted that the heavy reliance on legal and/or uninformed sources of information meant that the voices of advocates, service providers, researchers and women with lived experience – those most equipped to situate violence against women within the broader social context – were rarely heard. Fairbairn and Dawson (2013), who demonstrated positive changes in some aspects of newspaper reporting of intimate partner homicide over time, found no instances of women’s advocacy groups or researchers being cited in newspapers in either time period. A series of reports from New Zealand also found that while media responded positively to a social media campaign raising awareness of the high rates of family violence in that country, the use of judges, police and courts as sources of information remained the norm (McGuiness, 2007-2008). One of the key recommendations from a report compiled by the End Violence Against Women Coalition in response to the Leveson Inquiry into the culture, practices and ethics of the UK press, was for a “closer working relationship between media and experts, a register of possible commentators and representatives on equality, advisory and accountability structures” (p. 14).

Only one study found that qualified sources exceeded non- qualified ones in television news coverage of femicide. Comas- d’Argemir (2014), in a study of television news broadcasts of violence against women, found several different categories of qualified sources used as experts including members of government, women’s and feminist organisations and courts of justice. The study found that, overall, more women contributed opinion in televised coverage of violence against women than men; a finding that has not been replicated anywhere else in the world.

Although Easteal et al. (2015) noted that social media has the potential to bring new voices to the public debate, we found only one study that explored contributions from “citizen journalists” (Bou-Franch, 2013). Bou-French’s analysis of 102 posts to an online news site showed the 45 contributors largely posted socially conservative ideas about violence against women. More research in this area is clearly warranted, particularly in light of the proliferation of online news sites that select stories for comment. There is a sizable gap in the literature exploring what news stories are selected by news organisations for public comment, who comments and how those voices contribute to the public discourse on violence against women.

Women’s voices

A number of studies noted a lack of women’s voices generally in news coverage of violence against women, including contributions from women who have experienced violence. Jackson (2012) identified “mediated depictions of both racism and rape remain the purview of white males” (p. 58). Some studies questioned whether the invisibility of women in stories about their experiences reflected a lack of women in media generally and more specifically a lack of women in news-making. In the study examining how the sports section of two Australian newspapers reported sexual assaults by footballers, Toffoletti (2007) found that 64 percent of articles were written by men. In the tabloid press, all stories about sexual violence against women were written by men.

A recent report on the status of women in the US media found that women were outnumbered by men in all aspects of reporting by about one to three – “whether as paid full time, freelance writers, online, in print or on air – or as citizen journalists or as non-paid commentators” (Women’s Media Centre, 2014, p.6). The report also noted that female reporters were less likely than their male counterparts to cover crime stories.

Although counts of women and men are a superficial way to understand nuances in gendered reporting, it does tell a story about who shapes decisions about media content. In Australia, a recent report on women in the media by New Matilda, an independent news website, found that the “… balance of power was tipped overwhelmingly towards men”. When the report was compiled a little over two years ago (2013), two-thirds of all Australian newspaper editors were male. Of the 16 major national and metropolitan papers, there was only one female editor. Although they noted some limitations in how data were collected, the results point to a massive gender imbalance in who shapes the news media in Australia.

Summary

We identified and reviewed a large number of primary studies on media representations of violence against women. Although studies from the US dominated the field, accounting for over one third of all research output, contributions to this review were from 16 different countries. Newspapers were the most common type of media analysed. More recent studies explored both online and printed formats. In comparison to women’s common experiences of violence, there was a disproportionate focus of studies on media representations of intimate partner homicide. Studies collected data in a number of different ways with a uniformity in analytic approaches appropriate to the field of study: the analysis of text-based media discourse.

The synthesis of contemporary national and international literature illustrated a number of key themes in the way news and information media portray violence against women. By only including those studies that were critically appraised as being rigorous, credible and relevant, we are confident that our synthesis of outcomes covers only the best available evidence. Key themes included:

• not reporting the social context in which male perpetrated violence against women occurs, perhaps unintentionally by omitting necessary information;
• sensationalising stories through language or by disproportionately focusing on violent crimes that fit key news values, but which are statistically unlikely;
• perpetuating myths and misrepresentations so as to skew public perceptions about who perpetrates violence against women, who is most at risk of violence and where violence occurs;
• directly and indirectly shifting blame from male perpetrators of violence and assigning responsibility for violence to women by focusing on the behaviour of women and their role in both the etiology and prevention of violence; and
• relying on law enforcement as the expert “voices” that inform debate in the media to the detriment of women with lived experience, advocates, service providers and researchers.

A number of recent studies, including those exploring changes over time, found a shift in reporting that included more contextual information about the underlying causes of violence, increased use of help seeking information for women, a reduction in sensationalism and direct victim blaming, along with less salacious coverage of violence against women (Comas-d’Argemir, 2014; Morgan & Politoff, 2012; O’Hara, 2012). Unfortunately, reporting on violence against women that challenges, rather than reinforces cultural and social norms about gender, was still in the minority.

Finally, it is important to acknowledge that many researchers recognised that reporting on violence against women is not an easy space for journalists to work in. Newsmakers face increasing competition to cover all the pertinent stories and reach sources before their competitors. Pressure to capture audience attention and “break” the news as soon as it occurs creates a bias towards stories that are atypical and sensational.

Results: Studies of audience reception

Overview

While the research on media representations provides strong evidence about the way news and information media portray violence against women, this body of work cannot answer questions about the effect of those messages on audiences.

We identified ten primary studies that analysed the way audiences respond to news and information media reports on violence against women, including nine peer-reviewed research papers and one unpublished report. The studies examined how audiences responded to (experimentally controlled) prompts about violence against women including emotional responses (e.g. sympathy, empathy, anger), attributions of responsibility and punishment preferences. One study explored the relationship between televised news reports and behaviour. The studies originated from a narrow set of countries, with the majority conducted in the US (n=7); the remaining three were from Belgium, Spain and the UK. Three studies included in this review of audience reception were included in the previous synthesis of studies of media representation; these articles contained two-parts relevant to both areas of inquiry (Anastasio & Costa, 2004; Franiuk, Seefelt, Cepress, et al., 2008; Franiuk, Seefelt, & Vandello, 2008).

Study characterisations including country of origin, medium, forms of violence, study design and quality appraisal ratings are included in Appendices C and D.

Medium

Six studies explored the way audiences responded to print media (newspapers). Three studies explored audience responses to televised news stories on violence against women. One study explored audience reactions via their comments posted to an online news site. One examined the relationship between news reports on Spanish television and rates of violence against women in the community.

Forms of violence

Studies of audience reception investigated how media audiences responded to various forms of violence against women. Five studies explored the issue of sexual violence. One study explored intimate partner violence-related deaths and three studies explored audience responses to intimate partner violence in general (i.e. fatal and non-fatal).

Two of the five studies exploring the way audiences respond to media coverage of sexual violence used the “real life” case of a high-profile US basketballer, Kobe Bryant. One study analysed how news frames influenced audience perceptions of perpetrator guilt. The other examined the impact of newspaper headlines only.

Study designs and data collection

Studies of audience reception were typically descriptive or experimental studies that adopted either ecological (population-based) or individual-level study designs. The majority of studies collected and analysed individual-level data whereby an outcome for individuals in one group (for example, perceived level of anger or perceived level of sympathy) was compared with individuals in another group. In these studies the outcome variable of interest was typically measured by self-report surveys. Groups were differentiated by exposure to different types of mediated information on violence against women (e.g. one group might read and respond to headlines containing “rape myths”, while the other group read and respond to headlines without “rape myths”). In most of these studies, participants comprised undergraduate university students recruited via college campuses in the US with sample sizes ranging from n=62 to n=309. One study used focus groups (n=19) to gauge audience responses to different media stimuli.

One study adopted an ecological approach and used population data to assess the relationship between televised reports of intimate partner homicide and actual incidents of intimate partner homicide occurring in the community.

Analytic approach

Depending on the design, studies used content and/or frame analysis (described previously – see Studies on media representations) or other qualitative techniques to report on focus group discussions. The analytic studies (individual- level and ecological designs) primarily used quantitative descriptive statistics to measure the association between exposure (media) and the outcome (attitudes/behaviours).

It is important to note here that measuring media effects is known to be notoriously difficult. Researchers in other content areas have argued that methodological problems are common and arise because media effects accumulate over time and interact with different social and cultural frameworks (National Cancer Institute, 2008). An additional challenge is that media is not confined to particular geographical locations or populations making it difficult to control for potential confounders. Results should be interpreted with caution.

Quality appraisal

Three studies were assessed as being of a high quality, all of which were quantitative analytic studies. They were characterised as high quality because they included a clear statement of aims, appropriate selection of groups, exposure and outcomes, consideration of bias, combined with a clearly defined analytic approach and applicable findings.

Five studies of audience reception were categorised as moderate quality; one study was categorised as low quality and is not included in the synthesis of outcomes below (Worthington, 2008).

Outcomes

There were a number of key themes evident in the literature on audience reception. Overall, audience responses were dependent on how a story was framed; the way journalists attribute responsibility or sympathy is echoed in the emotional responses of the audience. As Palazzolo and Roberto (2011) explain, “…message content can elicit emotions in predictable ways” (p. 14).

Emotional response and attribution of responsibility

Palazzolo and Roberto (2011) used Weiner’s “attribution- emotion-action model” to explore how including or excluding information on male perpetrators of violence and female “victims” influence audiences’ emotional response (specifically, anger and sympathy) and how this translated to decision- making regarding attribution of blame and punishment (Palazzolo & Roberto, 2011; Weiner, 1980). Carlyle et al. (2014) explored a similar theme by analysing how contextual information about women and men in stories of intimate partner violence correlate with feelings of sympathy.

In both studies participants were randomly assigned to read a constructed news article on intimate partner violence (Carlyle, Orr, Savage, & Babin, 2014; Palazzolo & Roberto, 2011). Palazzolo and Roberto (2011) found that participants who were exposed to information designed to increase perpetrator responsibility (i.e. he had been drinking, he had a history of domestic violence) were more likely to attribute responsibility to the perpetrator and experience feelings of anger towards the perpetrator, with strong preferences for punishment (Palazzolo & Roberto, 2011). Carlyle et al. (2014) noted that mediated information purposively designed to assign responsibility to the male perpetrator was more likely to elicit feelings of sympathy among the audience for women who experienced the violence.

On the other hand, participants who were exposed to contextual information designed to decrease perceptions of perpetrator responsibility (i.e. he had a history of mental illness and left no visible injuries) were less likely to attribute responsibility to the perpetrator, more likely to be sympathetic towards the perpetrator and less likely to prefer to place the perpetrator in prison (Carlyle et al., 2014; Palazzolo & Roberto, 2011). The researchers found that manipulating information contained within constructed newspaper articles about “victims” of intimate partner violence had the same, predictable effect: being exposed to information designed to increase perceptions of “victim” responsibility (i.e. she had been drinking, she had been having an affair) resulted in audiences responding to stories with less sympathy and increased anger towards women.

Anastasio and Costa (2004) explored the way personalising news stories influenced audience emotional responses, particularly feelings of empathy and blame. The authors found that more personalised descriptions (i.e. including names and other personal information) were associated with greater empathy and reduced the tendency to assign blame. However, this relationship was only observed when the target of the crime was female (Anastasio & Costa, 2004). Other researchers also noted that audiences tend to mimic frames presented to them in the news and information media. Brossoie et al. (2012) suggested that comments posted to an online news site were largely influenced by news content in relation to intimate partner violence among older adults. They found, for example, that stories that contained photographic images were more likely to elicit reader comments, particularly images depicting the perpetrator in a poor light (i.e. distraught and disheveled). The researchers also found that the use of descriptive language and personal quotes influenced responses; when news stories used phrases like “tragic loss” and “unfortunate situation” comments were more likely to refer to the incident as a “sad story” without reference to the case as a crime (Brossoie, Roberto, & Barrow, 2012).

Rape myths and attitudes

Two studies specifically explored the issue of “rape myths” and the way these misrepresentations influence audience attributions of responsibility (Anastasio & Costa, 2004; Franiuk, Seefelt, Cepress, et al., 2008; Franiuk, Seefelt, & Vandello, 2008). In one study, participants read one of two articles about a sexual assault case involving the US basketballer, Kobe Bryant; one story contained rape-myth endorsing information, the other rape-myth challenging information (Franiuk, Seefelt, Cepress, et al., 2008). Results showed that participants were more likely to report that Bryant was not guilty and that the woman who accused him of sexual assault was lying, after reading a rape-myth endorsing article. By comparison, participants were more likely to report that Bryant was guilty after reading an article with content that challenged stereotypes about sexual assault.

In the second study, participants were randomly assigned to read one of four different newspaper headlines: (1) newspaper headlines about the Kobe Bryant case that contained rape- myth endorsing content; (2) rape-myth endorsing headlines that didn’t identify the athlete; (3) headlines about sexual assault that did not include rape-myths; and (4) headlines unrelated to sexual assault (control group) (Franiuk, Seefelt, & Vandello, 2008). Results showed a strong gender effect. Male participants were more likely to respond that Bryant was not guilty after being exposed to headlines containing rape myths than after being exposed to headlines not containing rape myths. Female participants’ attributions of blame and responsibility were largely unaffected by headline manipulation.

The unpublished report by Tranchese (2014) supported the view that media effects are not static but interact with audience experience, gender and understandings about violence against women. Results from her focus group study, in which participants were exposed to six “real-life” newspaper articles on the rape and murder of a British woman in India, found that participants with more awareness about issues of gender-based violence tended to interpret articles based on a wider context of gender discrimination. The author suggested that less-informed media audiences might be more vulnerable to media misrepresentations of violence against women.

Behavioural outcomes

We identified only one study that examined the relationship between televised news reports of intimate partner homicide and rates of intimate partner homicide in the community. The study was a retrospective ecological study design based on 340 intimate partner violence-related deaths and 3733 televised new items covering intimate partner violence on Spanish television between 2003 and 2007 (Vives-Cases, Torrubiano-Dominguez, & Alvarez-Dardet, 2009). To explore associations between these data, the authors calculated the difference between the number of intimate partner violence-related deaths in the 7 days following a news broadcast and the number of deaths in the 7 days before a broadcast. They found that television news of intimate partner violence was associated with an increase in intimate partner homicide. However, any observed increases in intimate partner homicides were mitigated by news items featuring preventive messages or including interviews with key advocates challenging the place of gender-based violence in Spanish society. The authors acknowledged the inherent limitations of ecological study designs and that caution should be warranted in claiming that media stories about violence against women cause violence to happen in the community.

Summary

We identified and reviewed a small number of primary studies on the way media reporting on violence against women influences audience responses. Seven of the ten studies were from the US. Most studies addressed the influence of print media on audiences’ emotional responses and attributions of responsibility. Studies typically exposed groups of individuals to different types of mediated information on violence against women and used self-report surveys to assess responses. Five studies – all from the US – drew their sample from undergraduate university student populations, who may not be representative of the wider community. Only one study examined the relationship between behaviour and exposure to media coverage of violence against women.

The effect of the media on audience attitudes, emotions and actions are formed by multiple sources, which means that identifying direct relationships between media portrayals of violence against women and public opinion is not straightforward. Despite limitations, the existing literature presents some common themes including how:

• audiences’ emotional responses and attributions of responsibility are affected, and can be manipulated, by how media frames stories;
• “rape myths” impact on attitudes towards perpetrators of sexual violence;
• images and language moderate audience responses; and
• media effects intersect with audience gender and prior knowledge and understandings of violence against women.

Results: Studies of news production

Overview

There was limited peer-reviewed literature on media production. We identified only four primary studies: all published by researchers in the US, however one explored news media production in Delhi, India. Two studies explored journalist practices in producing stories about violence against women and two studies evaluated the extent to which media guidelines impact on journalism practice.

Study characterisations including country of origin, medium, forms of violence, study design and quality appraisal ratings are included in Appendix E.

Medium

The two studies on news production explored televised news media. The two evaluation studies were concerned with print media (newspapers).

Forms of violence

The two studies that explored journalistic practices in media production did so in the context of reporting on rape and sexual violence. Among the two evaluation studies, one explored the effect of guidelines on media reporting of domestic violence in general and one examined print news stories about cases in which femicide (the murder of intimate partner) was followed by perpetrator suicide.

Study designs, data collection and analytic approach

The two evaluation studies were specifically designed to determine the extent to which guidelines on reporting of violence against women impact on the behaviour of media professionals. One used a pre-post study design whereby researchers collected and analysed newspaper articles on intimate partner homicide before and after dissemination of the Rhode Island Coalition Against Domestic Violence (RICADV) handbook for journalists. The second study examined newspaper coverage of femicide-suicide cases to determine the extent to which each story adhered to a set of pre-existing suicide-reporting guidelines.

The remaining two studies were qualitative case studies, primarily focused on a narrow set of stories and were able to delve deeper into journalistic practice. Both used interviews with television news producers or journalists to explore the practice of making news, including the internal and external pressures of newsworthiness and competition.

Quality appraisal

Both evaluation studies were categorised as high quality. They each clearly stated the aims of the research, implemented an appropriate methodology to address these aims, clearly stated their findings, and are both novel studies in a limited area of research. Both studies on news production were categorised as being of moderate quality.

Outcomes

A number of interrelated themes emerged from the studies included in this review of news media production. The two studies that interviewed news makers and presenters demonstrated the types of pressures faced in the modern newsroom including increasing competition, economic imperatives and other types of constraints that may impact on best practice reporting (Rao, 2014; Worthington, 2005). These findings provide context to those studies seeking to understand the extent to which media professionals adhere to, or change practice on the basis of, best practice guidelines (Richards, Gillespie, & Givens, 2014; Ryan, Anastario, & Dacunha, 2006).

The intersection of journalistic and market interests

The main theme identified in the primary studies of media production was the competing interests of news-making and profit-making. These are not necessarily always viewed as competing priorities, as noted by Ryan et al. (2006), “Market- driven media outlets are not opposed to covering stories with social significance; they simply want to make money doing so” (p. 224).

Worthington (2005) echoed this view stating that decisions about what to include on local television broadcasts about sexual assaults on a college campus in the US were an “intersection of traditional news values and a need to earn profits in a competitive industry” (p. 12). In this study, Worthington explored how coverage of violence against women can challenge typical news reports and “successfully” meet both journalistic and feminist criteria. Worthington suggested that the inherent newsworthiness of rape stories, rather than being a negative, was one of the reasons why the producer was allowed to explore the case through a gendered lens.

The intersection of journalistic endeavour and profit making was also raised by Rao (2014) in a study exploring media ethics in India’s newly globalised, liberalised, over-saturated and ultra-competitive news media market. The qualitative study focused on the way television news reported on the rape of a woman on a bus in Delhi in 2012: a story that gained global media attention. The author conducted face-to-face interviews with 38 television news journalists and found similar themes to studies exploring media representations. For example, women’s social class (caste), rather than the severity, frequency or nature of the crime determined whether the story was considered newsworthy. The researcher also noted that in India’s highly competitive news media environment journalists may be in danger of exchanging their ethical motives for “breaking news”.

Changing practice

The two qualitative studies, described above, on the practice of news-making illustrated the pressures and constraints that journalists, editors and producers frequently face in reporting on violence against women. In an attempt to respond to the research evidence, a number of agencies have developed guidelines to promote responsible reporting. The extent to which they impact on media professionals is the focus of the final two primary studies we identified for this review of news production. Prior to describing those outcomes, we present a summary of media guidelines, including content and processes of development and implementation.

Media guidelines

Media guidelines were identified by a search of the grey literature and are presented here on the basis that they were written in English and publically accessible. This is likely not a complete list, but reflects the range of activities in this area occurring around the world.

Overview

We identified 23 different sets of guidelines on reporting violence against women in the media. Eleven were published in the US, six in Australia and the remaining six were various international guides: one from New Zealand, two from the UK, one from the European Union and one which was published by the International Federation of Journalists. The majority of guidelines specifically focused on violence against women perpetrated by a current or former intimate partner. Three media guides specifically focused on media reporting of sexual assault.

The guidelines could be broadly divided into two main types: comprehensive report-style guides that were typically longer in length at around 20 to 50 pages or fact/tip sheets, typically under ten pages in length. One of the media guides identified was a section from a commercial radio code of practice document. One media guide was presented as a website only. Overall, we identified 14 report-style media guides and nine tip/fact sheets.

Sources were excluded from this review if reporting guidelines were not the primary focus. Some of these were media backgrounders or briefing papers that might have included recommendations for media reporting best practice, but the recommendations constituted a small part of the overall content (e.g. Australian Institute of Family Studies, 2015; United Kingdom. Home Office, 2015).

Content

The contents of the media guidelines are summarised in Appendix F and show a high level of consistency. Across all the media guides, we identified ten commonly occurring themes. These themes have significant overlap with outcomes previously identified in the studies of media representation.

Provide a social context for domestic violence

Every media guideline reviewed recommended providing social context when reporting on incidents of violence against women. Common examples of social context included providing the audience with statistics on prevalence and acknowledging that violence is rarely perpetuated by a stranger. The Connecticut Coalition Against Domestic Violence (2011) recommend, “It thus is helpful to place the story within the broader context of domestic violence…” and, to “Use local, state and/or national statistics to provide a point of comparison…”.

Don’t propagate myths and misconceptions

The majority of guidelines suggested that journalists avoid using common myths or misconceptions regarding violence against women. Thirteen of the 23 media guides proposed that journalists should explicitly include that domestic violence is not a private matter. Child and Family Services (2011) recommends to:

Report consequences of family violence and highlight the fact that family violence impacts on our community regarding, for example, neighbourhood and workplace safety, the health system, economic productivity and on children.

Other myths and misconceptions the media guidelines referred to included “domestic violence is unusual, bizarre, or uncommon”:

Avoid treating domestic violence homicides and homicide- suicides as inexplicable, unpredictable tragedies. They are not. In most cases, a little digging will uncover this truth (Washington State Coalition Against Domestic Violence, 2008).

“Certain cultures or classes are more violent than others”:

…it should be noted that the country of origin of the perpetrator is irrelevant, considering that gender-based abuse is a cross-cultural phenomenon, with no geographical or cultural boundaries (Against Violence and Abuse, 2013).

And, “alcohol, drugs or stress ‘cause’ domestic violence”:

Alcohol is neither an excuse, nor a cause for domestic or sexual violence. In cases of domestic abuse for example, responsible journalists could point out that men bear the responsibility for their actions, not alcohol, and that violent men use violence towards their partners even when they are sober. (Against Violence and Abuse, 2013)

Use correct language and terminology

Eighteen guidelines suggested appropriate language and terminology for journalists to use when referring to intimate partner violence. Guidelines commonly recommended use of the term “violence” or “abuse” rather than terms such as “domestic dispute”.

Always use the term “domestic violence” when it applies. Using language like “domestic dispute”, “volatile relationship” or “bashing” minimises and trivialises a violent situation. Plus if your audience consistently comes across this term they will get a better understanding of the extent of the problem (Our Watch, 2014).

A number of guidelines also suggested that some women prefer the term “survivor” to the term “victim”, while “accuser” should be avoided:

In the case of attack that has not resulted in murder, do not use the word “victim” unless the woman self-identifies as one. If she has survived the attack, she is a “survivor” (National Union of Journalists, 2013).

There was, however, a degree of conflicting information regarding language; one media guide recommended that the term “domestic violence” should be avoided in favour of “domestic abuse” (Zero Tolerance, n.d.) and another guide suggested that using the term “violence” may result in ambiguity when referring to sexual assault:

…even the word “violence” can be a barrier to understanding… sexual violence can also include pornography, internet abuse, sexual harassment, and other forms of harm that may not necessarily result in physical injury. The offender may use threats or coercive behaviour. Sometimes the offender grooms the victim over a long period of time using kind words, seemingly supportive actions and flattery. This is manipulation that leads to harm. (Minnesota Coalition Against Sexual Assault, 2013)

Avoid victim-blaming

Most media guidelines recommended avoiding both direct and indirect victim-blaming (n=18). Examples included focusing on the dress, behaviour and occupation of women. As the UK’s media guide Against Violence and Abuse (2013) notes:

The focus on the victim’s appearance…is not only irrelevant, but it is also implying that the appearance and/ or the behaviour of a woman might be the motivating factor behind the violence.

Don’t sympathise with, or offer excuses for the perpetrator’s behaviour

The majority of media guidelines suggested that best practice is to avoid shifting blame from the perpetrator by providing excuses for their behaviour (n=18). To highlight this, UK- based media guideline Against Violence and Abuse (2013) presented an example of a report that sympathises with perpetrators in relation to a high-profile sexual assault case involving athletes in the UK: “The trauma may not yet be over for some in the England squad as they arrive back at Heathrow today…” (The Daily Telegraph (2008) in Against Violence and Abuse, 2013). Furthermore, source selection (described below) can skew the framing of the news report to sympathise with perpetrators of violence:

Many words describe Trevor Saunders. His friends use words like sweet, accepting, joking, mentor, brother… “Maybe he just had a bad moment, a mistake [when he shot and killed his ex-girlfriend, her co-worker and himself],” said a close friend of Saunders. “We’ve all made mistakes we wish we could take back…I want people to know Trevor was the kindest guy. It wasn’t like Trevor to do anything to hurt anyone”. (The Moscow-Pullman News (2005) in Connecticut Coalition Against Domestic Violence, 2011)

Never minimise or sensationalise domestic violence

Minimising and sensationalising domestic violence can both be detrimental in their own right, and were commonly referred to in the media guidelines (n=17). Examples of sensationalism included concentrating on morbid or “gory” details and/or using sensationalist headlines, for example, “Who let beast into Ireland” (Against Violence and Abuse, 2013).

Consider how source selection shapes the story

As stated previously, an important consideration for reporting violence against women that can impact the framing of the story is “source selection”. Sixteen media guidelines included information for journalists to consider how sources skew media frames. As an example, the Child and Family Services guideline includes the following recommendation:

Be aware of how source selection will shape the story. Include interviews with experts…avoid using sources that have a close relationship with the person perpetrating the violence, or sources that are not well informed about the crime and those involved…give those who have experienced family violence a voice. (Child and Family Services, 2011)

Survivor safety and dignity is priority

For non-fatal domestic violence cases, 18 sets of media guidelines discussed taking appropriate measures for protecting the safety and preserving the dignity of women. Examples included understanding disclosure and privacy laws.

Include support and resources for survivors

Most media guides suggested that best practice is to provide a range of information for women about where to seek help, including telephone helplines, websites, local emergency services, short- and long-term housing, counselling, legal and court support, health or mental health services and other locally available crisis support services: “You should always include numbers for local support services where possible” (Our Watch, 2014).

Humanise the story

A less common theme among the media guides was to “humanise” news reports (n=6) by acknowledging that the victim and perpetrator were people and not simply a “man”, “woman”, “victim” or “perpetrator”. For example:

Acknowledge that this crime has both a victim and a perpetrator. Domestic violence is sometimes reported with headlines like “Woman assaulted”, or with stories that focus only on what happened to the victim. This can make it seem like violence is something that “just happens” to women. Emphasise that someone perpetrated this violence, and that it was a crime. (Women’s Centre for Health Matters, 2014)

Development and implementation

Despite similar content, the processes and sources used to develop the guidelines differed. Eleven media guides reported consulting local or national domestic violence advocate groups or experts, ten guidelines referenced one of the other media guides. The Rhode Island Coalition Against Domestic Violence (RICADV) handbook for journalists and the Washington State Coalition Against Domestic Violence (WSCADV) guides were each listed as a key reference in five other guidelines. Other sources were media professionals, including journalists, editors, reporters and publishers (n=8), government departments (n=7), journal articles or research papers (n=5), domestic violence survivors (n=4) and legal or law enforcement resources (n=2). Four media guides either listed no references, or it wasn’t readily apparent.

The RICADV handbook for journalists stands out among the other guides as it is the only one that provided a comprehensive overview of development, implementation and dissemination, reported in a peer-reviewed article. It is also the only guide, to date, to report on whether it was effective in changing practice (described below).

Changing practice

Of the four primary studies on news production, two focused on the extent to which journalists adhered to, or changed practice on the basis of media guidelines. The first was a study conducted in the US where the researchers assessed the degree to which newspaper coverage of femicide-suicide, adhered to existing suicide reporting recommendations. (Richards et al., 2014). They found that reporters adhered to some, but not all recommendations.

Of note is the second study that focused on the development, dissemination and evaluation of the Rhode Island Coalition Against Domestic Violence (RICADV) handbook (Ryan et al., 2006). In 1996, concerned that Rhode Island news coverage was perpetuating myths and misconceptions surrounding domestic violence, the RICADV embarked on a major state-wide strategic communications plan with the aim of changing the way domestic violence was covered in the news. Part of the program included an analysis of local print media coverage from 1996 to 1999 (reviewed previously, see Studies of media representation). Findings from this component of the project, with input from the police, the judiciary, legislative, health, educational and community sectors, survivors, service providers and court personnel, informed the development of the handbook, which was then disseminated to Rhode Island journalists. Researchers found that after the introduction of the handbook, the use of the term “domestic violence” to describe instances of intimate partner homicide increased from 51 percent to 87 percent, and the presence of advocates as sources in news coverage doubled. The researchers were careful to note that these outcomes could not be attributed to the introduction of the handbook alone, because the process of developing and disseminating it served as a catalyst to facilitate ongoing discussions between domestic violence advocates and media representatives.

Summary

We identified and reviewed only a small number of primary studies on the practices of gathering and reporting on violence against women; news production. All four studies were conducted by researchers in the US. Two studies explored the views of those who make and report televised news coverage of violence against women. Findings from these studies suggest that journalists are limited by structural constraints and caught between ethical considerations, newsworthiness and profitability. Despite the studies being conducted in two very different media environments – India and the US – journalists in both countries described their work in the context of competing demands.

The two other primary studies explored the extent to which journalists adhered to, or changed practice on the basis of media guidelines. Given the number of national and international guidelines available for journalists, future research and evaluative work is clearly required in this area. While the guidelines are similar in content, emphasising, for example, that violence against women should be reported within the social context in which it occurs, they differed in their development and dissemination strategies. Only eight of the 23 media guidelines noted explicitly that media professionals were key sources in development. No media guideline explicitly described whether women with lived experience contributed to development. This paves the way for future research to explore questions around development, acceptability, uptake and effectiveness.

Discussion and conclusion

Summary of key findings

There is growing international interest in the portrayal of violence against women in the news and information media and the impact that reporting may have on community understanding, attitudes and behaviours. To date, most research attention has focused on how media messages about violence against women and their children are framed in the news. This body of work provides strong evidence about predominant patterns in the way stories are presented; namely that media offer their audiences overly-simplistic, inadequate and distorted representations of the extent and nature of the problem.

The audience reception literature demonstrated an association between representations of violence against women in the news and audience attitudes and perceptions of blame and responsibility. There is also emerging evidence of an association between televised news reports of intimate partner violence and observed rates in the community. These two complementary areas of media inquiry – representation and reception – leave no doubt that the reporting of violence against women in the news and information media should be done in a responsible manner.

The few studies available on news production confirm that the pressures of newsworthiness and profitability present formidable challenges to the task of responsible and sensitive reporting of violence against women. One way to tackle this problem is by providing journalists with resources and guidelines, of which there are now many. These guidelines emphasise the need to provide context when reporting on individual incidents of violence, as well as accurate facts about causes of violence against women that do not shift the blame from male perpetrators to women. Most highlight the desirability of providing information for women about where to seek help. The limited evaluative work suggests guidelines can lead to improvements in the quality of reporting.

Despite an expanding body of research concerning violence against women and the media, gaps in our knowledge remain. There is a need to better align the studies of media representation reviewed here with the emerging work on audience reception and news production. Fostering cross- disciplinary partnership will likely help to strengthen our understanding of the imperatives that operate on media professionals when they are preparing stories on violence against women, the contested frames which shape these stories, and the way in which these stories are interpreted by different readers and viewers.

One of other key gaps in the literature is the online news media space. The way audiences consume news has rapidly changed in the past decade: people get their news updated throughout the day often delivered via social media platforms, such as Facebook or Twitter. News and events are discussed and shared among social network site users, with some users becoming news makers themselves by posting photos or videos of news events. Meanwhile, journalists are urged to embrace multi-media reporting. Journalists increasingly use platforms, such as Twitter, to break news quickly before writing up full articles. While the research to date suggests the demand for shorter, bite sized pieces of information is less compatible with responsible and sensitive coverage of violence against women, there is an urgent need to better understand online news production, reporting and audience contribution.

Implications

The key question is whether it is possible to modify the nature of Australian news production and more specifically, the nature of reporting on violence against women and their children. On the basis of evidence presented in this state of knowledge paper we believe the answer is yes, although success may depend on how this is facilitated.

In an attempt to respond to research evidence, a number of countries have developed media guidelines to encourage responsible, sensitive and ethical reporting of violence against women. In Australia, we identified six such guides and are aware of other resources available for journalists, such as the Australian Institute of Family Studies’ (2015) Reporting on Sexual Assault “backgrounder” developed by their Sexual Violence Research program (formerly the Australian Centre for the Study of Sexual Assault). The contents of these guides are broadly shaped by the evidence available in this review. To date, however, there has been no evaluation work undertaken on the extent to which journalists are aware of and/or use guidelines in their work, let alone whether they influence reporting practices. The only international evidence available on the effectiveness of guidelines (Rhode Island Coalition Against Domestic Violence Handbook) produced positive findings but is now over ten years old.

In Australia, the Mindframe Media and Mental Health Project provides a world’s best-practice template on developing, disseminating and supporting journalists to understand and use guidelines on reporting suicide and mental health (Skehan, Hazell, & Pirkis, 2006). This body of work suggests that collaboration and media “buy-in” are vital. In the area of violence against women, there are a number of outstanding research questions including:

are “journalist-led” guidelines more likely to influence journalist practices? In what way does dissemination impact uptake? Is the best strategy for adoption exposing emerging journalists through undergraduate courses or by editors promoting them to junior staff?

There are also outstanding questions about the role that survivors of violence might play in guiding journalistic practices.

The challenge, as identified in studies of news production, is that the communication of information about violence against women as a social problem does not always fit easily with expectations around what makes news. Targeting media via prescriptive media guidelines may not be the best approach to changing mainstream media reporting. Even in the well- developed area of suicide and media, researchers have noted that media guidelines, or certain elements contained within them, are not always easy to interpret (Machlin et al., 2012), nor are they necessarily used by journalists and editors (Pirkis, Blood, Beautrais, Burgess, & Skehan, 2006). As such, it is important to explore other avenues to effect change, specifically by way of targeted dissemination.

With newsmakers facing increasing competition to cover all pertinent stories in the 24-hour news cycle, public-relations generated stories are becoming more influential in journalists’ newsgathering and reporting practices (Lewis, Williams, & Franklin, 2008). What makes the news is increasingly influenced by those outside of formal media organisations and newsrooms. The heavy reliance on law enforcement and criminal justice perspectives in stories on violence against women likely reflects the higher engagement of these sectors with news media outlets. In Australia, each of the state and territory police forces operate media communications teams to coordinate media conferences and issue press releases. In some states, media organisations can subscribe to receive daily media releases. On any one day the NSW Police Force, alone, issues more than a dozen press releases (www.police. nsw.gov.au). In Victoria, the police have a website dedicated to breaking news with stories and images accessible via Facebook, Twitter, YouTube and Instagram (www.vicpolicenews.com. au). Police agencies provide a large and accessible output on crime – of which violence against women is one.

To counter this “police frame” it is important to increase output of media releases by other agencies – agencies that can provide expert input so that the news no longer deals with each case as just one more in a long line of similar events, but as an indicator of a broader social problem. Violence against women is a topic that journalists are increasingly required to cover, so there are key advantages of media advocacy. News and information media are tools that should be better utilised by advocates, services providers and researchers.

Limitations of the review

This state of knowledge paper was deliberately focused on portrayals of violence against women in the news and information media. We did not search for or include any primary studies on violence against women in entertainment media, including in magazines, films, television, music, plays and video games. Studies considering the way entertainment media depict violence against women would be an important accompaniment to this review. Unlike the international interest in mediated messages in the news, understanding media effects and intervening to promote responsible depictions of violence against women in entertainment media is not well advanced.

We also did not specifically search for media representations of some forms of violence against women and their children, such as forced marriage, female trafficking and prostitution, female genital cutting/mutilation, or “honor” crimes. While these forms of violence represent important contexts in which women experience violence, especially considering Australia’s large culturally and linguistically diverse communities, evidence suggests that these forms of violence are less common in Australia than those addressed in this paper. With this in mind, we did not exclude any articles focusing on these forms of violence, but did not specifically include them in our search strategy. Similarly, we didn’t explicitly search for studies exploring sexual harassment. While we acknowledge that being subjected to sexually harassing behaviours is a particularly common experience for women, it is also very complex to define, with a great deal of variation or inconsistency in behaviours that constitute sexual harassment. There are also a number of recent reviews that specifically deal with media depictions of sexual harassment (Easteal, Holland, et al., 2015; Judd & Easteal, 2013).

Similarly, we also note the absence of literature focusing on media representation, production and audience reception as related to culturally and linguistically diverse groups, or Aboriginal and/or Torres Strait Islander communities. Future research is needed to examine how diverse cultural communities interpret and engage with media reporting on violence against women generally and on violence against women from their own communities. There is little evidence to inform the accuracy and standards of media reporting of violence against women from diverse groups, and the impact of media reporting on the dynamics of different cultural communities is unclear. While the Australian Muslim Women’s Centre for Human Rights recently published a guide for media professionals reporting on female genital cutting, there is no equivalent in relation to guidelines for reporting violence against migrant and refugee women in Australia.

Conclusion

In today’s heavily mediated society, all forms of media have the ability to reach increasingly more people than ever before. The study on media coverage of the rape of a woman on a bus in Delhi in 2012 showed how quickly a local story can become global. News of the incident spread globally through professional media sources within just 2 days (Phillips et al., 2015). The story resonated in communities far from India showing the unpreceded capacity of today’s media to influence public perceptions that violence against women is neither exceptional nor acceptable.

While many factors are known to influence public perceptions about violence against women, in this state of knowledge paper we examined the role of news media– a dominant force in shaping the discourse on matters of public importance. We found that media representation studies dominate research in the area. Collectively, these studies illustrate that the media frequently mirrors society’s confusion and ambivalence about violence against women. Although the link between media reporting and behaviour is not well-established, studies of audience influence show that media can play a role in dispelling myths and reinforcing information about the true nature and extent of the problem.

References

Against Violence and Abuse. (2013). Guidelines for accurately covering male violence against women. London, UK: AVA.

Ahmed, S. (2014). Violence against women: Media representations of violent issues in the perspective of Pakistan. Science International, 26(1), 367-371.

Alat, Z. (2006). News coverage of violence against women: The Turkish case. Feminist Media Studies, 6(3), 295-314, doi: 10.1080/14680770600802041

Anastasio, P. A., & Costa, D. M. (2004). Twice hurt: How newspaper coverage may reduce empathy and engender blame for female victims of crime. Sex Roles, 51(9-10), 535-542. doi: 10.1007/s11199-004-5463-7

Australia. Department of Social Services. (2014). Second action plan 2013-2016 of the national plan to reduce violence against women and their children 2010-2022. Canberra: DSS.

Australian Bureau of Statistics. (2012). Personal Safety Survey 2012. Canberra: ABS.

Australian Institute of Family Studies. (2015). Reporting on sexual assault: Media backgrounder.Retrieved from http://www3.aifs.gov.au/acssa/media/index.html

Barnett, B. (2012). How newspapers frame rape allegations: The Duke University case. Women and Language, 35(2), 11-33.

Blood, R. W., Putnis, P., & Pirkis, J. (2002). Mental illness news as violence: A news frame analysis of the reporting and portrayal of mental health and illness in Australia media. Australian Journal of Communication, 29, 59-82.

Bonnes, S. (2013). Gender and racial stereotyping in rape coverage: An analysis of rape coverage in a South African newspaper, Grocott’s Mail. Feminist Media Studies, 13(2), 208-227. doi: 10.1080/14680777.2011.623170

Bou-Franch, P. (2013). Domestic violence and public participation in the media: The case of citizen journalism. Gender and Language, 7(3), 275-302. doi: 10.1558/genl. v7i3.275

Brofenbrenner, U. (1977). Toward an experimental ecology of human development. American Psychologist, 32(7), 513-531.

Brossoie, N., Roberto, K. A., & Barrow, K. M. (2012). Making sense of intimate partner violence in late life: Comments from online news readers.Gerontologist, 52(6), 792-801. doi: http://dx.doi.org/10.1093/geront/gns046

Brownmiller, S. (1975). Against our will: Men, women and rape. New York: Simon & Schuster.

Bullock, C. F. (2010). Framing domestic violence fatalities: Coverage by Utah newspapers. Women’s Studies in Communication, 30(1), 34-63.

Bullock, C. F., & Cubert, J. (2002). Coverage of domestic violence fatalities by newspapers in Washington state. Journal of Interpersonal Violence, 17(5), 475-499. doi: 10.1177/0886260502017005001

Burt, M. R. (1980). Cultural myths and support for rape. Journal of Personality and Social Psychology, 38, 217-230.

Carll, E. K. (2003). News portrayal of violence and women: Implications for public policy. American Behavioral Scientist, 46(12), 1601-1610. doi: 10.1177/0002764203254616

Carlyle, K. E., Orr, C., Savage, M. W., & Babin, E. A. (2014). News coverage of intimate partner violence: Impact on prosocial responses. Media Psychology, 17(4), 451-471. doi: 10.1080/15213269.2014.931812

Carlyle, K. E., Slater, M. D., & Chakroff, J. L. (2008). Newspaper coverage of intimate partner violence: Skewing representations of risk. Journal of Communication, 58(1), 168-186. doi: 10.1111/j.1460-2466.2007.00379.x

Child and Family Services. (2011). Family violence in the news: a media toolkit. Ballarat, VIC: Child and Family Services Ballarat.

Chung, D. (2013). Understanding the statistics about male violence against women. Sydney: White Ribbon Research Series.

Comas-d’Argemir, D. (2014). News of partner femicides: The shift from private issue to public problem. European Journal of Communications, 30(2), 121-136.

Connecticut Coalition Against Domestic Violence. (2011). Reporting on domestic violence in Connecticut: A guide for media. Wethersfield, CT: CCADV.

Council of Australian Governments. (2011). The national plan to reduce violence against women and their children: Including the first three-year action plan. Canberra: FAHCSIA.

Cripps, K., & Davis, M. (2012). Communities working to reduce Indigenous family violence. Sydney: Indigenous Justice Clearinghouse.

Department of Families Housing Community Services and Indigenous Affairs. (2009). Domestic violence laws in Australia. Canberra: FAHCSIA.

Domestic Violence Death Review Team. (2015). A report of the NSW domestic violence death review team. Sydney: NSW Government Department Attorney General and Justice.

Dwyer, T., Easteal, P., & Hopkins, A. (2012). Did she consent? Law and the media in New South Wales. Alternative Law Journal, 37(4), 249-253.

Easteal, P., Bartels, L., Nelson, N., & Holland, K. (2015). How are women who kill protrayed in newspaper media? Connections with social values and the legal system. Women’s Studies International Forum, 51, 31-41.

Easteal, P., Holland, K., & Judd, K. (2015). Enduring themes and silences in media portrayals of violence against women. Women’s Studies International Forum, 48, 103-113. doi: 10.1016/j.wsif.2014.10.015

End Violence Against Women Coalition. (2012). Just the women. London, UK: End Violence Against Women.

Evans, I. (2001). Desperate lovers and wanton women: Press representations of domestic violence. Hecate, 27(2), 147.

Exner, D., & Thurston, W. E. (2009). Understanding ‘passion killings’ in Botswana: An investigation of media framing. Journal of International Women’s Studies, 10(4), 1-16.

Fairbairn, J., & Dawson, M. (2013). Canadian news coverage of intimate partner homicide: Analyzing changes over time. Feminist Criminology, 8(3), 147-176. doi: 10.1177/1557085113480824

Family Violence Act 2004 (Tas). Retrieved from http:// www5.austlii.edu.au/au/legis/tas/consol_act/fva2004158/

Family Violence Protection Act 2008 (Vic). Retrieved from http://www.austlii.edu.au/au/legis/vic/consol_act/ fvpa2008283/

Flood, M., & Pease, B. (2009). Factors influencing attitudes to violence against women. Trauma, Violence, & Abuse, 10(2), 125-142.

Franiuk, R., Seefelt, J. L., Cepress, S. L., & Vandello, J. A.(2008). Prevalence and effects of rape myths in printjournalism: The Kobe Bryant case. Violence Against Women, 14(3), 287-309.

Franiuk, 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. doi: http:// dx.doi.org/10.1007/s11199-007-9372-4

Gillespie, L. K., Richards, T. N., Givens, E. M., & Smith, M. D. (2013). Framing deadly domestic violence: Why the media’s spin matters in newspaper coverage of femicide. Violence Against Women, 19(2), 222-245. doi: http://dx.doi. org/10.1177/1077801213476457

Grabe, M. E., Zhou, S., & Barnett, B. (2001). Explicating sensationalism in television news: Content and the bells and whistles of form. Journal of Broadcasting and Electronic Media, 45(4), 635-655.

Halim, S., & Meyers, M. (2010). News coverage of violence against Muslim women: A view from the Arabian Gulf. Communication, Culture & Critique, 3(1), 85-104. doi: 10.1111/j.1753-9137.2009.01059.x

Harries, G., & Bird, W. (2005). Keeping an eye on the campaign: Monitoring media coverage of the 16 days of activism: No violence against women and children campaign. Agenda, 66, 90-97.

Holt, S., Buckley, H., & Whelan, S. (2008). The impact of exposure to domestic violence on children and young people: A review of literature. Child Abuse and Neglect, 32(8), 797-810.

Iyengar, S., & Simon, A. (1993). News coverage of the Gulf Crisis and public opinion: A study of agenda-setting, priming, and framing. Communication Research, 20, 365-383.

Jackson, S. J. (2013). Framing Megan Williams: Intersecting discourses of race, class, and gender in television news coverage of racialized rape. Feminist Media Studies, 13(1), 46-63. doi: 10.1080/14680777.2011.647970

Judd, K., & Easteal, P. (2013). Media reportage of sexual harrassment: The credible complainant. Denning Law Journal, 25(1), 159-180.

Korn, A., & Efrat, S. (2004). The coverage of rape in the Israeli popular press. Violence Against Women, 10(9), 1056-doi: 10.1177/1077801204267378

Lewis, J., Williams, A., & Franklin, B. (2008). Four rumours and an explanation. Journalism Practice, 2(1), 27-45.

Lindsay-Brisbin, J., DePrince, A. P., & Welton-Mitchell, C.(2014). Missed opportunities: Newspaper reports ofdomestic violence. Journal of Aggression, Maltreatment & Trauma, 23(4), 383-399.

Machlin, A., Skehan, J., Sweet, M., Wake, A., Fletcher, J., Spittal, M., & Pirkis, J. (2012). Reporting suicide: Interpreting media guidelines. Australian Journalism Review, 34(2), 45-56.

Marhia, N. (2008). Just representation? Press reporting and the reality of rape. London, UK: EAVES.

Mason, P., & Monckton-Smith, J. (2008). Conflation, collocation and confusion: British press coverage of the sexual murder of women. Journalism, 9(6), 691-710. doi: 10.1177/1464884908096241

McGuiness, M. (2007-2008). Quarterly media analysis of family violence reporting. Wellington: McGuiness Institute.

McManus, J., & Dorfman, L. (2005). Functional truth or sexist distortion? Assessing a feminist critique of intimate violence reporting. Journalism, 6(1), 43-65. doi: 10.1177/1464884905048952

Minnesota Coalition Against Sexual Assault. (2013). Reporting on sexual violence: A guide for journalists. St. Paul, MN: MCASA.

Moffitt, P., Fikowski, H., Mauricio, M., & Mackenzie, A. (2013). Intimate partner violence in the Canadian territorial north: Perspectives from a literature review and a media watch. International Journal of Circumpolar Health, 72. doi: http://dx.doi.org/10.3402/ijch.v72i0.21209

Mogashoa, T. (2014). Understanding critical discourse analysis in qualitative research. International Journal of Humanities Social Sciences and Education, 1(7), 104-113.

Morgan, J., & Politoff, V. (2012). Victorian print media coverage of violence against women: A longitudinal study. Melbourne: VicHealth.

National Cancer Institute. (2008). The role of the media in promoting and reducing tobacco use (Vol. NIH Pub. No. 07-6242). Bethesda, MD: U.S Department of Health andHuman Services, National Institutes of Health.

National Council to Reduce Violence against Women and their Children. (2009). The cost of violence against women and their children. Canberra: FAHCSIA.

National Union of Journalists. (2013). NUJ guidelines for journalists on violence against women. London, UK: NUJ.

Naylor, B. (2001). Reporting violence in the British print media: Gendered stories. Howard Journal of Criminal Justice, 40(2), 180-194.

Noh, M. S., Lee, M. T., & Feltey, K. M. (2010). Mad, bad, or reasonable? Newspaper portrayals of the battered woman who kills. Gender Issues, 27(3-4), 110-130. doi: 10.1007/ s12147-010-9093-9

O’Hara, S. (2012). Monsters, playboys, virgins and whores: Rape myths in the news media’s coverage of sexual violence. Language and Literature, 21(3), 247-259. doi: 10.1177/0963947012444217

Our Watch. (2014). Reporting on domestic violence. Melbourne: Our Watch.

Oxman-Martinez, J., Marinescu, V., & Bohard, I. (2009). Shades of violence: The media role. Women’s Studies International Forum, 32(4), 296-304. doi: 10.1016/j. wsif.2009.05.008

Palazzolo, K. E., & Roberto, A. J. (2011). Media representations of intimate partner violence and punishment preferences: Exploring the role of attributions and emotions. Journal of Applied Communication Research, 39(1), 1-18. doi: 10.1080/00909882.2010.536843

Pan, Z., & Kosicki, G. M. (1993). Framing analysis: An approach to news discourse. Political Communication, 10(1), 55-75.

Phillips, M., Mostofian, F., Jetly, R., Puthukudy, N., Madden, K., & Bhandari, M. (2015). Media coverage of violence against women in India: A systematic study of a high profile rape case. BMC Women’s Health, 15. doi: http:// dx.doi.org/10.1186/s12905-015-0161-x

Pirkis, J., Blood, R. W., Beautrais, A., Burgess, P., & Skehan, J.(2006). Media guidelines on the reporting on suicide. Crisis, 27(2), 82-87.

Quadara, A., & Wall, L. (2012). What is effective primary prevention of sexual assault? Translating evidence for action. Melbourne: Australia Centre for the Study of Sexual Assault.

Rao, S. (2014). Covering rape in shame culture: Studying journalism ethics in India’s new television news media. Journal of Mass Media Ethics: Exploring Questions of Media Morality, 29(3), 153-167. doi: 10.1080/08900523.2014.918497

Restraining Orders Act 1997 (WA). Retrieved from http:// www.slp.wa.gov.au/legislation/statutes.nsf/default.html

Richards, T. N., Gillespie, L., & Smith, M. D. (2011). Exploring news coverage of femicide: Does reporting the news add insult to injury? Feminist Criminology, 6(3), 178-202. doi: 10.1177/1557085111409919

Richards, T. N., Gillespie, L. K., & Givens, E. M. (2014). Reporting femicide-suicide in the news: The current utilization of suicide reporting guidelines and recommendations for the future. Journal of Family Violence, 29(4), 453-463. doi: 10.1007/s10896-014-9590-9

Richards, T. N., Gillespie, L. K., & Smith, M. D. (2013). An examination of the media portrayal of femicide-suicides: An exploratory frame analysis. Feminist Criminology, 9(1), 24-44. doi: 10.1177/1557085113501221

Roberto, K. A., McCann, B. R., & Brossoie, N. (2013). Intimate partner violence in late life: An analysis of national news reports. Journal of Elder Abuse and Neglect, 25(3), 230-241. doi: 10.1080/08946566.2012.751825

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. doi: 10.1177/0886260505282285

Sampert, S. (2010). Let me tell you a story: English-Canadian newspapers and sexual assault myths. Canadian Journal of Women and the Law, 22(2), 301-328. doi: 10.3138/ cjwl.22.2.301

Saroca, C. (2013). The absent and silenced voice in media representations of Filipina victims of homicide in Australia. South East Asia Research, 21(3), 517-543. doi: 10.5367/ sear.2013.0161

Scheufele, D. A. (1999). Framing as a theory of media effects. Journal of Communication, 49(1), 103-122.

Skehan, J., Hazell, T., & Pirkis, J. (2006). Reach, awareness and uptake of media guidelines for reporting suicide and mental illness: An Australian perspective. International Journal of Mental Health Promotion, 8(4), 28-34.

Skilbrei, M. L. (2013). Sisters in crime: Representations of gender and class in the media coverage and court proceedings of the triple homicide at Orderud Farm. Crime, Media, Culture, 9(2), 136-152. doi: 10.1177/1741659012461040

Stemler, S. (2001). An overview of content analysis. Practical Assessment, Research & Evaluation, 17(7).

Sweeney, M. (2012). Listening rhetorically to textual silence: Intimate partner homicide media coverage. International Journal of Listening, 26(3), 146-158. doi: 10.1080/10904018.2012.712472

Taylor, R. (2009). Slain and slandered: A content analysis of the portrayal of femicide in crime news. Homicide Studies, 13(1), 21-49. doi: 10.1177/1088767908326679

Toffoletti, K. (2007). How is gender-based violence covered in the sporting news? An account of the Australian Football League sex scandal. Women’s Studies International Forum, 30(5), 427-438. doi: 10.1016/j.wsif.2007.07.004

Tranchese, A., & Zollo, S. A. (2013). The construction of gender-based violence on the British printed and broadcast media. Critical Approaches to Discourse Analysis Across Disciplines, 7(1), 141-163.

UNICEF. (2000). Domestic violence against women and girls. Florence, Italy: UNICEF.

United Kingdom. Home Office. (2015). Violence against women and girls: communications insight pack. UK: Her Majesty’s Government.

United Nations. (1993). Declaration on the Elimination of Violence against Women (UN Resolution 48/104). Paper presented at the proceedings of the 85th plenary meeting, United Nations General Assembly, Geneva.

VicHealth. (2007). Preventing violence before it occurs: A framework and background paper to guide the primary prevention of violence against women in Victoria. Melbourne: Victorian Health Promotion Foundation.

Vives-Cases, C., Torrubiano-Dominguez, J., & Alvarez- Dardet, C. (2009). The effect of television news items on intimate partner violence murders. European Journal of Public Health, 19(6), 592-596. doi: http://dx.doi.org/10.1093/ eurpub/ckp086

Washington State Coalition Against Domestic Violence. (2008). Covering domestic violence: a guide for journalists and other media professionals. Seattle, WA: WSCADV.

Waterhouse-Watson, D. (2012). Framing the victim: Sexual assault and Australian footballers on television. Australian Feminist Studies, 27(71), 55-70. doi: 10.1080/08164649.2012.648260

Wathen, C., & MacMillan, H. L. (2003). Interventions for violence against women: Scientific review. JAMA, 289(5), 589-600. doi: 10.1001/jama.289.5.589

Webster, K., Pennay, P., Bricknall, R., Diemer, K., Flood, M., Powell, A., . . . Ward, A. (2014). Australians’ attitudes towards violence against women: Findings from the 2013 National Community Attitudes Towards Violence Against Women Survey (NCAS). Melbourne: Victorian Health Promotion Foundation.

Weiner, B. (1980). A cognitive (attribution)-emotion-action model of motivated behavior: An analysis of judgments of help-giving. Journal of Personality and Social Psychology, 39(2), 186-200. doi: 10.1037//0022-3514.39.2.186

Women’s Centre for Health Matters. (2014). Reporting on domestic violence: A guide for ACT media. Canberra: WCHM.

Women’s Media Centre. (2014). The status of women in the US media 2014. New York: WMC.

World Health Organization. (2013). Global and regional estimates of violence against women: Prevalence and health effects of intimate partner violence and nonpartner sexual violence. Geneva: WHO.

Worthington, N. (2005). Negotiating news representations of rape: Reporting on a college sexual assault scandal. Media Report to Women, 33(4), 6-14.

Worthington, N. (2008). Encoding and decoding rape news: How progressive reporting inverts textual orientations. Women’s Studies in Communication, 31(3), 344-367.

Wozniak, J. A., & McCloskey, K. A. (2010). Fact or fiction? Gender issues related to newspaper reports of intimate partner homicide. Violence Against Women, 16(8), 934- 952.doi: http://dx.doi.org/10.1177/1077801210375977

Zero Tolerance (n.d.). Handle with care: A guide to responsible media reporting of violence against women. Edinburgh, Scotland: Zero Tolerance.

Appendix A

Primary studies of media representations of violence against women

Appendix B

Unpublished reports of media representations of violence against women

Appendix C

Primary studies of audience reception related to violence against women

Appendix D

Unpublished reports of audience reception related to violence against women

Appendix E

Primary studies of media production related to violence against women

Appendix F

International media guidelines

Endnotes

1. Note that an asterisk indicates a truncation operator so that the search engine will search for all terms beginning with the preceding string of text. For example for the search term televis*, the search engine will return results for televise, television, televisual etc. ↑

2. Note that indicates that the media guide recommends practice that contradicts recommendations from other media guides ↑