Abstract
Researchers have described cyber sexual harassment (CSH) as a range of sexually aggressive or harassing images or texts delivered through the use of digital mediums. CSH occurs in high proportions particularly among young populations, but the prevalence varies substantially across studies, largely due to variation in measures used and types of behaviors assessed. We describe the prevalence of CSH by types of CSH assessed, identify the measures used in previous research, and provide recommendations for future assessment. Future research is needed to ensure consistent and comprehensive measures that accurately capture CSH to assess prevalence and consequences.
Sexual harassment has social implications as a form of gender discrimination and sexual aggression; however, it also poses a serious public health concern for adolescent girls in the United States and worldwide, particularly affecting girls’ mental health and health behaviors such as substance use (Bucchianeri, Eisenberg, Wall, Piran, & Neumark-Sztainer, 2014; Fairchild & Rudman, 2008; Gruber & Fineran, 2008; Madan & Nalla, 2016). Electronic communication technology, such as social networking sites (e.g., Facebook, Instagram, Snapchat) and text messaging, is increasingly a new mechanism through which sexual harassment and coercion is experienced among girls (Henry & Powell, 2015). Technology, including social media and smart phones, has been changing the communication landscape for young people, particularly individuals between the ages of 14 and 34 years (Henry & Powell, 2015). In the United States, about 95% of 12- to 17-year olds use the Internet and 74% have mobile devices (Livingstone & Smith, 2014). These types of cyber technologies facilitate social connections, but also present new ways in which girls experience sexual harassment. Researchers have described cyber sexual harassment (CSH) as a range of sexually aggressive or harassing images or texts delivered through the use of digital mediums (Henry & Powell, 2018; Madigan, Ly, Rash, Van Ouytsel, & Temple, 2018). CSH is unique compared to in-person experiences of sexual harassment in that offenders can more easily target victims across geographic barriers, are able to reach multiple victims at once, can be more difficult to regulate, and can potentially retain their anonymity (Henry & Powell, 2015). Studies indicate that adolescents are at high risk for CSH, with girls and LGBTQ youth being at even greater risk (Henry & Powell, 2018).
Less is clear how pervasive CSH is among adolescent populations. There is substantial variation in prevalence of CSH across studies, from 1-59%, likely due to differences in populations sampled but also due to variation in the range of CSH behaviors being assessed through current measures (see Table 1). More comprehensive measures of ever experiencing CSH conducted with older samples, such as those seen in online polls, have yielded larger estimates, whereas measures assessing past-year CSH and those conducted with younger adolescents and males have reported smaller prevalence estimates. Further complicating the study of CSH, there are various terms used to describe CSH. The use of consistent and comprehensive terminology and measures will be key to assessing the prevalence and consequences of this relatively new public health threat.
Types of Cyber Sexual Harassment: Examples of Items Used to Assess CSH Types and Prevalence Reported.
Note. CSH = cyber sexual harassment; IM = instant messaging; RDD = random-digit-dial.
Some prevalence estimates combine several items of CSH, as indicated in the table.
Thus, the objective of this research note is to summarize the CSH literature to date to better understand what types of CSH are being measured and reported. Specifically, we will describe the prevalence of CSH by types of CSH assessed, identify the measures used in previous research, and provide recommendations for future assessment.
Measures Used for Each Type of CSH and Implications for Future Research
In Table 1, we outline the types of CSH that are most commonly identified in studies, examples of measures used, as well as the prevalence reported, with the goal of supporting clarity on how best to measure CSH in future research. The types of CSH most commonly reported in studies have included the following:
Unwanted sexual solicitation
Receiving unwanted sexual messages/images
Having sexual texts/images shared without permission
Unwanted Sexual Solicitation
Sexual solicitation involves requests to engage in sexual behaviors that are unwanted by the recipient (e.g., Helweg-Larsen, Schütt, & Larsen, 2012; Marret & Choo, 2017; Mitchell, Wolak, & Finkelhor, 2008). Items pertaining to CSH include unwanted solicitation to exchange sexual messages/images, to have sex in-person, and to perform sexually over a webcam (Table 1). Unwanted requests to talk about sex and unwanted requests for sexual information/images were measured in most studies (e.g., Chang et al., 2016; Mitchell, Jones, Finkelhor, & Wolak, 2013; Ybarra, Espelage, & Mitchell, 2007). Researchers used questions such as “In the past year, did anyone on the Internet ever try to get you to talk about sex when you did not want to?” and “In the past year, did anyone on the Internet ask you for sexual information about yourself when you did not want to answer such questions? I mean very personal questions, like what your body looks like or sexual things you have done?” Furthermore, questions to measure unwanted solicitation to have sex in-person, send sexual photos, or pose sexually via a webcam included “In the past year, did anyone on the Internet ask you to do something sexual that you did not want to?,” “[unwanted] hinting or asking you to send photos of a naked part of your body,” and if “someone has threatened you to pose for sexy pictures in front of the webcam.” However, the most commonly used measures asked about unwanted requests to talk online about sex, unwanted requests for sexual information, and unwanted requests to do something sexual. Prevalence estimates have varied given that some studies reported a combined prevalence (usually 20% or less) of these and a few reported the prevalence of each of these separately over the past year. (See Table 1 for examples of sexual solicitation and prevalence estimates.) Studies that assessed unwanted requests to send naked/sexual photos had especially high proportions (more than 20%) reporting these incidents; however, it is difficult to compare across different contexts and age groups.
While insightful, the most frequently used questions (unwanted requests to talk online about sex, unwanted requests for sexual information, and unwanted requests to do something sexual) may not adequately capture unwanted requests for or being pressured to send sexual messages or images, yet such scenarios have been reported in high proportions (Choi, Van Ouytsel, & Temple, 2016; Lippman & Campbell, 2014; Montiel, Carbonell, & Pereda, 2016). Nonconsensual sexting, which can include being pressured to send sexual images as well as to have sexual images shared without consent, has been identified in several studies on sexting (Drouin, Ross, & Tobin, 2015; Madigan et al., 2018; Reed et al., 2019; Walrave, Heirman, & Hallam, 2014). Based on these existing studies, CSH measures also need to specifically capture the prevalence of experiencing pressure or coercion to send sexual messages and/or images, as well as how often sexting occurs as a result of being pressured or coerced. While multiple studies have reported poor health outcomes related to sexting, particularly among adolescent girls, these consequences may be largely a result of high proportions reporting being pressured into sexting (Choi et al., 2016; Reed et al., 2019; Van Ouytsel, Walrave, Ponnet, & Heirman, 2015; Ybarra & Mitchell, 2014). Additional aspects that should be further explored to inform future measures include the perpetrator type and the contexts of sexual solicitations, such as on which digital platforms these occurrences are happening most often (e.g., texting, online messaging, video chat) and the ways in which images/videos may be shared and create vulnerability. In summary, most studies used items to assess unwanted requests to talk online about sex, unwanted requests for sexual information, and unwanted requests to do something sexual; however, items are also needed to specifically assess pressure or coercion to send sexual messages and/or images (Table 2).
Brief Summary of Recommendations for Measuring Cyber Sexual Harassment in Future Research.
Note. CSH = cyber sexual harassment.
Receiving Unwanted Sexual Texts or Images (“Unwanted Sexts”)
Another examined area of CSH involves assessing the receipt of unwanted sexts (e.g., Finn, 2004; Jones, Mitchell, & Finkelhor, 2011; Ybarra, Mitchell, & Korchmaros, 2011; Table 1). This measure encompasses events such as getting unwanted sexually explicit pictures, emails, text messages, or comments. An example of questions asked include “Has anyone sent you a naked picture without you asking?” As with sexual solicitation, the prevalence ranges substantially based on how this question is asked and especially whether the question implied that the message was sent privately versus a group email as well as whether it was a sexual photo taken of the sender or some other pornographic material, for example (see Table 1).
Again, given the high prevalence of all of these scenarios reported, this area warrants attention. In developing improved measures for the future, items need to be explicit in terms of whether the sexual image/message was sent as part of a mass email versus a personal message, as well as whether the sexual images were of the sender (versus something found online), as these may differ in terms of the sender’s intent and impact on the respondent. In addition, items also need to be better operationally defined to focus on communication that is unwanted to clarify the coercive nature of these behaviors. To summarize, future measures are needed to better assess context (private versus group message), content (e.g., whether or not sender is sharing a personal sexual photo), and consent, as sharing personal images with mutual consent may be a normal part of a sexual relationship (Table 2). These distinctions will allow researchers to better identify the scenarios that have a negative impact on recipients. In addition, like all CSH measures, the types of relationships between the recipient and sender (e.g., intimate or dating partner, friend, someone met online) as well as the varied digital platforms used will also be important to assess.
Sexual Texts/Images Shared Without Sender’s Permission (Also Known as Nonconsensual Sexting and “Revenge Porn”)
Another type of CSH involves disseminating sexual comments or pictures to the public or to other people without the sender’s approval (e.g., Helweg-Larsen et al., 2012; Jones et al., 2011; Priebe & Svedin, 2012; Table 1). An example of a question to measure this includes having a “nude or semi-nude image posted online/sent to others without permission.” Another examples is “Someone sharing explicit images of you without your consent.” Technology makes this behavior have a high potential for negative impact. Lifetime prevalence was usually assessed (versus past year) and most estimates were just below 10%.
Notably, few CSH studies have assessed this, and among those that did, some lacked specific assessment regarding whether the photos were shared without permission/consent. For example, questions that ask about whether sexual photos were shared by others need to detail whether this was done without the sender’s approval. In most online platforms, there is an expectation that anything posted publicly can be shared by others and usually remains under control (e.g., can be deleted) by the person who posted the content initially. As electronic communications evolve, the distinction between “public” and “private” communications requires careful delineation. Inaccurate assessment could conflate and confuse sharing of sexual images from CSH, impeding recognition of the coercive and abusive nature of CSH. Thus, in summary, future research should use measures to assess incidents when sexual messages/images are shared by others without permission/consent (Table 2).
Summary and Next Steps
Table 1 illustrates the wide range of CSH prevalence estimates, indicating that CSH is occurring in high proportions, but also warranting the need for consistent and comprehensive measures. While many studies included two to three of the above CSH types, only a minority of studies captured incidents of receiving unwanted sexual messages/images or when sexual messages/images are shared without permission, despite that these types of scenarios were prevalent (e.g., Chang et al., 2016; Lwin, Li, & Ang, 2012; Montiel et al., 2016). Furthermore, while multiple measures captured unwanted requests to talk about sex or for sexual information, more work is needed that specifically asks about incidents of being pressured or coerced to send sexual messages/images as well as to assess the proportion of sexual messages/images that are sent as a result of feeling pressured or coerced. Future research may also be needed to assess specific scenarios of cyber sexual solicitation, including requests for sexual webcam interactions and in-person sexual engagement, which was measured less frequently. Important to note for future research and measurement development is that some of these forms of cyber sexual solicitation may have implications for girls’ risk of sexual victimization off-line. Furthermore, the perpetrator type and platforms (e.g., video chat, texting, online messaging) in which the various types of CSH incidents are experienced across the population is largely unknown, but will be important in designing intervention approaches.
While decades of research has measured in-person experiences of sexual harassment, CSH is a new and emerging threat, often involving different scenarios and perpetrators. This relatively new research on CSH may not only highlight similarities between in-person and CSH but also identify unique aspects of each of these.
In summary, we identified the need for CSH items to include (a) receiving unwanted sexual messages/images, (b) having sexual messages/images shared without permission, and (c) unwanted sexual solicitation, which includes unwanted requests or pressure/coercion to send sexual messages/images, unwanted requests to talk about sex or for sexual information, and unwanted requests to do something sexual online or in-person. As part of studying CSH types and prevalence, it will be important for future research on sexting to distinguish scenarios that involve coercive or abusive sexting, including sharing sext messages without permission of the sender as well as the use of coercion to solicit sexual messages/images from others. Given the high proportion of these scenarios occurring, previous findings highlighting the relation between sexting and poor health outcomes may be due, in part, to the high proportion of sexting that is nonconsensual. The use of clear, operationally defined scenarios and behaviors are imperative to ensure the individual understands the question being asked and all the CSH behaviors it entails. Like other forms of sexual harassment and violence, CSH measures need to be sensitive to norms that promote sexual exploitation disproportionately among certain subgroups (e.g., as a result of gender or sexuality) and consider how stigma may influence CSH reporting. Finally, future work is needed to assess the prevalence of CSH perpetration and victimization, as well as the identification of perpetrators most commonly involved by CSH type.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The preparation of this manuscript was funded by a grant from the Bill & Melinda Gates Foundation (PI: Raj, Anita).
