Prevalence, age of initiation, and patterns of co-occurrence of digital dating abuse behaviors nationwide

Abstract

Introduction:

Little is known about the patterns of adolescent and young adult digital dating abuse (DDA) nationwide. This study characterizes (1) the lifetime prevalence, (2) the age of initiation, and (3) the patterns of co-occurrence of both using and experiencing DDA behaviors in dating relationships.

Methods:

A cross-sectional online survey was conducted among a sample of 696 U.S. young adults recruited from Prolific, an online research platform. The sample was 50.7% female, 43.7% male, and 5.6% gender non-binary or transgender. The average age was 18.7 years (SD = 0.63, range: 16–22).

Results & Conclusions:

Among those with dating experience, 76.1% (n = 530) reported either using or experiencing at least one DDA behavior in their lifetime. Overall, 42.9% of respondents reported using and 58.3% experiencing digital monitoring and control behaviors, 25.0% reported using and 49.2% experiencing digital direct aggression, and 12.4% reported using and 36.4% experiencing digital sexual coercion. The average age of initiation for most DDA behaviors was 16 years with respondents reporting experiencing these behaviors at 11 years of age at the earliest. Of those with any involvement with DDA, 59.2% report both using at least one DDA behavior and experiencing at least one DDA behavior (n = 314), 32.5% report experiencing at least one DDA behavior but not using any (n = 172), and 8.3% report using at least one DDA behavior but not experiencing any (n = 44). DDA behaviors are common, can occur at young ages. Our findings highlight crossover between those who use and those who experience DDA behaviors and suggest prevention should focus on underlying issues that drive both the use and experience of these behaviors.

Introduction

Approximately 400,000 adolescents in the U.S. experience dating violence, consisting of physical, sexual, psychological, or emotional violence by a current or former dating partner (Vagi et al., 2015; Wolitzky-Taylor et al., 2008). Recent research finds that digital dating abuse (DDA), “the use of digital media to harass, pressure, threaten, coerce, or monitor a dating partner,” is also pervasive in adolescent and young adult dating relationships (Borrajo et al., 2015b; Brown & Hegarty, 2018; Clancy et al., 2020; Hinduja & Patchin, 2020; Korchmaros et al., 2013; Reed et al., 2016, 2017, 2018, 2020; Stonard et al., 2014, 2015; Zweig et al., 2013) given the extensive use of technology among this population. Other terms used to describe this kind of dating violence include cyber dating abuse (Borrajo et al., 2015a, 2015b; Zweig et al., 2013) and electronic aggression (Bennett et al., 2011).

Technology can be used in several different ways that might constitute dating abuse. Different studies use different measures to understand the scope and severity of DDA, but no particular measure is the gold standard (Brown & Hegarty, 2018). A comprehensive measure of a variety of DDA behaviors was developed recently (Reed et al., 2016) and drew from national surveys (National Campaign to Prevent Teen and Unplanned Pregnancy and CosmoGirl.com, 2008; Picard, 2007) and the Psychological Maltreatment Inventory for Adolescents (PMI-A) (Tolman, 1999). This measure considers DDA behaviors as falling into three conceptual subgroups: digital monitoring and control, “using social media/technology to keep track of, intrude on the privacy of, and control the activities of a dating partner;” digital direct aggression, “intentional behaviors meant to hurt, humiliate, or threaten a dating partner” using social media/technology; and digital sexual coercion, “pressuring a dating partner for online or offline sexual behavior and engagement in unwanted distribution of sexual images” (Reed et al., 2017). Among a sample of 703 mid-Western high school students with dating experience, 53.8% report digital monitoring and control, 46.3% report digital direct aggression, and 32.2% report digital sexual coercion (Reed et al., 2017). This expansion of dating violence into the internet, social media, and cyber communication between dating partners presents new and unique challenges in combating dating violence. Existing studies among high school participants limit questions about DDA behaviors to the current or most recent dating partner or in the most recent year, providing prevalence estimates that may not include all prior dating partners (Reed et al., 2017; Zweig et al., 2013).

Estimates of the prevalence of DDA behaviors have varied due to study location, definitions of digital dating abuse or cyber dating abuse, and age of participants (Brown & Hegarty, 2018; Hinduja & Patchin, 2020; Reed et al., 2017; Zweig et al., 2013). Multiple studies find differences in DDA behaviors between genders as well as responses to these behaviors, but the evidence is mixed (Bennett et al., 2011; Lucero et al., 2014; Reed et al., 2016, 2017, 2018, 2020; Stonard et al., 2015; Tynes & Mitchell, 2014; Zweig et al., 2013). In general, those who identify as male are more likely to report using sexually coercive behaviors (Reed et al., 2016; Zweig et al., 2013). There is less consensus in differences across genders in the use and experiences of monitoring and control and direct aggression. The most detailed study focusing on gender differences of DDA behaviors finds that boys and girls experience DDA at similar rates of frequency except for sexual coercion. Girls were more likely than boys to report experiencing digital sexual coercion and much more likely to report experiencing distress for all three types of DDA than boys (Reed et al., 2017). Because these studies each have a particular participant population (e.g. middle school students, high school students, college students), the patterns of DDA behaviors across ages has not been thoroughly explored.

Several important factors for public health interventions to address DDA are still unknown. We sought to contribute to the literature by describing and characterizing (1) the reported lifetime prevalence of DDA behaviors to replicate prevalence estimates found in prior work in a national sample of young adults in the U.S., (2) the age of initiation of these behaviors, and (3) the patterns of co-occurrence of using and experiencing DDA behaviors. Prior surveys have asked about DDA behaviors among students in all high school grades (Reed et al., 2017), and those in grades 7–12 (Hinduja & Patchin, 2020; Zweig et al., 2013), but the earliest onset age of these behaviors is not known and is a critical measure that can inform when to deliver primary and secondary prevention interventions. Furthermore, it is well documented that the use and experience of intimate partner and dating violence as well as other adolescent violence like bullying is often mutual (Langhinrichsen-Rohling et al., 2012; Taylor & Mumford, 2016; Yang & Salmivalli, 2013). Prior work separately examined those who use and those who experience DDA behaviors, but studies have found associations between DDA behaviors and other forms of dating violence (physical and sexual) (Korchmaros et al., 2013; Stonard et al., 2014, 2015; Zweig et al., 2013). However, there is limited evidence on patterns of co-occurrence across DDA behaviors. To date, it is also not known whether the consequences of or reasons for DDA behaviors differ between those who only experience or only use them, and those who both use and experience them. Understanding co-occurrence, motivations, and consequences can provide critical information for primary and secondary prevention efforts including who to target, what to include, and addressing the specific motivations for involvement in DDA.

Methods

Design and Procedure

Participants were recruited using Prolific, a platform for online subject recruitment with detailed rules regarding the treatment of survey subjects on the platform, advanced integration with survey design platforms like Qualtrics, and a high rate of data quality compared to some online subject recruitment platforms (Palan & Schitter, 2018; Peer et al., 2017). Participants were provided with an online invitation to participate in the study and an online consent form that notified participants that some questions would be sensitive, asking about physical, sexual, and/or psychological abuse. Participants were informed that participation was voluntary, that all their responses were anonymous, and that they could withdraw their consent at any time. The consent form was provided at the beginning of the online survey. Both the consent form and the survey were administered through Qualtrics survey platform. As the research team did not have access to any identifiable information on participants, the Institutional Review Board at the University of Washington deemed the study exempt. Prior to fielding the study, our research team pilot tested the survey design with a group of 18 young adults ages 18–22. Responses from pilot testing were not retained or used in our analysis.

Our research team fielded the survey using the Prolific platform and Qualtrics survey software from July 11, 2020 to July 15, 2020 to a sample of U.S. young adults ages 18–19 years according to their Prolific registration. All participants who consented and completed the survey were automatically redirected back to the Prolific platform for compensation ($9.50 per hour). There were 3,751 participants from the Prolific survey pool who were eligible to participate in the study (i.e., were 18–19 years according to their Prolific registration). Our target enrollment was 1,000 participants. We met target enrollment, and 1,001 eligible individuals elected to participate in the study. Two participants chose not to consent to participation and were not surveyed. Among those who consented, 100% of participants completed the survey. All survey responses were anonymized, and 27% of participants (n = 274) responded “No” to “Have you ever had a dating partner?” and were excluded from our analyses. An additional 4% of participants (n = 29) were excluded based on survey data quality checks. The final sample for analysis contains 696 participants. A participant flow diagram is included in Appendix B and describes the data quality checks that were used on the sample.

Measures

Participants were asked a screening question to determine whether they have ever had a dating partner, and therefore could be at risk of DDA. Participants were informed that the term dating partner means any of the following: “a boyfriend or girlfriend, someone you are a “thing” with, someone you have dated or are currently dating (e.g., going out without being supervised), someone who you like or love and spend time with, or a relationship that might involve sex.” Participants were asked “Have you ever had a dating partner?” Participants who responded “Yes” to this question were given the remainder of the survey. Respondents who indicated “No” were directed to the end of the survey. This broad definition for dating partner is used for two reasons. First, prior work raises concerns about the use of more limited terms which may inadvertently include only serious relationships and exclude casual dating relationships (Vagi et al., 2015). Second, the terminology used to describe dating relationships may evolve over time and this approach has been validated in prior work (Reed et al., 2016, 2017, 2018, 2020).

The survey asked participants with prior dating experience about their lifetime experiences with DDA and covered 3 domains: (1) DDA behaviors used against them and the age of first occurrence, 2) the worst DDA behaviors used against them and their emotional and behavioral responses, including disclosure, safety concerns, and self-reported mental health consequences, and 3) DDA behaviors that they used against dating partners and the age of first use of these behaviors as well as their motivations or reasons for using these behaviors. Finally, participants were asked to share demographic information on their current age (in years), gender identity, race, and sexual orientation. The survey fielded to participants is provided in the Appendix A.

Digital dating abuse

Digital dating abuse (DDA) was measured with a modified instrument from previous DDA studies [Cronbach’s alphas for score variables were .76 for victimization and .73 for perpetration] (Reed et al., 2016, 2017, 2018, 2020). We chose to use this instrument defining DDA because of the breadth of behaviors it considers as well as the definitions of behaviors in the context of dating violence. In this measure, behaviors are grouped into three broad categories: digital monitoring and control, digital direct aggression, and digital sexual coercion. Participants were informed that many people use the internet and cell phones to communicate with dating partners, including social media websites, text messages, and other digital platforms. To assess whether a participant experienced the use of a DDA behavior by a partner, each DDA behavior was preceded by the following prompt: “We would like to know whether any dating partner (current or past) has ever done any of the following things to you using the internet or a cell phone. My dating partner has…” To assess whether a participant ever used a DDA behavior with a dating partner, each DDA behavior was preceded by the following prompt: “Have you ever done any of the following things to a dating partner (current or past) using the internet or a cell phone?” To ascertain the age of initiation for each DDA behavior, all DDA behavior prompts were framed as lifetime experience or use of the behavior.

The modified survey instrument in this study included six items in the monitoring and control category. These items include behaviors like monitoring a partner’s location or activities, and looking at a partner’s private information (e-mails, text messages) to check up on the partner without permission. There were seven items in the direct aggression category including behaviors like sharing an embarrassing photo of a partner with others, spreading rumors about a partner, and threatening a partner physically. The last category, sexual coercion, included five items. This includes behaviors like pressuring a partner to sext or engage in sexual activities and sending naked photos without consent. All participants who indicated the use or experience of a DDA behavior were also asked their age (in years) when this first happened.

Consequences of and responses to experiencing DDA behaviors

Participants were asked a series of questions after indicating that they experienced a DDA behavior(s). Participants who experienced multiple DDA behaviors were asked to indicate the “worst” one, selecting from the set of DDA behaviors they indicated they had experienced. Based on prior work (Reed et al., 2020), we inquired about the “worst” behavior to focus on the likely most meaningful DDA experience according to each participant. While some DDA behaviors may be more common than others, this allows us to explore consequences and responses to the experience perceived as most harmful. Those who indicated only one DDA behavior were only able to select that behavior as the worst one. Participants were asked five additional questions about the “worst” DDA experience. They were asked to select any emotions they felt, any actions they took, whether or not they disclosed the worst thing that a dating partner did to them to anyone else, and if so, to whom. Prior work describes emotional responses as being either distressed or dismissive and behavioral/active (action-oriented) responses as being either engaging or blocking (Reed et al., 2017). Distressed emotional responses included things like “I cried” or “I was embarrassed,” and dismissive emotional responses included “I laughed” or “I ignored it.” Engaging active responses included “I tried to talk to them about it” and “I yelled at them or argued with them,” and blocking active responses included things like “I blocked them on social media” and “I avoided them.” In addition, they were asked “Did you ever experience mental or emotional harm (for example, anxiety or depression) because of what a dating partner did to you?” and “Were you ever concerned for your safety because of what a dating partner did to you?”

Reasons for using DDA behaviors

Participants who indicated they used DDA behaviors against a dating partner were asked two additional questions after indicating that they used a DDA behavior(s) and providing the age of initiation. First, they were asked to indicate the reasons they did these things. We compiled a list of seventeen reasons based on prior work (Borrajo et al., 2015a; Clancy et al., 2020) and feedback from the pilot test with young adults. Examples of response options were “As a joke, to be funny,” “Because I was angry,” and “To get someone to like you.” Participants were able to check multiple reasons and were able to select “Other” and indicate any additional reason that was not available in the provided set. We also asked an open-ended question to allow participants to note any additional details they wanted to provide by asking “Is there anything else you want to add about these experiences with a dating partner?”

Data analysis

Lifetime prevalence estimates for each DDA behavior were calculated as the number of individuals who responded “Yes” to whether they used or experienced each DDA behavior divided by the total number of participants who reported having a dating partner. Prevalence estimates were calculated using all behaviors participants noted as using or experiencing, not just the worst. Separate prevalence estimates were calculated for each DDA behavior across four mutually exclusive groups: (1) those who only used the behavior, (2) those who only experienced the behavior, (3) those who both used and experienced the behavior, and (4) those who neither used nor experienced the behavior. DDA behavior group prevalence estimates were calculated by dividing the number of participants who reported at least one behavior in that category divided by all respondents who reported having a dating partner. These group prevalence estimates were also calculated across the four mutually exclusive groups. A heatmap was generated to illustrate correlation between the use and experience of each DDA behavior for each participant. Behavior pairs with a larger number of participants both using and experiencing the particular behavior were shaded a darker blue.

For age of initiation, we calculated the average age as well as the standard deviation and interquartile ranges for each DDA behavior. Proportions of disclosure were calculated as the number of participants who reported disclosure of the worst DDA behavior divided by the number of participants who indicated that DDA behavior as the worst behavior. Finally, we calculated the frequency and percent distributions for all categorical variables including mental health, safety, emotional and behavioral responses to experiencing DDA behaviors, and reasons for using DDA behaviors.

Results

Sample and Measures

There were 353 cis-female participants (50.7%), 304 cis-male participants (43.7%), and 39 gender non-binary or transgender participants (5.6%). The average age of respondents was 18.7 years (SD = 0.63, range: 16–22). Over a third (37.6%) of respondents were White non-Hispanic or Latinx (n = 262), 8.5% were Black non-Hispanic or Latinx (n = 59), 16.1% of participants identified as Hispanic or Latinx (n = 112), and 28.7% identified as Asian (n = 200). The remaining participants identified as two or more races (n = 59), Native Hawaiian or other Pacific Islander (n = 2), American Indian and Alaska Native (n = 2). Of the sample, 67.2% identified as heterosexual or straight (n = 468), 28.9% identified as LGBQ (n = 201), 0.7% of respondents selected a different identity (n = 5) and listed pansexual, asexual, or bicurious, and 1.9% of respondents (n = 13) responded “Not sure.” The Cronbach’s alpha for the items included in the experiencing DDA behaviors scale is 0.86, and the Cronbach’s alpha for the items included in the using DDA behaviors scale is 0.72.

Lifetime DDA Prevalence

Over three-fourths (76.1% of participants, n = 530) reported either using or experiencing at least one DDA behavior (Table 1). Overall, 42.9% of respondents reported using digital monitoring and control behaviors, 25.0% reported using digital direct aggression, and 12.4% reported using digital sexual coercion. Over half, 58.3%, of respondents reported experiencing monitoring and control, 49.2% reported experiencing direct aggression, and 36.4% reported experiencing sexual coercion. Respondents indicated three digital monitoring and control behaviors that they commonly used against dating partners with similar prevalence rates: pressuring a partner to respond quickly (n = 143, 20.5%), monitoring a partner’s location or activities (n = 148, 21.3%), and monitoring who a partner talks to or is friends with (n = 142, 20.4%). The most common digital monitoring and control behaviors participants reported experiencing were pressuring them to respond quickly to calls, texts, or other messages (n = 274, 39.4%) and monitoring a partner’s location and activities (n = 200, 28.7%) or who they talk to or are friends with (n = 200, 28.7%). The most common direct aggression behavior used by participants was sending a mean or hurtful private message (n = 105, 15.1%). The most common direct aggression behaviors that participants experienced was receiving a mean or hurtful private message (n = 226, 32.5%). The most common sexual behavior respondents used was sending a sexual or naked photo of themselves to a partner without consent (n = 56, 8.5%), and the most common sexual coercion behavior respondents experienced were being pressured to sext (n = 181, 26.0%).

Table 1.

Prevalence of DDA Behaviors

Digital Dating Abuse Behaviors	Used DDA behavior only N (%)	Experienced DDA behavior only N (%)	Both used and experienced behavior N (%)	Neither N (%)	Declined to answer N (%)
Overall	44 (6.3)	172 (24.7)	314 (45.1)	166 (23.9)	--
Digital Monitoring or Control	63 (9.0)	170 (24.4)	236 (33.9)	227 (32.6)	--
Pressured to respond quickly to calls, texts, or other messages	56 (8.0)	184 (26.4)	87 (12.5)	363 (52.2)	6 (0.9)
Monitored partner’s whereabouts and activities	55 (7.9)	106 (15.2)	93 (13.4)	438 (62.9)	4 (0.5)
Sent so many messages it made partner feel uncomfortable	43 (6.2)	151 (21.7)	26 (3.7)	474 (68.1)	2 (0.3)
Monitored who my partner talks to and is/was friends with	64 (9.2)	120 (17.2)	78 (11.2)	429 (61.6)	5 (0.7)
Pressured for passwords to access cell phone or online accounts	20 (2.9)	72 (10.3)	11 (1.6)	591 (84.9)	2 (0.3)
Used private information to check on partner without permission	45 (6.5)	92 (13.2)	46 (6.6)	505 (72.6)	8 (1.1)
Direct Digital Aggression	31 (4.5)	200 (28.7)	143 (20.5)	322 (46.2)	--
Shared embarrassing content with others without consent	29 (4.2)	105 (15.1)	21 (3.0)	539 (77.4)	2 (0.3)
Sent a mean or hurtful private message	20 (2.9)	141 (20.2)	85 (12.2)	446 (64.1)	4 (0.6)
Posted a mean or hurtful public message using social media	12 (1.7)	75 (10.8)	11 (1.5)	594 (85.3)	4 (0.6)
Spread a rumor about partner	10 (1.4)	102 (14.6)	12 (1.7)	563 (80.9)	9 (1.3)
Threatened to harm partner physically	2 (0.3)	42 (6.0)	3 (0.4)	646 (92.8)	3 (0.4)
Used cell phone or online account to pretend to be partner	7 (1.0)	40 (5.7)	8 (1.1)	638 (91.2)	3 (0.4)
Used information on social network sites to tease partner	20 (2.9)	168 (24.1)	11 (1.6)	491 (70.5)	6 (0.9)
Digital Sexual Coercion	35 (5.0)	202 (29.0)	52 (7.4)	407 (58.5)	--
Pressured partner to sext	20 (2.9)	168 (24.1)	11 (1.6)	491 (70.5)	6 (0.9)
Sent a sexual or naked photo of self to partner without consent	29 (4.2)	95 (13.6)	27 (3.9)	540 (77.6)	5 (0.7)
Sent a sexual or naked photo of partner to others without consent	12 (1.7)	41 (5.9)	2 (0.3)	633 (91.0)	8 (1.1)
Pressured partner to have sex/other sexual activity	14 (2.0)	150 (21.6)	6 (0.9)	523 (75.1)	3 (0.4)

Notes. Participants were permitted to decline to answer whether they used or experienced a given DDA behavior. If a participant declined to answer either the used or experienced question for a given DDA behavior, they were categorized as “Declined to answer” because it is not possible to observe a pattern of co-occurrence.

Total used for each behavior is calculated as sum of the number “Used DDA behavior only” and the number “Both used and experienced behavior.”

Total experienced for each behavior is calculated as sum of the number “Experienced DDA behavior only” and the number “Both used and experienced behavior.”

The majority of respondents, 56.1% of females (n = 353) and 59.8% of males (n = 304), reported experiencing digital monitoring and control, 49.9% of females and 46.7% of males report experiencing digital direct aggression (Table 1B in Appendix B). More females in our sample report experiencing sexual coercion (46.2%) than males (22.7%). A large proportion of transgender female, transgender male, and gender non-binary participants in our sample reported experiencing all three categories of DDA behaviors. Both transgender female participants in our sample (n = 2) reported both using and experiencing all three groups of DDA behaviors, digital monitoring and control, digital direct aggression, and digital sexual coercion. Specifically, 61.1% of transgender males (n = 18) reported experiencing monitoring and control and direct aggression, and 50% reported experiencing sexual coercion. Over half, 57.9%, of gender non-binary participants reported experiencing digital monitoring and control, 63.1% reported experiencing direct aggression, and 57.9% reported experiencing sexual coercion.

Age of initiation

Among respondents who reported using DDA behaviors against their partners, the average age participants reported the first use of monitoring and control, direct aggression, and sexual coercion was 16.4 (SD = 1.73), 16.1 (SD = 1.79), and 16.8 (SD = 1.80), respectively (Figure 1). Among respondents who reported experiencing digital monitoring and control, digital direct aggression, and digital sexual coercion, the average age of first experiencing these behaviors was 16.1 (SD = 1.69), 16.0 (SD = 1.73), and 16.0 (SD = 1.85) years of age, respectively (Figure 1). One respondent in the sample accounted for three of the four age of initiation outliers. Age of initiation does not differ substantially for individual DDA behaviors within these categories.

Figure 1.

Age of first occurrence

Patterns of Co-occurrence

One-third (33.9%) of respondents both used and experienced digital monitoring and control behaviors, 20.5% of respondents both used and experienced direct digital aggression behaviors, and 7.4% of respondents both used and experienced digital sexual coercion behaviors with a dating partner (Table 1). The most common behavior that participants both experienced and used was monitoring a dating partner’s location and activities (n=93, 13.4%). These most common DDA behaviors that co-occur are similar to those that are the most common as described above. Figure 2 provides individual-level patterns of co-occurrence for each DDA behavior using a heatmap. Broadly, digital monitoring and control is both prevalent and co-occurs at higher rates than the other two forms of DDA. These behaviors are largely used in conjunction with other digital monitoring and control behaviors. If participants use similar behaviors that they experience, a darker shade would be observed across the 45-degree line where each behavior used intersects with the same behavior experienced. However, this pattern is not evident. The clearest pattern in the heatmap indicates that the use of digital monitoring and control co-occurs with experiencing all three types of DDA, monitoring and control, direct aggression, as well as sexual coercion. Among those who use or experience any DDA behaviors, 9.6% of respondents report experiencing behaviors in all three groups and using monitoring and control as well as direct aggression (Figure 2B in Appendix B). There are differences across gender identities in the patterns of co-occurrence as well, with similar patterns for digital monitoring and control, but bigger differences for sexual coercion. The complete breakdown of patterns of co-occurrence across DDA groups by gender identity is available in Table 1B in Appendix B.

Figure 2.

Heatmap of Individual-level Patterns of Co-occurrence

Disclosing the worst experience of DDA behaviors

Among the 486 participants who experienced at least one DDA behavior, the survey also asked participants to select among the behaviors they experienced which one was “the worst thing a dating partner did to you.” Among the 482 respondents who replied to this question, 30.9% (n = 149) indicated the worst thing a partner did was a monitoring and control behavior, 38.0% (n = 183) indicated the worst thing was a direct aggression behavior, and 31.1% of participants (n = 150) indicated the worst thing was a sexually coercive behavior. Overall, 54.5% of respondents who reported experiencing a DDA behavior disclosed the worst experience to someone (n = 265). Many participants indicated they disclosed the worst DDA behavior experienced to more than one individual, most often friends (94.3%), parents (16.6%), and/or siblings (12.8%). However, whether a participant shared their worst DDA experience with someone else differed across DDA behaviors (Figure 3). Several DDA behaviors were disclosed to others at high rates (impersonating a partner using the internet or a cell phone, 83.3%, threats of physical harm, 80.0%, posting mean or hurtful messages publicly, 75.0%, sharing sexual or naked photos or videos of a dating partner with others, 72.7%, and spreading rumors, 71.9%). Disclosure rates are similar regardless of whether an individual also indicated a history of using DDA behaviors against a dating partner. Other behaviors were disclosed less frequently, especially those that involve sexual coercion or monitoring (pressured to have sex, 50.0%, pressured to respond quickly, 48.7%, monitoring partner’s friends and contacts, 44.4%, and pressured to sext, 34.2%, monitoring location and activities, 29.0%).

Figure 3.

Percent of worst DDA behaviors participants disclosed

Consequences of experiencing DDA behaviors

There were 314 participants who experienced at least one DDA behavior and use at least one DDA behavior. There were 172 participants who experienced at least one DDA behavior but did not report using any DDA behaviors, and there were only 44 participants who reported using a DDA behavior but not experiencing any (Table 1). Among those who experienced DDA behaviors but did not report using them (n = 172), 33.7% indicated it negatively impacted their mental health, and 15.7% reported they were concerned about their safety because of what a dating partner did. Among those who both experienced and used DDA behaviors (n = 314), 54.5% indicated it negatively impacted their mental health, and 28.7% reported they worried about their safety. Emotional and action-oriented responses to experiencing DDA behaviors are similar for those who only experience them and those who both use and experience them (Figure 4). Among the 486 participants who report experiencing a DDA behavior, the most common emotional responses are distressed including being sad or upset (n = 281), being angry (n = 235), and being embarrassed (n = 192). The most common action-oriented responses to experiencing DDA behaviors are engaging including talking to the partner (n = 229), telling someone else about it (n = 155), and telling the partner no (n = 142).

Figure 4.

Emotional and Behavioral Responses to Experiencing DDA

Reasons for using DDA behaviors

Figure 5 depicts the number of respondents that reported each reason for using DDA behaviors. The reasons respondents indicate they used these DDA behaviors do not differ substantially across the groups that only used the behaviors, and those who both used and experienced them. The four most common reasons given were that the participant “did not think it was a big deal” (n = 173), did it “to get noticed or attention” (n = 106), did it “because they were angry” (n = 105), or “to show that they cared” (n = 95). Participants who selected “Other” primarily indicated curiosity, jealousy, insecurity, and concern as contributing factors. Several also indicated that with the invention of location-sharing options in device applications, constantly knowing the location of dating partner is part of many relationships.

Figure 5.

Reasons for using DDA behaviors

Discussion

Summary and Discussion

In a national sample of 696 young adults, 76.1% of respondents reported either using or experiencing at least one DDA behavior during their lifetimes. 58.3% of respondents reported experiencing digital monitoring and control, 49.2% reported experiencing direct digital aggression, and 36.4% reported experiencing digital sexual coercion with lower rates reported for using these DDA behaviors on their dating partners. These lifetime prevalence estimates across DDA types are slightly higher than those in prior work, which is expected because the participants in our sample are slightly older than those in previous studies. In addition, participants in prior work may have only been indicating experiences with their current dating partner. This study adds to the body of evidence that DDA is prevalent among adolescents and young adults nationwide. Our estimates are similar to prevalence estimates in prior work for males and females (Reed et al., 2017). We also provide prevalence estimates for non-binary gender identity groups though these sample sizes are limited. Future work should consider focusing on DDA behavior use and experience among these underrepresented gender identities.

Among all three behavior groups, monitoring and control, direct aggression, and sexual coercion, the average reported age of initiation of these behaviors is about 16 years, with the bottom quartile (earliest) indicating these behaviors started between 11 and 15 years of age. Many participants who reported experiencing these behaviors also report using them on dating partners. This group of mutual users and experiencers also report negative effects of DDA on their mental health and safety. The detrimental health and safety effects noted in participant responses and any mental health interventions to address them may benefit from using terminology that does not alienate individuals who both use and experience DDA behaviors and may need both trauma-informed care as well as behavioral interventions. We recommend a transition from the terms “victim” and “perpetrator” to “individuals who used” and “individuals who experienced” DDA behaviors to avoid alienating or stigmatizing these individuals.

These results overlap intuitively with extensive work in bullying which also finds that mental health consequences of bullying are common, and often more severe among the group of individuals who both use and experience the behaviors (Arseneault, 2017; Benedict et al., 2015; Lereya et al., 2015; Wolke et al., 2014). We find a high proportion of those who both use and experience DDA behaviors which is quite similar to work on the co-occurrence of other forms of violence including dating violence and bullying (Langhinrichsen-Rohling et al., 2012; Taylor & Mumford, 2016; Yang & Salmivalli, 2013). Furthermore, DDA behaviors often also resemble bullying behaviors. Our study finds that public DDA behaviors, behaviors that involve additional people outside of the dating relationship, are disclosed to others at high rates. Prior work also finds that these public DDA behaviors are often described as the worst experience with DDA (Reed et al., 2020). Because these behaviors involve individuals outside of the dating relationship, those who experience DDA may be more willing to disclose them because individuals outside of the relationship are already aware of the behavior. They may also be considered just as harmful as more private behaviors when used as a public tool for humiliation and embarrassment (Reed et al., 2020). More private DDA behaviors, ones that only involve individuals within the dating relationship, especially those that are focused on pressuring or monitoring partners, are disclosed to others less frequently. Both public and private uses of DDA behaviors can be detrimental. Future work is needed to assess whether bullying interventions (Evans et al., 2014; Vreeman & Carroll, 2007) could be appropriate and effective for addressing public DDA behaviors, and IPV/dating violence interventions (Jennings et al., 2017; Koker et al., 2014) for more private expressions of DDA. Future work should also longitudinally explore longer-term outcomes for individuals who both use and experience DDA behaviors.

Limitations

There are three primary limitations of our study. First, the use of an online sample may exclude underrepresented minorities due to the digital divide (Dilmaghani, 2018; Lenhart et al., 2005; Tynes & Mitchell, 2014). Recent literature identifies Black youth as having lower access and using the internet at lower frequency than non-Black youth as well as being more likely to visit websites that teach kids and adolescents about being safe online and being more likely to talk to others about sex online (Tynes & Mitchell, 2014). In addition, the population of our study is not nationally-representative by race as illustrated in the high percentage of Asian respondents, and the low percentage of Black and white respondents, relative to the U.S. population. More work is needed to characterize the accuracy of these findings in a nationally-representative sample. Our findings are also likely affected by social desirability bias (Dillman et al., 2014) that may be present for all questions asking about the use of DDA behaviors. This is reflected in the lower validity Cronbach’s alpha for perpetration items compared to victimization items. We tried to minimize this effect in the survey design by carefully choosing wording that did not include negative perceptions about using these behaviors, but it is possible that asking the DDA behavior use questions after the questions about how harmful their DDA experiences were could have increased the degree of social desirability bias in the responses about using DDA behaviors. Further, in our assessment of lifetime co-occurrence patterns it is not possible to know whether the behaviors used and experienced occurred in the same relationship. Therefore, the patterns of co-occurrence should be interpreted as DDA behaviors that co-occur across relationships for a given individual. Finally, reported gender identity was provided at the time of the survey which may be different than at the time that some or all of the DDA behaviors occurred. It is documented in both our results and in prior work, that both gender minority and sexual minority individuals disproportionately experience violence. The causes and consequences of this disproportionate exposure should be a focus of future work.

Implications

Given our findings highlighting the crossover between those who use and those who experience DDA behaviors, prevention and intervention programs should focus on underlying issues that drive both the use and experience of these behaviors. Interventions for DDA must recognize and adapt to changing norms in adolescent dating relationships. Several participants in our study responded to the open-ended question asking for other comments about these experiences with dating partners by sharing that “I didn’t think these things were a problem,” and “These experiences are normal between both of us.” While these monitoring behaviors may be common in adolescent dating relationships, interventions would benefit from clearly addressing common experiences in dating relationships and putting them in the context of violence risks. For example, mutual location sharing on devices and apps was noted as a common behavior in dating relationships. Interventions should address that this type of continual access to a partner’s location can be a risk factor if a someone chooses to stalk their dating partner or escalates violence in the relationship. Future work should seek to describe both predictors of DDA and understanding the factors that put adolescents at risk of experiencing and/or using DDA behaviors. Further, the high rate of those who use these behaviors reporting they did so because “it wasn’t a big deal” is of particular importance in designing public health interventions. Primary prevention efforts should focus on educating adolescents on the potential harms of these behaviors on their peers and dating partners, as well as on what is considered legal and illegal communication using the internet or a cell phone, especially in the context of explicit content. In addition, these prevention efforts to educate adolescents on healthy and safe communication using the internet and cellphones with dating partners should begin as early as 10 years of age, with secondary prevention strategies focused on those ages 12–16.

• Most youth report using or experiencing at least one digital dating abuse behavior.

• The average age of initiation of digital dating abuse behaviors is 16 years of age.

• Digital monitoring and control is the most common digital dating abuse behavior.

• Many participants who report experiencing these behaviors also report using them.

• Mutual users and experiencers report reductions in mental health and safety.

Abbreviations:

DDA

digital dating abuse