The Victim-Offender Overlap in Technology-Facilitated Abuse: Nationally Representative Findings Among U.S. Young Adults

Abstract

There has been limited examination of the phenomenon of the victim-offender overlap in the field of technology-facilitated abuse (TFA). To design effective strategies to prevent TFA, it is important to understand which individuals are most at risk of victimization, perpetration, and to what extent a subset of people both experience victimization and engage in perpetration. This study drew on Cyber-Abuse Research Initiative (CARI) data, a nationally representative U.S. sample of adults ages 18–35. TFA measurement consisted of parallel scales for victimization and perpetration, each with 27 items assessing forms of technology-facilitated surveillance, monitoring/tracking, interference/communications, reputational harm, controlling/limiting access, and fraud. A bivariate probit of TFA perpetration and TFA victimization, as separate outcomes, was fit to allow for joint estimation of regression coefficients and robust standard errors. Analyses confirmed that TFA, similar to other forms of interpersonal aggression, is characterized by a substantial victim-offender overlap, with 30 percent of the sample reporting involvement both as a victim and as a perpetrator. Internet/social media use and social isolation did not distinguish victimization and perpetration. However, positive and negative affect as well as Lesbian, Gay, Bisexual, Queer, Asexual, or other sexual orientation (LGBQA+) were positively correlated with victimization, whereas female gender and having postsecondary education were positively associated with perpetration. These results may be used to design interventions and anticipate service needs. TFA, as a new topic of research, should capitalize on the theoretical and empirical article related to other forms of the victim-offender overlap.

Introduction

Technology-facilitated abuse (TFA) involves violations of privacy and safety, and may be perpetrated by intimates, acquaintances, or strangers.¹ Research and services related to TFA address cyberstalking, cyber-harassment, and cyberbullying, as well as technology-facilitated reputational harm, sexual abuse, financial abuse, surveillance abuses, and in-person harm.^2–4 TFA is a growing problem with associated mental health impact,⁵ material costs,⁶ and substantial risks of IRL (in real life) harm.^7–10

And yet, research regarding a TFA victim-offender overlap lags behind explanatory research conducted on other forms of interpersonal abuse and violence.^11,12 Understanding the prevalence of a victim-offender overlap in TFA, and characteristics of individuals involved in TFA both as victims and offenders, is critical for the purpose of universal prevention efforts as well as tailoring services for both victims and offenders.

Victim-offender overlap in TFA

The TFA-related research that has investigated the overlap between victims and offenders previously has not focused on adult interactions, but has generally focused on the concept of cyberbullying among youth.^13,14 Previous research on cyberbullying has consistently identified that there is a subset of youth who can be classified as both victims and aggressors,¹⁴ referred to as a “victim-offender overlap.”^15,16 TFA studies that have been conducted in adult samples have often focused on TFA in intimate or dating partnerships.^8,17–20

In intimate relationships, prevalence estimates of self-reported TFA perpetration are similar to those of TFA victimization,¹⁰ but not all studies clarify if individuals who self-report TFA perpetration have also experienced TFA victimization in interactions with the same counterpart.¹⁷ In other words, estimates of a victim-offender overlap may refer to behavioral experiences within a single relationship, or to behavioral experiences across different relationships.

However, whether the behavior is bidirectional in a single relationship or reflects involvement across relationships, the victim-offender overlap is a significant concern for prevention and services. In fact, in a study of past-year cyberaggression by or against a dating partner (not necessarily the same relationships), approximately three-quarters of participating college students reported both experiencing TFA victimization and perpetrating TFA.²¹

However, technology increases the opportunity for abuse outside of close relationships by removing the need for proximity or even being acquainted.¹ For example, a study of cyberbullying on Facebook not focused on dyadic partnerships documented the significant association between TFA perpetration and victimization among university students,²² a finding duplicated in college-based research supplemented by an MTurk sample.²³ As such, additional research that uses representative samples is needed. Moreover, research that uses a comprehensive measurement instrument to assess TFA is essential to improve precise estimation of the percentage of youth who are both victims and aggressors.^24,25

Risk and protective factors for TFA

A recent systematic review highlighted time spent online and on social media as commonly identified predictors of both TFA victimization and perpetration.²⁶ Research has shown, however, that online and social media activity are not constant across different identities in a population. Online activity—as measured by chosen platforms, intensity of Internet use, and forms of use—varies by race, gender, and other personal characteristics.^2,22,23,27 Moreover, having an online presence as a social media influencer, journalist, politician, or other public role, including scholarship, also correlates with exposure to TFA risks.^28–32 Evidence of varying online and social media patterns underscores the importance of conducting TFA research in representative samples.

In places where there is reliable Internet service and the users are technologically skilled, the Internet has had positive impacts such as fostering a sense of community, allowing people to meet each other, and reducing social isolation. But positive experiences are not universal, and community exclusion or TFA experiences can enhance feelings of social isolation.^33,34 Multiple studies have documented that TFA can create a sense of isolation for victims^20,35 (as well as negative affect and depression⁵). At the same time, feelings of loneliness or social isolation may also lead to TFA perpetration.³⁶

In addition to the correlation of TFA victimization and mental health, positive and negative affect may also be associated with a role in TFA perpetration. Neurobiological research suggests that there are both shared and interacting neuroreceptors for depression and aggression,^37,38 and immunology studies point to shared underlying responses for both depression and aggression.³⁹ This evidence is consistent with empirical studies of a positive correlation between these constructs.^40,41 There is every reason to believe that this body of research would extend to TFA perpetration as well.

Certainly, widespread research and public media attention has been paid to the impact of social media on mental distress and aggression.⁴² Less attention has been paid to positive affect, although where technology-facilitated interactions are dyadic, complicated feedback loops may inform participants' behavior and affective outcomes.^43,44 The current data and study design are an opportunity to understand how these risk and protective factors relate to TFA victimization and to TFA perpetration, adjusting for involvement in both.

The present study

The current research was guided by two research questions. First, we asked whether there is commonly young adult involvement in TFA as both victims and perpetrators, that is, whether there is a TFA victim-offender overlap. To answer this research question, we investigated the extent to which there is an overlap between reported experiences of TFA victimization and perpetration in a representative sample of U.S. young adults. We hypothesized that the extent of the victim-offender overlap would be fairly high, consistent with the overlap in psychological abuse in intimate as well as nonintimate relationships.⁴⁵

Second, we asked whether and how TFA victimization and perpetration may be distinguished by individual characteristics, adjusting for the degree of victim-offender overlap. Our hypothesis, based on evidence of shared correlates in other studies of the victim-offender overlap,⁴⁶ was that we would not find substantive distinctions. However, with few exceptions,^22,24,25 there has been limited study of a victim-offender overlap for adult TFA, and thus further investigation of differentiating characteristics or underlying relying risk factors might provide constructive guidance for prevention or treatment.

Methods

Data

Study data are drawn from a nationally representative cross-sectional data set collected via online survey through the Cyber-Abuse Research Initiative (CARI) from November 2020 to May 2021. A sample of adults ages 18–35 was drawn from the AmeriSpeak^® Panel, representative of ∼97 percent of U.S. households.⁴⁷ Black and Hispanic panelists were oversampled, and the survey was offered in English and Spanish. Out of 8,620 invited panelists, 2,752 (32 percent) completed surveys.

Respondents were asked to select gender identity; the subsample of respondents identifying as transgender, nonbinary gender, or other gender identity (some of whom marked more than one gender identity response option) does not support multivariable modeling. Descriptive results regarding TFA reports by this subsample are available elsewhere.² The proportion of missing data for each variable for cisgender respondents was observed to be <5 percent. Therefore, we employed list-wise deletion for our analysis under the assumption of data missing completely at random.⁴⁸ The current analytic sample includes 2,576 respondents.

Measures

TFA experiences

The 27-item CARI scale assessed lifetime exposure to TFA victimization, predicated on the introduction “if anyone ever frightened, angered, or annoyed you by doing any of the following without your consent, without your upfront knowledge (e.g., consistent with family rules or company policy), or when you did not want them to.” TFA measurement included items assessing monitoring/tracking (e.g., monitored or spied on me using spyware/stalkerware), interference/communications (e.g., encouraged other people to “troll,” attack, or harass me online), reputational harm (e.g., distributed, or posted online, an intimate image of me without my consent), controlling/limiting access (e.g., changed the password to my online accounts or social media), and fraud (e.g., accessed or manipulated my technology or accounts resulting in financial harm)—the full scale is available through prior research.²

The same set of 27 items, modified for directionality, was fielded to all respondents to measure TFA perpetration “without someone's explicit consent, knowledge, or desire.” The perpetration items followed a preamble that stated, “The next questions are asking if you have engaged in some online behaviors involving other people, with the understanding that this is in no way intended to shame or judge you for things that you may have done in response to someone's cyber-abuse against you. If you have experienced cyber-abuse, it is not your fault.” Summary indicators of any TFA victimization and any TFA perpetration are the two dependent variables.

Affect

We employed the emotional distress/depression scale (Short Form 4a) from the Patient-Reported Outcomes Measurement Information System (PROMIS), validated for use in general population samples.⁴⁹ We also fielded a 9-item validated measure of positive affect and well-being (PAW; from the Neuro-QoL™ measurement system).⁵⁰ In addition, respondents' social isolation was assessed through a 4-item measure (PROMIS Short Form 4a). Response categories ranged from “Never” (1) to “Always” (5) for all three measures, and higher scores (T-scores converted from raw summary scores) indicated higher levels of depression, PAW, and social isolation, respectively.

Online activity

We measured respondents' frequency of use of 10 different types of sites/apps (e.g., social media, dating sites, gaming, video or streaming, and messaging sites/apps). Mean frequency of use for each set of sites/apps was calculated on a scale of 0–6, ranging from “Don't use this type of site/app” (0) to “Use more than once a day” (6). Marking any use of a set of sites/apps with an indicator of “1,” a summary count of the number of platforms used (ranging from 0 to 10) was created. Respondents were also asked if they felt they had a public following, as “an influencer, prominent figure, or leader in any online or social groups,” coded 1 as a social media influencer for affirmative responses.

Sociodemographics

Characteristics measured for this study included respondent age, gender (female), race/ethnicity, education, sexual orientation (Lesbian, Gay, Bisexual, Queer, Asexual, or other sexual orientation [LGBQA+]), and household income (Table 1).

Table 1.

Sample Description (N = 2,576)

	N ^a	Frequency/mean (SD)
Age (years)	2,576	27.4 (4.9)
Gender^a
Female	1,302	50.6%
Male	1,274	49.4%
Ethnicity and race
Hispanic	504	19.6%
Non-Hispanic White	1,418	55.1%
Non-Hispanic Black	355	13.8%
Non-Hispanic Asian	146.8	5.7%
Non-Hispanic, Multiracial, or Other Race	152	5.9%
Sexual orientation^b
Heterosexual	2,155	83.6%
LGBQA+	421	16.4%
Education
High school equivalent or less	745	28.9%
Some college	848	32.9%
Bachelor's degree or above	984	38.2%
Household Income^c (18-point scale of income categories)	2,576	9.5 (4.3)
No. of sites/apps used	2,576	5.5 (2.0)
Frequency of use across sites/apps	2,576	2.2 (1.0)
Social media influencer/public figure	296	11.5%
Emotional distress-depression	2,576	56.0 (9.2)
Social isolation	2,576	50.5 (10.0)
Positive affect and well-being	2,576	51.2 (8.0)
No TFA perpetration or victimization	624	24.2%
TFA perpetration only	129	5.0%
TFA victimization only	1,055	41.0%
Co-occurring perpetration and victimization	768	29.8%

Weighted N (rounded to nearest integer) is limited to cisgender respondents given that neither the Current Population Survey nor the American Community Survey captures gender minority identities for the entire U.S. population data to adjust the CARI weights. The sample of transgender, nonbinary, or other gender identity respondents otherwise eligible for this study is n = 50.

Respondents identifying as lesbian or gay; bisexual, pansexual, or queer; questioning; asexual; or demisexual coded as 1 vs. heterosexual respondents coded as 0.

Income measured in 5,000 increments ranging from less than $5,000 to $200,000 or more.

CARI, Cyber-Abuse Research Initiative; LGBQA+, Lesbian, Gay, Bisexual, Queer, Asexual, or other sexual orientation; SD, standard deviation; TFA, technology-facilitated abuse.

Analysis plan

After descriptive analyses, we conducted a McNemar's chi-square test for paired data to examine whether there is a significant relationship between each respondent's report of TFA perpetration and TFA victimization. We then conducted separate logistic regression analyses of victimization and perpetration as outcomes, including each as an independent variable for the other outcome, adjusting for TFA correlates.

A significant correlation between perpetration and victimization was observed in these analyses, even when controlling for additional covariates,^a and thus we hypothesized that the outcomes were realizations of highly correlated underlying processes. Since standard regression analyses are not well-equipped to model two dependent variables simultaneously, we applied a bivariate probit approach,⁵¹ which allows for joint estimation of regression coefficients for two dependent variables.

The overlap coefficient “rho” ( $ρ$ ) denotes the correlation or dependency among TFA perpetration and victimization outcomes.^51,52 We first fit two unadjusted models to understand if there was significant overlap between the two outcomes. The results from the naive model demonstrated that the estimate of $ρ$ was significantly different from 0, implying that the underlying stochastic processes are indeed correlated and that the selected statistical model is appropriate.^b Thus, we included the covariates of interest in the model to study the relationship between these factors and the two outcomes.

Using R version 4.2.1, we applied bootstrapping to generate robust standard errors for the estimates. All analyses use weights created through the iterative proportional fitting method such that the weighted marginal distributions of age, gender (male or female), race/ethnicity, Census division, education, housing tenure, household phone status, and cross-tabulations of age, gender, race/ethnicity and that of race/ethnicity, gender and census region in the sample match with that in the U.S. Current Population Survey (CPS).

Results

Association between TFA perpetration and victimization

Approximately one in three respondents (30 percent) reported both TFA perpetration and TFA victimization, whereas 41 percent identified only as victims, and 5 percent identified only as perpetrators, and 25 percent did not experience nor perpetrate any form of TFA (Table 1). McNemar's test indicated a strong association between victimization and perpetration (χ² = 722.66, p < 0.001).

In the fully adjusted bivariate probit models (Table 2), we found that ρ = 0.36 and the 95 percent confidence interval (CI) of ρ is (0.29–0.42). Hence, it was evident that there is a nonignorable overlap between perpetration and victimization for which the factors in the bivariate model cannot completely account.

Table 2.

Bivariate Probit Regression Models of Technology-Facilitated Abuse Victimization and Perpetration

	Perpetration	Victimization
	Beta (SE)
Age (years)	0.01 (0.01)	0.02 (0.01)^*
Female^a	0.30 (0.08)^***	0.16 (0.09)
Ethnicity and Race^b
Hispanic	0.03 (0.1)	−0.05 (0.11)
Non-Hispanic, Black	−0.15 (0.12)	−0.13 (0.12)
Non-Hispanic, Asian	−0.19 (0.18)	−0.28 (0.17)
Non-Hispanic, Multiracial, or Other Race	−0.09 (0.16)	0.11 (0.16)
LGBQA+	0.05 (0.1)	0.27 (0.11)^**
Education^c
Some college	0.26 (0.12)^*	0.31 (0.12)^**
Bachelor's degree or above	0.31 (0.11)^**	0.07 (0.11)
Household income (18-point scale of income categories)	−0.01 (0.01)	−0.01 (0.01)
No. of sites/apps used	0.13 (0.04)^**	0.15 (0.04)^***
Frequency of use across sites/apps	−0.09 (0.08)	−0.03 (0.08)
Social media influencer/public figure	0.27 (0.12)^*	0.48 (0.16)^**
Emotional distress-depression	0.00 (0.01)	0.03 (0.01)^***
Social isolation	0.02 (0.01)^***	0.01 (0.01)^*
Positive affect and well-being	−0.01 (0.01)	0.02 (0.01)^**

Reference categories as follows: ^aMale; ^bNon-Hispanic White; ^cHigh School or less.

p < 0.05; ^**p < 0.01; ^***p < 0.001.

SE, standard error.

Correlates of TFA victimization and perpetration

Individuals with some college or an associate degree were more likely to report victimization ( $β$ = 0.31, p = 0.009) and to report having perpetrated TFA ( $β$ = 0.26, p = 0.030) compared with those with at most a high school education. Individuals with at least a bachelor's degree were also more likely to report perpetrating TFA ( $β$ = 0.31, p = 0.006). Older individuals ( $β$ = 0.02, p = 0.032), influencers ( $β$ = 0.48, p = 0.002), and individuals who identified as LGBQA+ ( $β$ = 0.27, p = 0.009) also were more likely to report victimization.

Individuals with a higher depression score ( $β$ = 0.03, p < 0.001), greater positive affect ( $β$ = 0.02, p = 0.004), and greater social isolation ( $β$ = 0.01, p = 0.025) were at a higher risk of TFA victimization. Mirroring findings in the victimization model, influencers ( $β$ = 0.27, p = 0.027) were also more likely to be a perpetrator but, as for affect, only individuals with a higher score on the measure of social isolation ( $β$ = 0.02, p < 0.001) were at increased risk of TFA perpetration.

Finally, online activity as measured by the number of different types of online platforms used was positively associated with victimization ( $β$ = 0.15, p < 0.001) and with perpetration ( $β$ = 0.13, p = 0.001).

Discussion

Four out of ten U.S. young adults reported only experiences of TFA victimization, whereas 1 in 20 reported being involved only in TFA perpetration, a ratio of 8:1. The ratio of victims to offenders remains twofold when the 30 percent of adults who reported both victimization and perpetration are accounted for. That 30 percent of young adults have been involved at some point in both TFA perpetration as well as experienced TFA victimization confirms that the victim-offender overlap is a prevalent characteristic of TFA overall, consistent with other interpersonal conflict behaviors.⁴⁶

However, the TFA victim-offender overlap in this sample was not as high as what has previously been detected among U.S. young adults reporting a victim-offender overlap for other forms of abuse, contradicting our hypothesis. For example, recent nationally representative studies have found a 44 percent victim-offender overlap for past-day abusive experiences (psychological, physical, or sexual),⁵³ and the victim-offender overlap in verbal abuse over the prior 6 months was 68 percent within intimate partnerships and 54 percent in nonintimate relationships.⁴⁵

Taken together, these results highlight the ease with which some offenders may spread TFA across more than one target. Notably, it is possible even for repeat offenders to self-identify as victims of TFA themselves. Understanding the motivation and needs of perpetrators, and thus the overlap between victimization and perpetration, is critical to TFA prevention efforts.

Whereas we had expected to find few substantive distinctions in the predictors of TFA perpetration and TFA victimization, our findings related to our second research question point to some nuances meriting closer investigation. Adjusting for the correlation between TFA perpetration and victimization, the descriptive findings for these two outcomes are somewhat, but not entirely, consistent with previous research that broader use of social media and online platforms was associated with TFA overall.²⁶ However, our distinction of the frequency of use of each of these platforms suggests that it was not time online, but exposure to different communities (indicated by different platforms) that represents greater risk of TFA involvement.

The increased risk of TFA involvement for individuals who reported that they were public figures or social media influencers may also reflect broader community interactions.⁵⁴ Technology experts and other professionals responding to individual distress as a result of TFA may use these findings to assist private individuals—as well as individuals who are seeking broader engagement for professional purposes and thus cannot avoid a broad online presence—to make informed choices regarding their online presence or establish protective protocols to reduce exposure to TFA risks.

Family, peer, and online and IRL community factors may play a role in generating patterns of behavior.⁵⁵ In the current sample, both TFA victims and perpetrators indicated a greater degree of social isolation than young adults without a history of TFA. That social isolation is associated with TFA perpetration, an externalizing behavior, is consistent with literature associated loneliness with maladaptive social skills and poor coping mechanisms.⁵⁶ Moreover, people who are looking to mitigate their feelings of loneliness or isolation by searching for community online may concurrently be exposing themselves to greater risk.

In either case, individuals who are feeling isolated may need targeted support for healthy and safe ways to use technology.⁵⁷ Especially since the COVID-19 pandemic, there has been widespread awareness of the risks of social isolation for mental health, and young people who faced pandemic-related restrictions as part of their adolescent development are a new generation of young adults. Entering the adult world often means leaving a more socially integrated circle characterized by regular interactions with family, and yet there is an opportunity for parents and other intimate friends to check in as to whether their loved ones' overall online presence and interactions are positive or damaging.

At the individual level, TFA victims differ from TFA perpetrators in both positive and negative affect. Our results show that emotional distress, adjusting for perpetration, distinguishes TFA victimization, but not vice versa. And yet, in other research, some TFA victims have expressed positive outcomes of their experience,⁵⁸ and where there is a victim-offender overlap, there may be a direct association of positive affect and aggression.⁴³

The benefits of technology and technology-facilitated communications that attract users are potent; higher perceived benefits may lead to tolerance for some degree of TFA as well as enhanced self-protective behaviors.^59,60 The growing prevalence of TFA, while devastating to some, may be creating new coping mechanisms and conflict management strategies¹² that complicate recognition and support from practitioners.

This study highlights the importance of practitioners being prepared to support victims of TFA through nonblaming approaches that explore a range of technological and skill-based self-protective behaviors. School-based cyberbullying programs, where collective intervention is possible before launching youth into adulthood, will be more effective if led by technology experts rather than classroom teachers.⁶¹

Accounting for the victim-offender overlap, ethnicity and racial identity distinguished neither perpetration nor victimization in this representative U.S. sample. This finding stands in contrast with prior research in a single Southern U.S. college setting, which found that White students were more likely to be TFA perpetrators and to be involved in the TFA victim-offender overlap.²⁵ Regarding gender, prior research in university settings found no difference in the TFA victim-offender overlap by gender,^22,25 whereas in the current representative sample, adjusting for the victim-offender overlap, women were more likely to report perpetrating TFA, whereas no more likely than men to have experienced TFA victimization.

To interpret this gender finding, it may be useful to consider how cyber interactions differ from in-person interactions. A high proportion of cyber interactions are communications allowing for relational aggression, a behavior more commonly perpetrated by women than men.^62,63 Furthermore, the greater safety implied by cyber interactions may diminish women's self-control online such that they might more actively apply to IRL interactions.

However, this is not to imply interpretations regarding the severity or outcomes of TFA, particularly as public online harassment⁶⁴ and abuses that translate to IRL risk, such as doxing,⁶⁵ may build on gender-based grievances. In terms of other populations at particular risk across the U.S. culture, the current findings clearly highlight sexual minorities as a population requiring TFA information and services, consistent with other evidence that sexual minorities are particularly at risk of TFA victimization.^66–68

This study adds to the literature by bringing attention to the potential complexity of involvement in TFA, potentially as perpetrators as well as victims, for young adults who identify as sexual minorities. Organizations and individual professionals focused on supporting groups traditionally at greater risk for interpersonal abuses—namely women, individuals who identify as LGBQA+, and individuals identifying as transgender, nonbinary, or another gender identity—may find it useful to consider the possibility that TFA is often bidirectional.

Finally, it is ironic that so much of the cyber-abuse research among young adults has been conducted in college populations.⁶⁹ Our results suggest that TFA perpetrators are more likely to have attended college than to be limited to a secondary education. This finding may reflect technical education leading to more sophisticated skills or crimes, or the norms or hubris of a more privileged socioeconomic class. To date, the emphasis on prevention of cyberbullying has been on youth programming.⁷⁰ However, the current results indicate that college orientation programs and prevention training should include content on TFA as much as alcohol use and sexual misconduct.⁷⁰

Limitations and implications for future research

First, the study does not distinguish the relative intensity or severity of either TFA victimization or perpetration. Because the relative harmfulness of TFA experiences might impact a TFA survivor's subsequent behavior, it would be helpful to document frequency, intensity, severity, and TFA-related trauma in future studies of the TFA victim-offender overlap. Second, the CARI data are cross-sectional and thus results cannot be interpreted as causal. Third, because respondents were asked to report on lifetime TFA experiences, results cannot be interpreted as the overlap of TFA victimization and perpetration within a singular relationship.

In future research, dyadic samples beyond intimate partnerships would be of particular utility to attend to characteristics and perspectives reported by both parties and the timing of initiation of TFA behaviors (although this form of research would preclude studies of TFA perpetrated by casual acquaintances, strangers, or those participating in a coordinated group attack). Much TFA research to date among adults has focused on intimate partnerships,⁸ but more research is needed regarding the victim-offender overlap for TFA within and beyond intimate partnerships. Fourth, it is also possible that these experiences did not overlap in time, and associations with age at interview may be spurious.

To capture a victim-offender overlap with more precision, future research design should specify whether participants' involvement in TFA perpetration or experiences as a TFA victim are (a) contemporaneous and (b) involving the same counterpart(s)—studies may be designed to address one or both criteria, answering different research questions. Fifth, data collection from individuals living outside the nationally representative household sampling frame (e.g., in the military, in incarceration, or unhoused) are not included.

A focus on the potential for a TFA victim-offender overlap in these populations is warranted, particularly for in military populations given the preponderance of young adults and recognition that there is a TFA problem.⁷¹ Sixth, data collection was conducted in a young adult population that was experiencing the COVID-19 pandemic before most were likely eligible for vaccination; this context may have colored participant experiences and survey responses.

Research has shown that staying home more than usual because of pandemic-related mandates or restrictions was protective against TFA (Sheridan-Johnson J, Mumford EA, Maitra P, et al. Perceived impact of COVID-19 on cyberabuse, sexual aggression, and intimate partner violence among U.S. young adults. J Interpers Violence [Manuscript in preparation]); however, the measure of TFA in this study captured cumulative experiences before the pandemic, mitigating the impact of the timing of this survey. Future qualitative research might also help elucidate how public health restrictions impact TFA involvement.

Conclusion

With the potential for interpersonal abuses expanding alongside growth in the technology and communication sectors, research in TFA victimization and perpetration is critical. This study highlights the need to attend to the potential victimization of TFA perpetrators. Researchers in this field can benefit from past theoretical and empirical article regarding the overlap between victimization and offending related to other forms of interpersonal aggression. The TFA victim-offender overlap is robust and should be acknowledged both in prevention efforts and in professional response protocols, be they law enforcement, victim services, health services, or technological support.

Notes

In the logistic regression for TFA victimization controlling for covariates, individuals who perpetrated TFA had a higher likelihood of also being a victim than did nonperpetrators (β = 1.09, p < 0.001). Similarly, in the logistic regression for TFA perpetration controlling for covariates, individuals who were victimized had a higher likelihood of also perpetrating TFA (β = 1.1, p < 0.001).

There was a significant relationship among the overlapping categories of TFA ( = 0.43 and 95 percent CI of ρ = [0.38–0.49]). Furthermore, since the 95 percent CI of rho was significantly different from 0, we concluded that the error terms are correlated, and the bivariate probit model was appropriate to study the factors associated with each of the two outcomes.

Ethical Approval

The questionnaire and methodology for this study was approved by the Institutional Review Board of NORC at the University of Chicago (Protocol No. 20.01.13). Informed consent was obtained from all individual participants included in the study.

Footnotes

Authors' Contributions

All authors contributed to the study conception and design. The first draft of the article was written by E.A.M. and P.M., and all authors commented on previous versions of the article. All authors read and approved the final article.

Disclaimer

Points of views in this document are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice or any other organization.

Author Disclosure Statement

The authors have no competing interests to declare that are relevant to the content of this article.

Funding Information

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Grant number 2019-X1499-IL-SI awarded by the Office on Violence Against Women, U.S. Department of Justice.

References

Agustina

. Understanding cyber victimization: Digital architectures and the disinhibition effect. Int J Cyber Criminol, 2015; 9(1):35.

Mumford

, Maitra

, Sheridan-Johnson

, et al. Technology-facilitated abuse of young adults in the United States: A latent class analysis. Cyberpsychology, 2023; 17(3):7.

Patel

, Roesch

. The prevalence of technology-facilitated sexual violence: A meta-analysis and systematic review. Trauma Violence Abuse, 2022; 23(2):428–443.

DeKeseredy

, Stoneberg

, Lory

. Polyvictimization in the Lives of North American Female University/College Students: The Contribution of Technology-Facilitated Abuse. In: The Emerald International Handbook of Technology-Facilitated Violence and Abuse. (Bailey J, Flynn A, Henry N. eds.) Emerald Publishing Limited: Bingley; 2021; pp. 65–81.

Stevens

, Nurse

JRC

, Arief

. Cyber stalking, cyber harassment, and adult mental health: A systematic review. Cyberpsychol Behav Soc Netw, 2021; 24(6):367–376.

Borwell

, Jansen

, Stol

. Comparing the victimization impact of cybercrime and traditional crime: Literature review and future research directions. J Digit Soc Res, 2021; 3(3):85–110.

Marganski

, Melander

. Intimate partner violence victimization in the cyber and real world: Examining the extent of cyber aggression experiences and its association with in-person dating violence. J Interpers Violence, 2018; 33(7):1071–1095.

Fissel

, Graham

, Butler

, et al. A new frontier: Development and validation of an intimate partner cyber abuse measure in an adult-based sample. Soc Sci Comput Rev, 2021; 40(4):974–993.

Duerksen

, Woodin

. Technological intimate partner violence: Exploring technology-related perpetration factors and overlap with in-person intimate partner violence. Comput Hum Behav, 2019; 98:223–231.

10.

Watkins

, Maldonado

, DiLillo

. The cyber aggression in relationships scale: A new multidimensional measure of technology-based intimate partner aggression. Assessment, 2016; 25(5):608–626.

11.

Jennings

, Piquero

, Reingle

. On the overlap between victimization and offending: A review of the literature. Aggress Violent Behav, 2012; 17(1):16–26.

12.

Berg

, Mulford

. Reappraising and redirecting research on the victim–offender overlap. Trauma Violence Abuse, 2020; 21(1):16–30.

13.

Kowalski

, Giumetti

, Schroeder

, et al. Bullying in the digital age: A critical review and meta-analysis of cyberbullying research among youth. Psychol Bull, 2014; 140(4):1073.

14.

Estévez

, Cañas

, Estévez

, et al. Continuity and overlap of roles in victims and aggressors of bullying and cyberbullying in adolescence: A systematic review. Int J Environ Res Public Health, 2020; 17(20):7452.

15.

Chan

, Wong

DSW

. The overlap between cyberbullying perpetration and victimisation: Exploring the psychosocial characteristics of Hong Kong adolescents. Asia Pac J Soc Work Dev, 2020; 30(3):164–180.

16.

Kerstens

, Jansen

. The victim–perpetrator overlap in financial cybercrime: Evidence and reflection on the overlap of youth's on-line victimization and perpetration. Deviant Behav, 2016; 37(5):585–600.

17.

Taylor

, Xia

. Cyber partner abuse: A systematic review. Violence Vict, 2018; 33(6):983–1011.

18.

Fernet

, Lapierre

, Hébert

, et al. A systematic review of literature on cyber intimate partner victimization in adolescent girls and women. Comput Hum Behav, 2019; 100:11–25.

19.

Brown

, Sanci

, Hegarty

. Technology-facilitated abuse in relationships: Victimisation patterns and impact in young people. Comput Hum Behav, 2021; 124:106897.

20.

Rogers

, Fisher

, Ali

, et al. Technology-facilitated abuse in intimate relationships: A scoping review. Trauma Violence Abuse, 2023; 24(4):2210–2226.

21.

Melander

, Hughes

. College partner violence in the digital age: Explaining cyber aggression using routine activities theory. Partner Abuse, 2018; 9(2):158–180.

22.

Marcum

, Higgins

, Freiburger

, et al. Exploration of the cyberbullying victim/offender overlap by sex. Am J Crim Justice, 2014; 39(3):538–548.

23.

Kowalski

, Dillon

, Macbeth

, et al. Racial differences in cyberbullying from the perspective of victims and perpetrators. Am J Orthopsychiatry, 2020; 90(5):644–652.

24.

Weulen Kranenbarg

, Holt

, van Gelder

J-L

. Offending and victimization in the digital age: Comparing correlates of cybercrime and traditional offending-only, victimization-only and the victimization-offending overlap. Deviant Behav, 2019; 40(1):40–55.

25.

Nodeland

. The effects of self-control on the cyber victim-offender overlap. Int J Cybersecur Intell Cybercrime, 2020; 3(2):4–24.

26.

Marcum

, Higgins

. A systematic review of cyberstalking victimization and offending behaviors. Am J Crim Justice, 2021; 46(6):882–910.

27.

Anderson

, Steen

, Stavropoulos

. Internet use and problematic internet use: A systematic review of longitudinal research trends in adolescence and emergent adulthood. Int J Adolesc Youth, 2017; 22(4):430–454.

28.

Gosse

, Veletsianos

, Hodson

, et al. The hidden costs of connectivity: Nature and effects of scholars' online harassment. Learn Media Technol, 2021; 46(3):264–280.

29.

Lewis

, Rowe

, Wiper

. Online abuse of feminists as an emerging form of violence against women and girls. Br J Criminol, 2016; 57(6):1462–1481.

30.

Relly

JE.

Online Harassment of Journalists as a Consequence of Populism, Mis/Disinformation, and Impunity. In: The Routledge Companion to Media Disinformation and Populism. ( Tumber

, Waisbord

eds.) Routledge: London; 2021; pp. 178–187.

31.

Krook

, Restrepo Sanín

. The cost of doing politics? Analyzing violence and harassment against female politicians. Perspect Politics, 2020; 18(3):740–755.

32.

Eckert

. Fighting for recognition: Online abuse of women bloggers in Germany, Switzerland, the United Kingdom, and the United States. New Media Soc, 2018; 20(4):1282–1302.

33.

Best

, Manktelow

, Taylor

. Online communication, social media and adolescent wellbeing: A systematic narrative review. Child Youth Serv Rev, 2014; 41:27–36.

34.

Primack

, Karim

, Shensa

, et al. Positive and negative experiences on social media and perceived social isolation. Am J Health Promot, 2019; 33(6):859–868.

35.

Reiss

, Curbow

, Wang

. Young adults' perceptions of intimate partner cyberstalking: Behaviors, severity, and associations with depressive symptoms and social isolation. Sex Cult, 2022; 26(1):176–203.

36.

Marciano

, Schulz

, Camerini

A-L

. Cyberbullying perpetration and victimization in youth: A meta-analysis of longitudinal studies. J Comput Mediat Commun, 2020; 25(2):163–181.

37.

Popova

, Tsybko

, Naumenko

. The implication of 5-HT receptor family members in aggression, depression and suicide: Similarity and difference. Int J Mol Sci, 2022; 23(15):8814.

38.

Rosell

, Siever

. The neurobiology of aggression and violence. CNS Spectr, 2015; 20(3):254–279.

39.

Takahashi

, Flanigan

, McEwen

, et al. Aggression, social stress, and the immune system in humans and animal models. Front Behav Neurosci, 2018; 12:56.

40.

Berg

, Rogers

, Liu

, et al. The interpersonal context of depression and violent behavior: A social psychological interpretation. Aggress Behav, 2019; 45(4):437–449.

41.

O'Donnell

, House

, Waterman

. The co-occurrence of aggression and self-harm: Systematic literature review. J Affect Disord, 2015; 175:325–350.

42.

Ferguson

. Does the internet make the world worse? Depression, aggression and polarization in the social media age. Bull Sci Technol Soc, 2021; 41(4):116–135.

43.

Chester

. The role of positive affect in aggression. Curr Dir Psychol Sci, 2017; 26(4):366–370.

44.

Chester

, DeWall

. Combating the sting of rejection with the pleasure of revenge: A new look at how emotion shapes aggression. J Pers Soc Psychol, 2017; 112(3):413–430.

45.

Taylor

, Mumford

, Liu

, et al. Young adult reports of the victim–offender overlap in intimate and nonintimate relationships: A nationally representative sample. Crim Justice Behav, 2019; 46(3):415–436.

46.

Berg

, Schreck

. The meaning of the victim–offender overlap for criminological theory and crime prevention policy. Annu Rev Criminol, 2022; 5(1):277–297.

47.

Montgomery

, Dennis

, Ganesh

. Response Rate Calculation Methodology for Recruitment of a Two-Phase Probability-Based Panel: The Case of AmeriSpeak. University of Chicago National Opinion Research Center White Paper. 2016. Available from: https://www.norc.org/content/dam/norc-org/pdfs/WhitePaper_ResponseRateCalculation_AmeriSpeak_2016.pdf [Last accessed: November 17, 2023].

48.

Dong

, Peng

C-YJ

. Principled missing data methods for researchers. Springerplus, 2013; 2(1):222.

49.

Cella

, Choi

, Condon

, et al. PROMIS^® adult health profiles: Efficient short-form measures of seven health domains. Value Health, 2019; 22(5):537–544.

50.

Salsman

, Victorson

, Choi

, et al. Development and validation of the positive affect and well-being scale for the neurology quality of life (Neuro-QOL) measurement system. Qual Life Res, 2013; 22(9):2569–2580.

51.

Greene

Econometric Analysis (3rd ed.). Prentice Hall: Englewood Cliffs, NJ; 1997.

52.

Ashford

, Sowden

. Multivariate probit analysis. Biometrics, 1970; 26:535–546.

53.

Taylor

, Mumford

, Maitra

. Correlates of the victim–offender overlap based on daily conflict-level data from digital diaries. Crim Justice Behav, 2022; 49(8):1210–1232.

54.

Hassan

, Yacob

, Nguyen

, et al. Social Media Influencer and Cyberbullying: A Lesson Learned from Preliminary Findings. Knowledge Management International Conference (KMICe) 2018, Miri Sarawak, Malaysia, 2018; pp. 200–205.

55.

Cross

, Barnes

, Papageorgiou

, et al. A social–ecological framework for understanding and reducing cyberbullying behaviours. Aggress Violent Behav, 2015; 23:109–117.

56.

Heinrich

, Gullone

. The clinical significance of loneliness: A literature review. Clin Psychol Rev, 2006; 26(6):695–718.

57.

Nowland

, Necka

, Cacioppo

. Loneliness and social internet use: Pathways to reconnection in a digital world?. Perspect Psychol Sci, 2018; 13(1):70–87.

58.

Vakhitova

, Alston-Knox

, Reeves

, et al. Explaining victim impact from cyber abuse: An exploratory mixed methods analysis. Deviant Behav, 2021; 43(10):1153–1172.

59.

Vakhitova

, Alston-Knox

, Mawby

. Online routine activities and self-guardianship against cyber abuse. Vict Offenders, 2023; 18(4):623–645.

60.

Vakhitova

, Mawby

, Alston-Knox

, et al. To SPB or not to SPB? A mixed methods analysis of self-protective behaviours to prevent repeat victimisation from cyber abuse. Crime Sci, 2020; 9(1):24.

61.

, Chua

JYX

, Shorey

. The effectiveness of educational interventions on traditional bullying and cyberbullying among adolescents: A systematic review and meta-analysis. Trauma Violence Abuse, 2022; 23(1):132–151.

62.

Björkqvist

. Gender differences in aggression. Curr Opin Psychol, 2018; 19:39–42.

63.

Barlett

, Coyne

. A meta-analysis of sex differences in cyber-bullying behavior: The moderating role of age. Aggress Behav, 2014; 40(5):474–488.

64.

Sweeny

. Trapped in public: The regulation of street harassment and cyber-harassment under the captive audience doctrine. Nev LJ, 2016; 17:651.

65.

Calabro

. From the message board to the front door: Addressing the offline consequences of race-and gender-based doxxing and swatting. Suffolk UL Rev, 2018; 51:55.

66.

Vogler

, Kappel

, Mumford

. Experiences of technology-facilitated abuse among sexual and gender minorities. J Interpers Violence, 2023; 38(19–20):11290–11313.

67.

Powell

, Scott

, Henry

. Digital harassment and abuse: Experiences of sexuality and gender minority adults. Eur J Criminol, 2020; 17(2):199–223.

68.

Johns

, Lowry

, Haderxhanaj

, et al. Trends in violence victimization and suicide risk by sexual identity among high school students—Youth Risk Behavior Survey, United States, 2015–2019. MMWR Suppl, 2020; 69(1):19.

69.

Lund

, Ross

. Bullying perpetration, victimization, and demographic differences in college students: A review of the literature. Trauma Violence Abuse, 2017; 18(3):348–360.

70.

Polanin

, Espelage

, Grotpeter

, et al. A systematic review and meta-analysis of interventions to decrease cyberbullying perpetration and victimization. Prev Sci, 2022; 23(3):439–454.

71.

Department of Defense. Integrated Resilience Training and Support for the Prevention of Cyber Harassment and Technology Facilitated Sexual Harassment and Sexual Assault. Notice ID: FA300223R00202023.