Online abuse of women: an interdisciplinary scoping review of the literature

ABSTRACT This paper outlines the interdisciplinary scoping review undertaken into the online abuse of women. The review assesses literature published between 2000 and 2020, using a scoping review methodology, to discover how online abuse is defined, and recommendations proposed to ameliorate it. Using five databases and a selection of keywords generated 61 studies meeting the criteria. The majority defined online abuse as a structural issue (52/61). This paper determines that contemporary literature describes online abuse as a manifestation of gender-based violence, also evidencing the role of misogyny. The paper concludes that tackling online abuse requires action at three levels—the personal, the organizational, and the societal. Furthermore, utilizing the scoping review method to assess the literature identifies innovative multi-disciplinary solutions to a complex issue.


Introduction
Online abuse, as communicated via social networking sites (SNS), has increased considerably in recent years (Fiona Vera-Gray 2017), and criminal justice agencies have been slow to respond to the immense change that has occurred as a result of the way that individuals interact in the digital space (Emma A. Jane 2017a).The emerging nature of the phenomena has created a policy vacuum, suggesting that a lag in institutional responsiveness leaves victims without adequate protection or recourse (Jane 2017a).
The scoping review presented here demonstrates that the online abuse directed at women in public facing occupations is misogynistic, frequently includes violent threats, and dismisses women's contributions to online discussions.Ultimately, online abuse is not about image, political opinions, religious beliefs or sexual orientation-it is about gender and is the consequence of being a woman on the internet.
This scoping review analyses the literature on the gendered online abuse of women published between 2000 and 2020.This time period was chosen as it dovetails with the introduction of "Web 2.0," and the introduction of social media platforms such as Facebook and Twitter.Whilst older literature exploring the nature of computer mediated communication (CMC) offers broadly individualized explanations for malign interaction,

Background
Scholarly investigation into the nature of technological communication emerged in the 1980s, straddling the disciplines of social psychology, culture and commerce (Emma A. Jane 2015).This assemblage determined that abuse in online communication was insignificant, infrequent, or entertaining (Jane 2015).This changed with the introduction of "Web 2.0," the dramatic leap in technological development that occurred in 2004.Web 2.0 enabled the mass participatory collaborations (Grant Blank and Bianca C. Reisdorf 2012) that facilitated the introduction of SNS such as Twitter, Linked In and Facebook (Bridgette Wessels 2009).The introduction of Web 2.0 led to a significant increase in the academic interest in CMC.
Online abuse encompasses a range of behaviors that targets strangers on the internet for harassment (Megan Todd 2017).Gendered online abuse frequently consists of seven key elements, which appear in whole or in part, within every abusive encounter.These seven key elements are defamation and/or libel, emotional harm and harassment, and threat (Soraya Chemalyand Anna Louie Sussman 2020); along with a silencing of women's voices, the belittling and undermining of women in a professional context, and criticism of their appearance, age or other physical characteristics (Emma Louise Backe, Pamela Lilleston, and Jennifer McCleary-Sills 2018;Jaigris Hodson et al. 2018;Mona Lena Krook 2020;Ruth Lewis, Michael Rowe, and Clare Wiper 2017;Karla Mantilla 2015; Alice E. Marwick and Ross Miller 2014;Sarah Sobieraj 2018).When viewed together, these seven elements provide a representation of and explanation for online abuse directed at women in public facing occupations.
Online abuse has been facilitated by the introduction of the technologies that facilitate direct access to politicians (David S. Wall 2007) and others working in public facing occupations.Harassment of this nature can compromise the civil rights of those targeted and has emotional, financial, physical, and professional impacts.The often-criminal nature of these communications brings online abuse within the realm of cybercrime and genderbased violence, although this categorization has frequently failed to be reflected in the development of policy.
The literature on policymaking around gender-based violence analyses the numerous initiatives and legislation that has been promulgated.Scrutiny of the corpus reveals a wide range of policy choices, all of which profess to tackle gendered abuse.Such calls for action have frequently been introduced piecemeal, often emerging in response to public pressure (e.g., Jonathon W. Penney 2020;Cat Zakrzewski 2020;Mantilla 2015); or, as Sylvia Walby, Jude Towers, and Brian Francis (2014, 188) have vividly described (with reference to Jimmy Saville and Dominic Kahn), as a response "to the violence that emerges into public view in the form of 'scandals,' when some famous man is accused of perpetrating gendered violence." Work by Cynthia Southworth et al. (2007) provides evidence of the use of technology in gender-based violence, particularly in relation to stalking.Other examples of genderbased violence that would not exist without the medium of technology include image based sexual abuse-where intimate images are circulated without consent (Michael.Salter 2017); and "technology facilitated sexual violence" (Nicola Henry and Anastasia Powell 2015, 105).There is also clear evidence of technology being used in offences of coercive control, whether in the installation of tracking devices on victims' mobile phones or enabling remote access to home computers (Cassandra Wiener 2017); or by removing victims' access to the technology that facilitates relationships with others, such as text messaging and access to SNS (Mylène Fernet et al. 2019).
In England and Wales, there remains no structural approach to addressing the gendered violence perpetrated online.A pertinent example is provided by the Domestic Abuse Bill, which became law in April 2021 (Home Office 2021).Whilst this legislation provided the first ever legal definition of the offence of domestic abuse (Carolyn Stephens et al. 2021), there is no mention of either technology facilitated sexual violence (Henry and Powell 2015), or technology facilitated coercive control (Molly Dragiewicz et al. 2018) within the Act.
Growing concerns about the safety of women and girls in the physical space, particularly following the murders of Sarah Everard and Sabina Nessa in London in 2021 (Heidi Stöckl and Zara Quigg 2021); alongside the increase in domestic homicide during multiple COVID-19 lockdowns (Hannah I. Rochford, Kaleb Brooks, Mark Berg and Cori Peek-Asa 2021) have led to calls for a wider public discussion about the impact of male violence (e.g., Julia R. DeCook and Megan Kelly 2021; Kerry Dungay 2021; Jennifer Grant 2021; Irene Zempi and Jo Smith 2021), in both the physical world and the online space.
The paucity of academic research into online abuse has undoubtedly delayed the development of policy to tackle the issue (Charlotte Barlow and Imran Awan 2016) and has contributed to the languid introduction of legislation in this area (Claire Hardaker and Mark McGlashan 2015), as demonstrated by the absence of online abuse from the Domestic Abuse Bill (2021).Legislation as evidenced in the form of Acts of Parliament is important, as it provides key definitions of offences that can be operationalized by criminal justice agencies, whilst in the longer term changing social norms, and ultimately containing the potential to "transform online subcultures of misogyny to those of equality" (Danielle.K. Citron 2009, 404).However, as Citron (2009) confirms, the failure of institutions to take decisive action in this area sends a message to those engaged in acts of online violence and abuse that such behavior is trivial, and will not be investigated robustly, whilst simultaneously signaling to women that their abuse will be ignored.In this way, online abuse both perpetuates the inadequate treatment of gendered violence witnessed in the physical space, and also allows technology to act as an amplifier of misogynistic abuse.
Explanations for this lackadaisical approach can be traced back to the advent of CMC.There was a belief that the introduction of new communication platforms would foster notions of freedom of speech and enable participation (Jack M. Balkin 2004), with some advocating complete anarchy online, with no constraints on behavior (Susan C. Herring et al. 2002).This has sustained an enduring commitment to libertarianism from many advocates of emerging technologies (Fred Turner2010).The preponderance of libertarian beliefs within CMC has had a particular impact on online behavior (Emma A. Jane 2014).This is most evident in "flaming" -the "hostile and aggressive interactions via text-based computer mediated communication" (Patrick B. O'sullivan and Andrew J. Flanagin 2003, 69).Research from this period advocates that instances of "flaming" serve to build group identity, are infrequent, or are a source of entertainment (Jane 2015).
The motivations underpinning the decision to engage in "flaming" have been subject to numerous interpretations from multiple academic disciplines.This scoping review reveals that there has been a significant change in the tone and nature of online abuse over the last fifteen years.The expansion in the use of online platforms instigated by Web 2.0 has seen abuse increase exponentially (Seohee Sohn, Ho Chung, and Namkee Park 2019), whilst such behavior has moved from a discrete online activity, to a more generalized verbal violence that targets individuals' personal or occupational life (Jane 2015).Furthermore, there is evidence that "flaming," now commonly described as "online abuse," has become increasingly vituperative (Antigoni-Maria Founta et al. 2019), containing misogynistic condemnation, threats and descriptions of sexual violence.
Nevertheless, a key benefit correlated with the widespread adoption of online technologies is the opportunity to bring opinion formers and members of the public together (Wessels 2009).This development extends notions of the public sphere, and the concept that there is a space between society and the state where the public can organize, and opinion be formed (Chris Barker and Emma A. Jane 2016).There was a hope that an increase in online platforms would advance women's equality, with the internet "hailed as a place where offline prejudices and abuse could be negated and destroyed" (Bailey Poland 2016, 159).Whilst it is true that SNS have offered feminist and other campaigning groups an effective platform from which to develop their aims (Sarah Banet-Weiser 2018; R. Weathers Melinda and C. Hopson Mark 2015), this has come at the cost of online abuse.When considering the impact of online abuse on women's involvement in the public sphere, there is evidence that the incessant deluge of abuse is causing women to withdraw from the public arena (Susan Watson 2019).Many believe that the majority of violent and aggressive online abuse is received by women (e.g., Simin Kargar and Adrian Rauchfleisch 2019).On the rare occasion that men are the targets of online abuse, this abuse is in itself gendered, with attacks focused on the women relatives of targets (Jane 2014).The issue of intersectionality is crucial, with women often members of multiple demographic groupings that frequently intersect, creating myriad levels of discrimination (Linda J. Beckman 2014).There is evidence that women aged 18 to 29 are most likely to be the target of online abuse (Amanda Lenhart et al. 2016), along with women of color (Mona Lena Krook 2017) and lesbians (Becky Gardiner 2018).It is clear that the advancements made by women in public facing occupations (who have fought to challenge traditional hierarchies), have been made predominantly by white, middle-class women, who remain over-represented (Chrysa Lamprinakou et al. 2016;Fiona Mackay 2004).
The body of academic work investigating gendered abuse in CMC remains patchy, with the vacuum often filled by journalists (e.g., Hannah Cranston 2015).The academic literature that does attempt to investigate the phenomenon of sexualized, gendered violence occurring online (Jane 2017a) continues to straddle numerous disciplines, and consequently, obtaining an overview of the corpus has previously proven challenging.However, subjecting the literature to a scoping review enables a rigorous interdisciplinary analysis, whilst also demonstrating how the volume of investigation has increased over time.
The scoping review presented here is narrative, adapting the existing methods literature (e.g., Angela M. Boland, Gemma Cherry, and Rumona Dickson 2017;Colin Robson and Kieran McCartan 2016;Hilary Arksey and Lisa O'Malley 2005) to ensure that the process is sufficiently flexible to ensure that evidence can be extracted, whilst maintaining a high degree of methodological rigor.

Devising the research question
This review follows the York Framework (Susan E. Brien et al. 2010) for scoping reviews (Arksey and O'malley 2005).This framework provides the stages to be followed to ensure that the review process is robust (Brien et al. 2010).The first stage was to define the research question.The question posed to guide this review was: The decision to focus attention solely on the experiences of women is worthy of further explanation.Whilst there has been research focusing on the online experiences of both men and women serving in public facing occupations (e.g., Stephen Ward and Liam McLoughlin 2020; Amy Binns 2017), there is a robust rationale for excluding men from this research.Firstly, women's online experiences are overwhelmingly underpinned by misogyny, violence and threat (Poland 2016), which is reinforced by a consistent underestimation of the scale of gender-based violence from the academy (Saifuddin Ahmedand Dani Madrid-Morales 2021; Walby, Towers, and Francis 2014).Secondly, there is a clear link between online abuse, gender-based violence (Salter 2017), and the misogynistic aim to silence women's contributions in the public sphere (Mantilla 2015).Whilst men in public sphere occupations may also experience threats of violence, they do not do so because they are men.Therefore to include men in this sample would risk creating a "false symmetry between men's and women's experiences" (Krook 2020, 107).Furthermore, a gendered comparison of experience risks drawing misleading conclusions, whilst also ignoring the risks navigated by women (Karen Lumsden and Heather M. Morgan 2018).
Once the research question had been identified, it was possible to devise a search strategy to employ.The keywords and databases used in the scoping review are provided at Figure 1.
A data extraction table was maintained, in order to record the different stages of the search process.A copy of this is provided at Table 1.

Results
The initial search of the selected databases and additional journals yielded a total of 3672 papers.This was reduced to 3551 once duplicates and articles not written in English were removed.Whilst the decision to exclude journal articles written in other languages risks losing insights made by scholars in other countries, it was unavoidable given language and resource constraints.These 3551 articles were then subjected to further analysis using the first tranche of inclusion criteria, which specified that articles should be published between January 1 2000 and December 31 2019; study an adult research population (18 +); and must investigate CMC or online abuse.This reduced the number of papers requiring further analysis to 533.At this point, the abstracts of the 533 papers were read and subjected to a further set of selection criteria.This revised criterion stipulated that the articles chosen for inclusion focused on the experiences of women; investigated online abuse within public facing occupations; and were written from a victim perspective, in order to gain an understanding of how receiving abusive communication can affect an individual, both professionally and personally.The application of this revised criteria dramatically reduced the number of articles selected for inclusion in the review, creating a second tranche of data of 61 studies.This process is summarized at Figure 2.
The 61 studies included in the scoping review were further analysed to extract five pieces of information: (1) the theoretical framework underpinning the study; (2) the methodology adopted; (3) whether the explanation for online abuse was individual or structural; (4) the geographical location of the research population; and (5) the online platform being investigated.These dimensions were chosen as they best expressed the information required to answer the research question.
In order to further interrogate the data from the scoping review, the final selection of papers was imported into the NVivo qualitative data analysis program.By using the various tools within the software, it was possible to achieve a higher level of validity and reliability than a traditional literature review (Maureen M. O'neill , Sarah R. Booth, and Janeen Therese Lamb 2018).52 of the 61 studies analyzed in the scoping review defined online abuse as a structural issue.The nine studies that adopted an individualized perspective were located within the disciplines of Law, Psychology, and Information Systems.The findings emanating from the scoping review can be grouped into three overlapping themes, as illustrated in Figure 3.

Online abuse as an expression of gender-based violence
Much of the literature written about online abuse prior to 2010 adopted individualized explanations for the phenomenon (Henry and Powell 2015), often informed by libertarian theories of free speech (Dragiewicz et al. 2018).However, this study reveals that this has changed significantly over the last decade, with over 85% of the studies in time period two and time period three defining online abuse as a structural issue linked to wider gender inequality.It would appear likely that this change in approach has been driven by the recognition that women receive the majority of abuse perpetuated via CMC (Backe, Lilleston, and McCleary-Sills 2018), much of it sexual in nature (Dunja Antunovic 2019).Work by feminist academics (e.g., Jane 2017a; Mantilla 2015; Danielle K. Citron 2014) has demonstrated that the nature of this communication reinforces gender stereotypes and consolidates gender role models (Sue A. Barratt 2018).The technological advances instigated by SNS to extend the reach of the patriarchy in this way (Elaine Campbell 2017) has links with other forms of sexism (Andrea Carson 2018), reflecting social inequality that persists offline (Dragiewicz et al. 2018).
In addition to disseminating sexism, online platforms have often created new ways to perpetrate gender-based violence (Barlow and Awan 2016), and the nature of the threat, descriptions of sexual assault and the fear that this engenders, has parallels with violence against women that occurs offline (Lewis, Rowe, and Wiper 2017).This finding has led to the creation of a number of theoretical descriptions of gender-based violence expressed through online abuse.Clare McGlynn, Erika Rackley, and Ruth Houghton (2017, 26) Rackley, and Houghton (2017) reinforces that sexual assault does not require physical contact, highlighting that the law as currently formulated is too narrow.Dragiewicz et al. (2018, 1) have devised a similar framework, entitled "technology facilitated coercive control," which includes manifestations of online abuse such as harassment via SNS, stalking using GPS data, clandestine and conspicuous audio and visual recording, threats made via the medium of text messages, the monitoring of email, accessing an individual's online accounts without permission, impersonating a partner, and publishing private information (doxxing) or sexualized content without consent (Dragiewicz et al. 2018, 4).When located within the literature defining online abuse as a structural issue, these two frameworks provide a basis for better defining online abuse as a range of malign behaviors that meet the definition of gender-based violence.Much of the literature adopting an institutional explanation for online abuse highlights a desire from many criminal justice and social policy agencies to downplay the impact of online abuse (Maria Edström2016).It is argued that such an approach fails to recognize the harm of such behaviors (Henry and Powell 2015) and has clear parallels with the treatment of sexual crime (Jane 2014) effects of online abuse are borne out of an institutional failure to address gendered discrimination.It also emphasizes that this structural deficiency is demonstrated by the failure of criminal justice agencies to prosecute episodes of online abuse.

Misogyny
The scale of online abuse against women uncovered in the scoping review reaffirms that misogyny is a critical factor in the production and dissemination of online abuse (Barlow and Awan 2016).Indeed, there is evidence that the emergence of new forms of CMC have created new forms of misogyny (Debbie Ging and Eugenia Siapera 2018), allowing those who share similar views to connect and maximize hate and hostility, phenomena that are amplified online (Majid Khosravinik and Eleonora Esposito 2018).The scoping review has confirmed that rather than being an individualized phenomenon, online abuse has a number of commonly occurring characteristics.The markers of online abuse include communication that is highly aggressive, labels women as weak, and frequently contains threats of physical or sexual violence (Sady Doyle 2011).This reinforces that online abuse is a structural issue.This finding has important ramifications, removing the potential for blame from individual victims of online abuse and negates any discussion of what they may have done to deserve vicious invective (Jane 2017a), and instead makes it a societal issue.The link between online abuse and misogyny is particularly evidenced in the way that it attempts to silence women (Barlow and Awan 2016), an aim and function of wider misogyny (Carson 2018) which becomes a consequence when law enforcement agencies advise women to curtail online activity.The act of silencing women in this way, has the potential to wreak huge economic and personal consequences.CMC is now so firmly embedded in social and economic participation (Henry and Powell 2015), that to deprive victims of online abuse of this crucial locus of professional and social interaction is hugely damaging.This position has only increased following the changes to work and domestic routines caused by Covid-19.The cost of such silencing is so substantial that Jane (2018, 576) has termed it an act of "economic vandalism."Associated with this is the potential for an even wider negative impact, affecting women's very equity and citizenship (Jane 2017a), as the online abuse received by women in the public sphere receive causes them to withdraw completely from public facing occupations.Barratt (2018) further defines gender-based violence enacted in the online space as a backlash against feminism, and in particular the support and campaigning mechanisms that women have created to ameliorate the online abuse they receive (Jane 2017a).

Discussion
The findings of the scoping review confirm that there is a failure in tackling online abuse at a structural level, which is both exemplified and exacerbated by inaction emanating from the policy sphere.
The responsibility for the inertia in tackling online abuse is shared between governments, law enforcement agencies and private technology firms (Emma A. Jane 2016), whilst the relative paucity of academic research in this area has undoubtedly contributed to the delay in the development and implementation of policy and legislation designed to tackle online abuse (Barlow and Awan 2016).Hodson et al. (2018) have emphasized the need for a multi-level policy response, with action required at three levels-the personal, the organizational, and at a societal or cultural level (Hodson et al. 2018).In order to offer an approach for the devising of potential policy solutions, the Hodson et al. (2018) framework has been used to divide recommendations into three.

Personal responses
It is impossible to make policy recommendations that will realistically be adopted by everyone at risk of encountering online abuse.Nonetheless, given that the vast majority of those receiving online abuse deal with it without involving any external organizations (Jane 2017a), the actions taken by women at an individual level require recognition.These strategies include seeking informal advice and support from family members and friends (Hodson et al. 2018), or from a wider feminist "sisterhood," created to directly challenge online abuse (Antunovic 2019).Many women have chosen to respond to receiving online abuse in this way because of the paucity of the legislative, occupational, or criminal justice sector response (Jane 2017b).The management of online abuse at a personal level, results in online abuse being perpetually under recorded (Backe, Lilleston, and McCleary-Sills 2018), and contributes to the often-hidden nature of the offence (Campbell 2017), and its links with public shaming and victim blaming (Lewis, Rowe, and Wiper 2017).

Organizational and governmental responses
It is at an organizational level that the greatest degree of action is required.Ideally, this would occur at a global level, with development of an international consensus (Dragiewicz et al. 2018).Literature in this review has emphasized the need for interdisciplinary academic research designed to contribute to the development of such policies, in order to capitalize on the various strands of existing scholarship (Backe, Lilleston, and McCleary-Sills 2018).
The research suggests that in many countries, the laws available to tackle online abuse effectively are inadequate (Henry and Powell 2015).However, as illustrated by the situation in England and Wales, even where there is legislation in place that facilitates the prosecution of online abuse (e.g., Crown Prosecution Service 2018), the number of prosecutions remains very low, and abuse continues to grow (Salter 2017).This scoping review concurs that the policing of online abuse needs to be improved.There needs to be greater collective pressure applied to police forces to tackle the issue (Laura Bliss 2019), along with an improved awareness of the scale and consequences of online abuse provided to individual police officers.A fundamental part of this process is better training for police, whose lack of technical competence (Edström 2016) around online platforms has long been proposed as a reason for the lack of action in dealing with criminal activity (Stine Eckert 2018).However, it is important to note that if every case of online abuse was to be reported to the police, then the entire criminal justice system would be overwhelmed (Barlow and Awan 2016).This raises other policy and resourcing dilemmas, particularly in the UK, where the social policy landscape has been under-resourced for many years.
It is not just the legislative and social policy sectors that need to develop better policy responses when tackling online abuse.The technology companies similarly have a responsibility to put robust and responsive reporting mechanisms in place (Barlow and Awan 2016), as at present such provision appears patchy.Given the huge advances made in CMC, and the large number of tasks that are now undertaken online, it seems unlikely that the reasons for a fragmentary response to online abuse are purely technical.Instead it is proposed that this is an issue of the prioritization of resources (Meserve 2014).Meserve (2014) is one of many commentators to ponder that with such highly skilled technical experts in their employ, there must be more options to halt online abuse available than the existing "report button." Even when action is taken at a corporate level, the effects are not always noticeable.As Jane (2017a) has highlighted, when Facebook banned anonymous accounts, in an attempt to tackle online abuse, it made very little difference.The amount of online abuse that was perpetrated via the platform did not reduce, with the men engaging in this damaging invective continuing to do so, without the benefits of anonymity.This suggests that identification did not serve as a deterrent, possibly because the likelihood of subsequent criminal sanction was miniscule.Furthermore, when SNS act to remove and delete sexist, criminal or defamatory posts, this can have the unintended consequence of destroying evidence needed for a criminal prosecution.This places the onus for evidence gathering on individual women, who become responsible for both capturing and storing their own abuse via screenshots, if they wish to pursue the matter beyond the confines of the individual platform (Melinda C. Burgess et al. 2017).

Societal and cultural change
This is arguably the most difficult level at which to initiate change, as it challenges ingrained attitudes and biases.However, if policy change were to be enacted at the first two levels, then this may hasten the necessary cultural shift.The type of change that is required at a societal level could be incorporated if a "woman defined understanding" (McGlynn, Rackley, and Houghton 2017, 38) were adopted to categorize the types of abuse that is defined as threatening and potentially violent (McGlynn, Rackley, and Houghton 2017).This would provide a much clearer understanding of both the nature and consequences of online abuse.At the same time, having more women leading technology companies (Carson 2018) would also do much to promote cultural and institutional change moving forwards.

Conclusion
Undertaking a scoping review of the literature that has been published over the last twenty years investigating the issue of online abuse has made it possible to access a rich interdisciplinary corpus.The highly systematic scoping review process has enabled the creation of a narrative synthesis (Brien et al. 2010) of this body of work, revealing how this literature defines online abuse, and the recommendations proposed for its amelioration.The latest literature in this area defines online abuse as a structural issue, and a manifestation of gender-based violence.There is also clear evidence of the role of misogyny in the perpetuation of abuse via online platforms.Solutions to the problem have been proposed at three levels: the individual, the organizational, and the societal.Each of these has strengths and weaknesses, and in reality, it will require a combination of all three approaches to achieve anything like meaningful change.

FEMINIST MEDIA STUDIES
The scoping review process has proven to be useful in accessing and analyzing literature that would not have been discovered otherwise.Consequently, this method is recommended to researchers who are keen to explore beyond traditional disciplinary boundaries.However, undertaking this project has revealed a number of limitations of this approach, and it would be worth other researchers being aware of these before deciding to embark upon a scoping review.Of particular importance to lone researchers, is the hugely time-consuming nature of this method.Even with the assistance of electronic databases, the iterative process of searching and modifying took months, and was very resource intensive.Once a suitable search strategy was identified, the sheer amount of data that was produced meant that the analytical process was more rudimentary than originally intended, with the risk that important information was overlooked.This particularly applies to the hand searches of journals, where it proved impossible to read all the abstracts that were identified.Inevitably, this means that the scoping review provides an illustrative snapshot and cannot be regarded as an exhaustive analysis of the existing corpus on online abuse.Some of the issues associated with amassing such a large amount of data were addressed by using qualitative data analysis software (NVivo) as a tool to scrutinize the content of the 61 papers that made up the scoping review.Nevertheless, despite the judicious use of this software, there will have been information missed, which is a matter of regret.
The protracted nature of the scoping review had numerous repercussions, both for the scoping review itself, which risked becoming out of date even before it was completed; and for the wider study, where time spent on the scoping review had an opportunity cost, resulting in less time available for other elements of the wider analytical process.
Furthermore, it should be remembered that whilst the scoping review process provides a mechanism to access literature produced across disciplinary boundaries, it takes no account of the work produced outside of the academic domain.This is particularly important in the study of online abuse, where the gaps in academic investigation have often been filled by the reports produced in the media, and by non-governmental organizations such as the United Nations (e.g., UN Broadband Commission for Digital Development Working Group on Broadband and Gender 2015), and campaigning organizations like Amnesty International (e.g., Amnesty International UK 2017).A number of these issues could have been better anticipated if there was a wider methodological literature available focusing upon how to undertake scoping reviews, especially for lone researchers, who are typically working without the benefit of significant resources.It is hoped that this paper, as well as providing useful information about the issue of online abuse, will go some way to filling that gap.
How does the existing literature define gendered online abuse of women, and what recommendations are proposed to tackle it?

Figure 1 .
Figure 1.Keywords and databases adopted for the search strategy.

Figure 2 .
Figure 2. The paper selection process (adapted from Norman Archer et al. 2011).

Figure 3 .
Figure 3. Themes arising from the scoping review.

Table 1 .
Data extraction table compiled for scoping review.
The role of social media in gender-based violence How