Electronic Aggression Among Emerging Adults: Motivations and Contextual Factors

Abstract

The present study used quantitative and qualitative methods to investigate contextual factors and motivations associated with emerging adults’ electronic aggression victimization and perpetration with friends and dating partners. Participants (N = 226) reported online about electronic aggression occurrence and motivations, family risk, support from friends, and emotion regulation. Males reported more victimization than perpetration overall, whereas females reported more victimization than perpetration only with friends. Jealousy/insecurity emerged as the most common motivation for electronic perpetration; second most common was humor for males and negative emotion for females. Overall, risky family environment was associated with electronic aggression; yet, support from friends and emotion regulation each moderated this association. Discussion addresses potential miscommunications that can occur in electronic communication and the need to look at the interplay between in-person and online interactions.

Although mobile and online forms of communication provide many social benefits (Valkenburg & Peter, 2010), they have received extensive media and research attention as possible forums for harm (David-Ferdon & Hertz, 2009). A rapidly growing literature suggests that electronic modalities, such as e-mail, text messaging, and social networking websites can be used to hurt, embarrass, or threaten others (David-Ferdon & Hertz, 2009; Raskauskas & Stolz, 2007). Yet, considerable variability characterizes these electronic behaviors, which range from slight teasing to significant cruelty and originate from both known and anonymous sources (Wolak, Mitchell, & Finkelhor, 2007). Harmful electronic actions, which are variously labeled cyberbullying, online harassment, and electronic aggression, have been associated with various psychosocial concerns for adolescents, including suicidal thoughts (Hinduja & Patchin, 2010), depressive symptoms (Ybarra, 2004), and substance use (Hinduja & Patchin, 2008; for reviews on the prevalence and risks associated with cyberbullying, see Kiriakidis & Kavoura, 2010; Tokunaga, 2010).

There is preliminary evidence that emerging adults also engage in hurtful online behaviors (Dilmac, 2009), though less is known about the risk factors for or nature of these actions. Emerging adults, particularly those in college, may be vulnerable to electronic aggression, given their heavy utilization of electronic forms of communication (Angster, Frank, & Lester, 2010; Smith, Ramie, & Zickuhr, 2011) and the creation of college websites dedicated to anonymously gossiping about others. Theoretically, emerging adulthood marks a time when individuals gain autonomy and explore and develop their identities (Arnett, 2000, 2004). Electronic forms of communication present accessible ways for emerging adults to express themselves. The manner in which they use social media may also shape how they think of themselves and their relationships (Brown, 2006; Padilla-Walker, Nelson, Carroll, & Jensen, 2010). Furthermore, the deepening of relationships with friends and romantic partners is a developmental hallmark of emerging adulthood (Arnett, 2000, 2004; Fincham & Cui, 2011), but it may also be a time of heightened relationship instability and discord. Indeed, high rates of college students’ dating aggression (Straus, 2008) and relational aggression (Basow, Cahill, Phelan, Longshore, & McGillicuddy-DeLisi, 2007) draw attention to considering electronic media as tools for aggression because they offer readily available mechanisms for engaging in aggressive behaviors at any time and from any location. The public nature of social networking websites may also open up avenues for embarrassing people on a wide scale or incessantly monitoring their activities and interactions. Based on the premise that studying electronic media offers a window into emerging adults’ social relations, the present study investigates emerging adults’ electronic victimization and perpetration with friends and dating partners. The broader term electronic aggression is used because we conceptualize these behaviors as extensions of relational aggression, rather than characteristically cyberbullying behaviors that constitute repeated behaviors with the intent to harm (Tokunaga, 2010).

Bennett, Guran, Ramos, and Margolin (2011) quantitatively investigated electronic victimization among college students, conceptualizing electronic aggression as including acts of hostility, humiliation, exclusion, and intrusiveness conducted through electronic text and picture media. They found that 92% of participants reported being recipients of electronic aggression from friends or dating partners within the past year. Males reported more electronic victimization than females, which is contrary to previous studies using adolescent samples that have found more female victimization (Smith et al., 2008) or no gender differences (Ybarra & Mitchell, 2004). The inconsistent results in this literature regarding gender differences underscore the value of investigating whether males and females differ in their use of and motivations driving electronic aggression. To further explore the role of electronic aggression in emerging adults, the present study investigates electronic perpetration and not just victimization. We additionally expand knowledge by identifying motivations underlying electronic perpetration and investigating associations between family risk and electronic aggression as well as potential buffers of these relations.

Motivations for Electronic Perpetration

The literature has mainly focused on the frequency of electronic aggression, while overlooking motivations behind such actions. Yet, understanding these reasons could inform possible prevention points. A qualitative study of nonelectronic relational aggression among young adolescents suggests that maintaining social standing, dealing with negative internal emotional states, and simple dislike of the other person may be primary motivations for aggressing against peers (Pronk & Zimmer-Gembeck, 2010). Compared to younger, less socially experienced adolescents, emerging adults may possess greater skill in identifying self-ascribed intentions for aggressive behavior. The present study utilizes a qualitative approach in order to gather a more in-depth understanding of the motivations driving electronic perpetration.

Role of Family Environments in Electronic Aggression: Intergenerational Transmission of Violence Theory

In addition to more immediate circumstances surrounding electronic aggression, early family environment, a distal factor, may serve as a risk correlate, as the level of conflict, and quality of early family interactions can inform the types of interpersonal bonds that emerging adults experience (Collins & van Dulmen, 2006; Conger, Cui, Bryan, & Elder, 2000; Kinsfogel & Grych, 2004; Linder & Collins, 2005). The intergenerational transmission of violence theory (Pears & Capaldi, 2001) potentially explains how early exposure to a hostile family environment can influence emerging adults’ aggression. This theory proposes that those who grow up in aggressive and chaotic families are vulnerable for becoming involved in their own aggressive adult relationships (Kaufman & Zigler, 1987; Widom, 1989). There are several hypothesized pathways of transmission: parents’ modeling of aggression leading to children’s learned aggression and viewpoints that violence is a legitimate response to conflict; a family context marked by chaos and lack of warmth; absence of constructive problem-solving models contributing to children’s own problem-solving deficits; and lack of interpersonal skillfulness creating problems forming adult relationships as well as affiliations with aggressive peers and dating partners (Dodge, Bates, & Pettit, 1990; Ehrensaft et al., 2003; Smith, Ireland, Park, Elwyn, & Thornberry, 2011). Through these multiple interwoven pathways, a common end point is the increased likelihood of being in later aggressive relationships both as the perpetrator and as the recipient (Bank & Burraston, 2001; Ehrensaft et al., 2003; Linder & Collins, 2005).

It is likely that the transmission of violence spills over into electronic conflicts, given that emerging adults in college use electronic media to communicate with others and maintain relationships (Subrahmanyam, Reich, Waechter, & Espinoza, 2008). Research on family risk and electronic aggression in adolescents indicates that suboptimal parent-child relationships are associated with heightened risk for both electronic victimization and perpetration (Ybarra, Diener-West, & Leaf, 2007; Ybarra & Mitchell, 2004). Yet, there do not appear to be studies on whether associations between family risk and electronic aggression extend into the college years when family relationships become less salient and friendships and individual competencies might reduce the impact of a risky family environment.

Supportive Friendships and Emotion Regulation: Potential Moderators of Intergenerational Transmission

Despite support for intergenerational transmission theories, many individuals from risky families do not become involved in aggressive peer and dating partner relationships (Serbin & Karp, 2004; Smith et al., 2011). Support from friends and emotion regulation are two factors that have been shown to attenuate the influences of family risk on traditional forms of aggression and thus may be also protective factors against electronic aggression. Supportive friendships provide a sense of intimacy, social competency, and opportunities for developing positive conflict resolution strategies (Hartup & Stevens, 1997). Positive peer support has been shown to buffer children exposed to harsh family environments from peer victimization (Schwartz, Dodge, Pettit, Bates, & The Conduct Problems Prevention Research Group, 2000) and externalizing problems (Lansford, Criss, Pettit, Dodge, & Bates, 2003).

Receiving positive support from friends may be an important mitigating factor for emerging adults who spend increasing amounts of time with friends and whose peer relationships may influence the formation of other types of intimate relationships (Collins & van Dulmen, 2006). Having positive peer relationships predicts less aggression victimization and perpetration in emerging adult romantic relationships, even after taking into account family factors (Linder & Collins, 2005). Concerning electronic aggression, adolescents who perceive their friends as more trustworthy and caring engage in less cyber-aggression than those with lower levels of peer support (Williams & Guerra, 2007). Research has yet to investigate the moderating influence of friend support on associations between aversive family environments and electronic aggression in emerging adults.

Emotion regulation, or the processes influencing how emotions are experienced and expressed (Gross, 2002), may also protect against the effects of risky family environments (Curtis & Cicchetti, 2007). Difficulties regulating negative emotions, particularly anger, have been identified as a link between family violence and dating aggression (Jouriles, McDonald, Mueller, & Grych, 2012). Yet, good emotion regulation is associated with social competence in children (Eisenberg et al., 1997) and might interrupt connections between family violence and later aggression. In general, emotion regulation has been shown to play an important role in emerging adult relationships (Laurent & Powers, 2007). With respect to electronic communication, where emotional exchanges can be rapid, emotion regulation may be critical for preventing aggression. Individuals who can effectively manage negative emotional states, regardless of their previous family history, might be able to curtail harmful electronic exchanges, thereby lowering their probabilities of being a victim or perpetrator of electronic aggression.

Present Study

The present study addressed three aims in order to better understand electronic victimization and perpetration among emerging adults in college. First, we examined electronic aggression occurrence across relationships (friend vs. dating partner), roles (victimization vs. perpetration), and gender. Building on Bennett et al.’s (2011) study, we hypothesized that there would be more electronic aggression involvement with friends than dating partners and that males would report more electronic victimization and perpetration than would females. Second, using qualitative data, we investigated emerging adults’ electronic perpetration motivations to provide phenomenological background and generate hypotheses about gender differences in rationales for electronic perpetration. Third, we investigated associations between risky family environments and electronic aggression and tested whether these relations varied according to levels of peer support and emotion regulation. We predicted that more adverse early family environments would be associated with higher levels of electronic victimization and perpetration but that greater support from friends and better emotion regulation would each moderate these links. Gender was also examined in these family risk-electronic aggression analyses in an exploratory fashion.

Method

Participants

A total of 226 undergraduate students (138 females, M_age = 19.96 years, SD = 1.12) from an urban university participated in this study. To be eligible for this study, participants needed to be between the age of 18 and 23 years to represent the typical age range of college students. Participants also needed to be proficient in English and answer “Yes” to the question Have you been in a romantic relationship, at any point, during the past year? In all, 244 participants met these criteria; 18 participants were later excluded due to incomplete information. Of the 226 participants, the ethno-racial composition was 24.8% Asian, 45.6% Caucasian, 15.5% Hispanic/Latino, 11.2% multiracial/other, and 3.1% unknown. Participants completed online questionnaires, were guaranteed anonymity, and received research credit in exchange for participation. The university Institutional Review Board approved study procedures.

Measures

Electronic Aggression Victimization and Perpetration.

Participants reported on 19 electronically aggressive behaviors (Bennett, Guran, Ramos, & Margolin, 2011). These items, developed through focus groups with college students, represent behaviors that are potentially hostile (e.g., sent a mean text message), intrusive (e.g., intrusively called or texted me to monitor or check on me), humiliating (e.g., posted an embarrassing photo of me on a social networking site), and exclusionary (e.g., blocked me on an instant messaging program such as AOL instant messenger (AIM) or Google Messenger). For the present study, we developed a parallel perpetration version of these items, which asks participants to report on their own electronically aggressive behaviors against others (e.g., I sent a threatening text message). Participants first provided answers to all victimization items before completing the perpetration version. For each item, participants indicated how often the event occurred with friends and with dating partners during the past year: 0 = never happened, 1 = happened 1–2 times, 2 = 3–5 times, 3 = 6–10 times, 4 = more than 10 times.¹ Composite scores of friend victimization (α = .86), friend perpetration (α = .87), dating partner victimization (α = .86), and dating partner perpetration (α = .81) were created by averaging values across the 19 items.²

Electronic Aggression Motivations.

Participants provided open-ended responses to the following probe: If you have ever done any of the above (electronic aggression) items, please describe what motivated you. We used qualitative thematic analysis methods (Boyatzis, 1998) to draw reliable themes from participants’ open-ended responses. Having reviewed the literature on motivations for adolescent and adult aggression (Hamberger, Lohr, Bonge, & Tolin, 1997; Hinduja & Patchin, 2009; Pronk & Zimmer-Gembeck, 2010), the primary investigator conferred with the second author to generate hypotheses about what might motivate emerging adults to perpetrate electronic aggression. A preliminary set of themes was identified based on these hypotheses. After a preliminary categorization of approximately 25% of the data, we developed a more refined classification system (including definitions and examples) that included six themes: (1) retaliation; (2) jealousy/insecurity; (3) humor; (4) negative emotions; (5) nonmalicious behaviors; and (6) self-protection or privacy. Table 1 provides definitions, κs, and examples of each of these six themes. Two coders independently classified all available responses into one or more categories (average κ = .80), and a third coder settled coding discrepancies. Of the 226 participants, 171 (103 females) responded to the open-ended question about their motivations for carrying out these electronic acts. Of these 171 participants, the majority (64.9%) provided a response that contained only one of the six themes, whereas 45 participants (26.3%) described more than one theme. Reponses from 15 participants (8.8%) did not fall into any of the six themes.

Table 1.

Motivation Themes for Electronic Aggression: κs, Descriptions, and Example Responses.

Theme (κ)	Description	Examples
Retaliation (κ = .75)	Revenge or reacting in response to a perceived aggressive act	• Mostly to get back at them • Because I do not feel that those people treat me the way I should be treated and to let those people understand how it feels
Jealousy/insecurity (κ = .87)	Engaging in behaviors due to jealousy, suspicion, mistrust, or insecurity about self or a relationship	• Being insecure or jealous about a situation • I only do somewhat hurtful things in an attempt to make myself feel better and it never even works
Humor (κ = .92)	Actions are thought to be fun, entertaining, or merely jokes between friends	• I thought the embarrassing photo was funny, and hopefully they would feel the same way rather than insulted or embarrassed • Mostly joking with friends
Negative emotion (κ = .69)	Behaviors are due to anger, frustration, hurt, or other negative emotions	• Deep hatred • My girlfriend frustrates me sometimes. Out of anger or frustration, I have said something mean to her through a text while in a heated argument that was unresolved in person
Nonmalicious (κ = .52)	Behaviors were not done with the intentions of hurting someone, but may have been interpreted more negatively	• I have sent e-mail messages that I did not try to make sound mean but he interprets them as me being too sharp. Same with text messages, but I do not mean to • I never maliciously spread stories, just people would tell me and I would repeat funny stories
Self-protection (κ = .75)	Desire to keep oneself safe and one’s information private	• I block other people for being creepy • I blocked someone on Facebook because I was feeling bullied by them

Note. κ = kappa.

Family Risk.

The family risk variable was the mean value across 14 items (α = .86) assessing verbal and physical aggression among family members, chaos in the household, neglect, having a substance abusing family member, as well as positive dimensions, such as physical affection, feeling supported, and living in an organized household. Twelve of these items were taken from the Family Risk Questionnaire (Taylor, Lerner, Sage, Lehman, & Seeman, 2004), a measure developed for 18- to 25-year-olds and patterned after the categories of childhood abuse and household dysfunction described by Felitti et al. (1998). We added two questions to further assess family conflict. Participants answered questions on a 5-point scale ranging from 0 (not at all) to 4 (very often), with positive items reverse scored and higher scores representing greater family risk.

Positive Support From Friends.

We used the 25-item Peer Attachment scale from the Inventory of Parent and Peer Attachment-Revised (Armsden & Greenberg, 1987). This questionnaire has been normed on college students and measures perceptions of positive peer relationships, with specific items assessing alienation, communication, and trust with friends. In the present study, answers were based on a 4-point scale ranging from 0 (almost never/always) to 3 (almost always/always true). Example items include, I feel my friends are good friends and It seems as if my friends are irritated with me for no reason. Negatively worded items were reverse coded, and the support from friend score represents the average value across items (α = .93). Higher scores represent greater support from friends.

Emotion Regulation.

We assessed emotion regulation with an adapted version of the Emotion Regulation Checklist (Shields & Cicchetti, 1997), which includes 10 items that measure individuals’ abilities to cope with emotional arousal, recover from stress, and be responsive to others. Participants rated each item on a 4-point scale ranging from 0 (never true) to 3 (almost always true). We reverse scored negative items and averaged across items to create an emotion regulation score, with higher scores representing better emotion regulation (α = .69).

Results

Descriptive Statistics

Table 2 displays item-level endorsement of electronic victimization and perpetration in friends and dating partners. Overall, 86.3% of participants reported at least one act of friend victimization, 68.6% friend perpetration, 71.7% dating partner victimization, and 71.7% dating partner perpetration within the past year. Nearly half of the participants reported sending or receiving mean or hurtful text messages with both friends and dating partners. Intrusively calling or texting to monitor or checkup on someone was frequently reported with dating partners and being the recipient of embarrassing and unwanted photo posts was commonly reported with friends.

Table 2.

Percentage Endorsing Experiencing Electronic Aggression Item at Least Once in Past Year.

	Victimization		Perpetration
Item	Friend	Boyfriend/Girlfriend	Friend	Boyfriend/Girlfriend
Bullied through text messaging	26.7	21.4	17.3	21.7
Wrote something mean or hurtful on a website or through a chatroom/“bashboard”	15.0	7.6	8.4	6.2
Put a picture on website that did not want there	47.3	8.0	23.1	9.7
Took phone picture to embarrass	32.7	14.6	29.2	14.2
Sent a mean or hurtful e-mail	18.6	14.6	10.2	14.2
Sent a mean or hurtful text message	48.2	47.8	34.5	46.9
Sent a threatening e-mail	5.3	6.2	2.7	3.1
Sent a threatening text message	10.2	10.2	5.3	7.5
Intrusively called or texted to monitor or check up on me/someone	35.0	43.4	17.7	39.4
Used a fake online profile to interact with me/someone	11.1	3.1	3.5	2.7
Posted an insulting/hurtful comment or poke on a social networking site, e.g., Facebook or MySpace	27.6	10.2	21.2	7.5
Posted an embarrassing photo on a social networking site, e.g., Facebook or MySpace	57.3	11.1	24.1	12.0
Circulated an embarrassing but true story online	17.7	8.4	14.6	7.1
Circulated an embarrassing but untrue story online	10.6	4.4	2.2	2.7
Logged into my/someone’s e-mail to make trouble for or check up on me/them	8.4	15.5	7.5	19.5
Checked up on me/someone by being deceitful on the Internet	15.2	18.7	9.3	18.1
Blocked on AIM or on a website such as MySpace or Facebook	20.4	16.4	31.9	18.1
Excluded or shut out from top friend list	12.2	7.2	9.4	5.8
Tried to make jealous by commenting/posting on someone else’s Internet page	29.6	30.7	25.2	29.6

Table 3 presents intercorrelations between study variables as well as mean values for the total sample and for males and females separately for each variable. All four electronic aggression categories were positively correlated with one another. Friend victimization, dating partner victimization, and dating partner perpetration all correlated positively with family risk and correlated inversely with support from friends and emotion regulation; friend perpetration did not correlate with family risk or support from friends but correlated inversely with emotion regulation. Family risk was negatively related to emotion regulation and to support from friends. T test comparisons on all variables showed that support from friends was the only variable on which males and females significantly differed, t(155.21) = −3.91, p < .001. Females reported significantly higher levels of positive support from friends than did males, which is consistent with previous research (Armsden & Greenberg, 1987). One-way analysis of variances (ANOVAs) indicated no significant differences between the four ethno-racial groups for friend victimization, F(3, 218) = 1.16, p = .33, η² = .02; friend perpetration, F(3, 218) = 1.56, p = .20, η² = .02; dating partner victimization, F(3, 218) = .88, p = .45, η² = .01; and dating partner perpetration, F(3, 218) = 2.04, p = .11, η² = .03.

Table 3.

Intercorrelations and Mean Values for Electronic Aggression, Family Risk, Support From Friends, and Emotion Regulation Variables.

	1	2	3	4	5	6	Total M (SD)	Females M (SD)	Males M (SD)
1. Friend victimization	—						.35 (.35)	.32 (.32)	.39 (.42)
2. Dating partner victimization	.64**	—					.26 (.34)	.20 (.29)	.27 (.37)
3. Friend perpetration	.71**	.47**	—				.23 (.32)	.23 (.30)	.31 (.42)
4. Dating partner perpetration	.54**	.66**	.63**	—			.23 (.28)	.24 (.29)	.22 (.29)
5. Family risk	.23*	.27**	.12	.24**	—		.97 (.67)	.95 (.67)	1.00 (.68)
6. Support from friends	−.19*	−.23*	−.09	−.18*	−.36**	—	2.20 (.48)	2.30a (.43)	2.04 (.51)
7. Emotion regulation	−.19*	−.28**	−.19*	−.27**	−.32**	.49**	2.15 (.41)	2.18 (.40)	2.11 (.42)

^aFemales’ mean support from friends was significantly higher than males.’

^*p < .01.

^**p < .001.

Participant Gender, Relationship, and Role Differences Regarding Electronic Aggression

We conducted a 2 (participant gender) × 2 (relationship: friends vs. dating partner) × 2 (role: victimization vs. perpetration) mixed effects ANOVA, with gender as a between-group variable and relationship and role as within-group variables. As anticipated, a significant main effect indicated more electronic aggression with friends than dating partners, F(1, 224) = 8.71, p = .003, η² = .02, and more victimization than perpetration, F(1, 224) = 27.12, p < .001, η² = .04. There was a significant Role × Relationship interaction, F(1, 224) = 19.26, p < .001, η² = .01, but this result is best considered in the context of the 3-way interaction for gender, relationship, and role, F(1, 224) = 6.64, p = .011, η² = .001, which indicated different patterns of involvement for males and females. Males reported more victimization than perpetration, F(1, 87) = 14.02, p < .001, η² = .07, and more electronic aggressive involvement with friends than with dating partners, F(1, 87) = 8.94, p = .004, η² = .04. Females also reported more victimization than perpetration, F(1,137) = 11.59, p = .001, η² = .03. For females, there was no main effect of relationship, F(1, 137) = 1.54, p = .22, η² = .005, but there was a significant Role × Relationship interaction, F(1,137) = 31.28, p < .001, η² = .03. Pairwise comparisons indicated that females reported more victimization than perpetration regarding friends, t(225) = 6.95, p < .001, but not dating partners, t(225) = 1.64, p = .10.

Motivations Underlying Electronic Perpetration

Females Versus Males.

Approximately half of the females reported reasons for electronic perpetration associated with jealousy/insecurity (50.5%). Negative emotion (21.4%) was the next most common theme, followed by retaliation (17.5%), self-protection/privacy (17.5%), humor (11.7%), and nonmalicious intentions (10.7%). For males, jealousy/insecurity (33.8%) was also the most common motive (33.8%), followed by humor (27.9%), negative emotion (25%), retaliation (14.7%), nonmalicious intentions (5.9%), and self-protection/ privacy (4.4%). Percentages sum to more than 100% because some participants gave responses that contained more than one theme. Females, compared to males, were more likely to describe motives of jealousy/insecurity, χ²(1, 171) = 4.62, p = .04, and self-protection/privacy, χ²(1, 171) = 6.49, p = .02, whereas males were more likely to report humorous motives, χ²(1, 171) = 7.32, p = .01.

Female motivations suggested that electronic aggression often reflected extensions of offline difficulties in romantic relationships. A 21-year-old Caucasian female who admitted to logging into her boyfriend’s e-mail to make trouble for or check on him responded, “I have trust problems with my boyfriend, so from time to time I will check up on him.” A multiracial 19-year-old female explained, “I was jealous of my boyfriend’s ex-girlfriend, with whom he continued talking [… ]” Sometimes, online activity created trust issues:

I know my boyfriend’s Facebook password, and I can’t help but check to make sure he’s not being deceitful. I don’t know what’s motivating me because I have never found ANYTHING to make me think something bad is going on, but I can’t help myself. (Caucasian female, 19)

Female motives of self-protection/privacy at times appeared to be appropriate self-defense measures. For example, a 20-year-old Asian female blocked others on a website due to, “Harassment. I only block people when they really bother me (usually creepers).”

For males, jealousy/insecurity was relatively common and sometimes was combined with retaliation: “Being hurt and feeling rejected and feeling jealousy and wanting to get back at an ex, or at least make her feel the same way” (Caucasian, 20). Males’ endorsement of electronic aggression often originated from attempts to be humorous. For example, a 19-year-old Hispanic/Latino male who posted embarrassing photos of his friends on a social networking website, reasoned, “A lot of times, I did it in humor. A lot of my friends and I have the same meaning of funny and we would mess with each other in good fun.” A multiracial 21-year-old male insisted that his humorous behaviors were innocuous: “It was probably funny, and the person probably didn’t care.” Yet, not all online comments were made in jest: “I crack lots of jokes online on forums and such. I’ve said a few downright hateful things on [anonymous college gossip forum] somewhat ironically/hypocritically because the posts there sicken me” (Caucasian male, 18).

Intended and Unintended Aggression.

The desire to inflict emotional pain was evident in some instances. A 19-year-old Caucasian female who endorsed trying to make her friend jealous by commenting on someone else’s Internet page and sending a threatening text message to her dating partner stated, “I wanted to hurt someone else because they hurt me. Make them jealous because they made me jealous.” A 19-year-old Asian male who engaged in several types of electronic perpetration, such as circulating an embarrassing but true rumor about a friend online, attributed some of his behaviors to “revenge, anger.”

Responses indicated that electronic aggression sometimes resulted from attempts, albeit perhaps misguided, to demonstrate affection: “Any embarrassing pictures of friends have been out of love” (Caucasian female, 21). A 19-year-old Asian female explained, “I posted embarrassing pictures of friends just for fun, to remind them of the things we did the day before with no malicious motives.” Internet communication, often lacking immediate feedback from the recipient, may lead to unintentional humiliation: “Pictures I have posted seemed funny until I found out the other person was embarrassed” (Caucasian female, 20).

Electronic Aggression, Risky Families, Supportive Friendships, and Emotion Regulation

Ordinary least squares regression analyses showed that family risk was significantly related to friend victimization, β = .23, t(215) = 3.39, p = .001; dating partner victimization, β = .27, t(215) = 4.09, p < .001; and dating partner perpetration, β = .24, t(215) = 3.61, p < .001, when including gender as a covariate. Family risk was not related to friend perpetration, β = .12, t(215) = 1.78, p = .08.

Next, we conducted eight hierarchical multiple regression analyses to examine whether either support from friends or emotion regulation moderated the link between risky family environments and each of the electronic aggression variables. For each analysis, main effects of family risk, the moderator (support from friends or emotion regulation), and gender were entered in the first step, the two-way interactions in the second step, and the three-way Family Risk × Support From Friends/ Emotion Regulation × Gender interaction in the third step. Variables were grand mean-centered to reduce nonessential collinearity and increase the ease of interpretation of the intercept. Gender was not significant in the three-way interaction, two-way interactions, or as a main effect and thus was included as a covariate rather than a moderator in the final analyses.

Table 4 presents the regression analyses testing support from friends as a moderator in the association between family risk and electronic aggression and Table 5 presents parallel analyses for emotion regulation. There were significant 2-way interactions between family risk and support from friends for friend victimization, β = −.24, t(211) = −3.48, p = .001; friend perpetration, β = −.16, t(211) = −2.29, p = .02; and dating partner victimization, β = −.20, t(211) = −2.97, p = .003, but not for dating partner perpetration. Similarly, significant interactions were found between family risk and emotion regulation for friend victimization, β = −.16, t(213) = − 2.50, p = .01; friend perpetration, β = −.14, t(213) = −2.04, p = .04; and dating partner victimization, β = −.21, t(213) = −3.41, p = .001, but not for dating partner perpetration.

Table 4.

Hierarchical Multiple Regression Analyses Examining Support From Friends as a Moderator Between Family Risk and Electronic Victimization and Perpetration With Friends and Dating Partners.

	Friend		Dating Partner
Predictor	ΔR2	β	ΔR2	β
Victimization
Step 1	.01		.01
Gender		−.05		−.06
Step 2	.06**		.09***
Family risk		.11		.15*
Friend support		−.14		−.17*
Step 3	.05**		.04**
Family Risk ×		−.24**		−.20**
Friend Support
Total R2	.12*** (.10)		.14*** (.12)
Perpetration
Step 1	.01		.001
Gender		−.09		.07
Step 2	.02		07***
Family risk		.06		.16*
Friend support		−.05		−.16*
Step 3	.02*		.01
Family Risk ×		−.16*		−.08
Friend Support
Total R2	.05* (.03)		.08** (.06)

Note. All βs represent standardized coefficients at the final step. ΔR² represents change in variance accounted for at each step, with * next to ΔR² representing significance in model improvement and * next to “Total R²” representing final model significance. Values in parentheses represent adjusted R² values. df = 4, 211.

^*p < .05.

^**p < .01.

^***p < .001.

Table 5.

Hierarchical Multiple Regression Analyses Examining Emotion Regulation as a Moderator Between Family Risk and Electronic Victimization and Perpetration With Friends and Dating Partners.

	Friend		Dating Partner
Predictor	ΔR2	β	ΔR2	β
Victimization
Step 1	.01		.01
Gender		−.08		−.09
Step 2	.07**		.12***
Family risk		.17*		.18**
Emotion regulation		−.13		−.21**
Step 3	.03*		.05**
Family Risk ×		−.16*		−.21**
Emotion Regulation
Total R2	.10*** (.09)		.17*** (.16)
Perpetration
Step 1	.01		.001
Gender		−.10		.05
Step 2	.04*		.10***
Family risk		.06		.16*
Emotion Regulation		−.16*		−.22**
Step 3	.02*		.01
Family Risk ×		−.14*		−.09
Emotion Regulation
Total R2	.07** (.05)		11*** (.09)

Note. All βs represent standardized coefficients at the final step. ΔR² represents change in variance accounted for at each step, with * next to ΔR² representing significance in model improvement and * next to “Total R²” representing final model significance. Values in parentheses represent adjusted R² values. df = 4, 213.

^*p < .05.

^**p < .01.

^***p < .001.

Figure 1 displays simple slopes of these interactions for each analysis, which were probed using procedures recommended by Aiken and West (1991). The relationship between family risk and electronic aggression is plotted at 1 SD above the mean and 1 SD below the mean of the moderator—either support from friends or emotion regulation. Wald tests were used to test the significance of slopes. When support from friends was the moderator (Figure 1a), family risk and electronic aggression were significantly positively associated with low levels of support, β = .35, F(1, 211) = 17.80 for friend victimization; β = .22, F(1, 211) = 6.54 for friend perpetration; and β = .35, F(1, 211) = 17.91, for dating partner victimization (all ps < .05) but not at high levels of support from friends, β = −.13, F(1, 211) = 1.56 for friend victimization; β = −.11, F(1, 211) = .95 for friend perpetration; and β = −.05, F(1, 211) = .36 for dating partner victimization (all ps > .05). For dating partner perpetration, there were significant main effects of family risk β = .16, t(211) = 2.10, p = .04, and support from friends, β = −.16, t(211) = −2.11, p = .04, with higher family risk and lower support from friends associated with higher dating partner perpetration. The Family Risk × Support From Friends interaction was not significant, β = −.08, t(211) = −1.21, p = .23.

Figure 1.

Associations between family risk and electronic aggression by low and high levels of support from friends and emotion regulation.

As illustrated in Figure 1b, a nearly identical pattern of results emerged when emotion regulation was the moderator. Simple slopes examining the three significant interactions suggested associations between family risk and electronic aggression at low but not high levels of emotion regulation. At low levels of emotion regulation, β = .33, F(1, 213) = 13.36 for friend victimization; β = .19, F(1, 213) = 4.28 for friend perpetration; and β = .39, F(1, 213) = 20.20 for dating partner victimization (all ps < .05). In contrast, at high levels of emotion regulation, β = .003, F(1, 211) = .01 for friend victimization; β = −.08, F(1, 213) = .59 for friend perpetration; and β = −.04, F(1, 213) = .11 for dating partner victimization (all ps > .05). Again, for dating partner perpetration, there were main effects of family risk, β = .16, t(211) = 2.34, p = .02, and emotion regulation, β = −.22, t(211) = −3.25, p = .001, but no interaction, β = −.09, t(211) = −1.36, p = .18.

Discussion

Emerging adulthood is a period of relationship development and transitions, which may introduce the risk for considerable interpersonal conflict. Cell phones and social networking sites offer a convenient way for emerging adults in college to communicate or convey conflict in relationships. Studying aggression through electronic media may provide unique insights into emerging adults’ maladaptive relationships as well as how they express aggression. Thus, this study sought to identify factors contributing to and protecting against electronic aggression. Findings indicate that electronic aggression is common and should not be overlooked among emerging adults. Participants reported more electronic aggression with friends than with dating partners. Males reported more victimization than perpetration overall, whereas females reported more victimization than perpetration with friends, not dating partners. Qualitative analyses suggest that emerging adults offer a diverse array of motivations for their electronic perpetration, ranging from goodwill to hatred. The most commonly found motivation for electronic perpetration was jealousy/insecurity, which emerged more often in females’ than in males’ responses. Females were also more likely to attribute their electronic aggression to desired self-protection and privacy compared to males. Males were more likely than females to indicate that they were having fun and being humorous. Overall, early risky family environments were associated with electronic victimization and perpetration, but the links between family risk and electronic aggression were not significant for those with high support from friends and emotion regulation.

Importance of Electronic Victimization and Perpetration

Several noteworthy patterns were found in the present study. First, there was a moderate positive correlation between electronically aggressive behaviors in friendships and dating relationships. Second, in line with electronic aggression research with adolescents (Ybarra & Mitchell, 2004), results of this study suggest a high association between electronic victimization and perpetration. Studies often focus exclusively on either electronic victimization or electronic perpetration, but the relation between the two suggests the need to investigate the reciprocal nature of electronic aggression.

Overall, participants reported higher mean rates of electronic victimization than perpetration, suggesting that participants may be less likely to recognize the messages they send as aggressive but more likely to perceive the online behaviors of others as hurtful. Lacking facial and vocal cues, text messages, and social network website postings may carry ambiguous meanings, thus, being differentially perceived by the recipient and sender. The qualitative analyses illustrate that some students engaging in electronic acts labeled as aggressive may not actually ascribe hostile intentions to their behaviors. Thus, despite previous findings that victims described these behaviors as distressing (Bennett et al., 2011), individuals who acknowledged perpetrating the behaviors sometimes indicated that they did not intend to cause harm. Alternatively, when providing retrospective explanations of their own behaviors, perpetrators might have minimized their intended harm.

Electronic Perpetration Motivations

The electronic aggression literature has paid scant attention to perpetrators’ motives for aggression. As demonstrated here, motivations for electronic perpetration can range in valence from negative to neutral to positive. Analyses hinted at the possibility for a discrepancy between the actor’s intention and the recipient’s reaction, supporting previous suggestions that even behaviors intended as jokes might have negative consequences (Sears, Byers, Whelan, & Saint-Pierre, 2006). Investigating the consequences of electronic aggression within the context of the varied motivations identified in this study could increase the understanding of situations under which electronic aggression might have the most harmful impact for emerging adults.

Family Risk and Emerging Adult Electronic Aggression

The association between risky family environments and electronic aggression with friends and dating partners extends intergenerational transmission findings that aggression in the family-of-origin spills over to young adult relationships (e.g., Linder & Collins, 2005). Individuals raised in hostile, unsupportive families may have social and emotional deficits, placing them at risk for using aggressive and maladaptive strategies when faced with potential tension in relationships (Repetti, Taylor, & Seeman, 2002; Taylor et al., 2004), as well as for receiving aggression in either face-to-face or electronic settings. Growing up in aggressive families may also sensitize individuals to conflict. Thus, participants with riskier family environments could be more likely to label ambiguous electronic messages as aggressive or hurtful and to respond with aggression.

Yet, strong support from friends and good emotion regulation skills appear to moderate the link between family environment and three of the four electronic aggression variables: friend victimization, friend perpetration, and dating partner victimization. Positive support from friends and emotion regulation skills may characterize individuals who are socially competent, who have close interpersonal relationships, and who show low levels of aggression (Eisenberg, et al., 1997; Repetti et al., 2002). These attributes may be important in the rapid-fire context of electronic communication. For example, good emotion regulation skills might enable an individual to step back from an escalating electronic negative exchange. Similarly, having confidence in the good nature and trust of friends could lead to more benign interpretations of ambiguous messages via the Internet and text messages and a greater willingness to take the perspective of others, thereby circumventing an aggressive response.

The mitigating effects of social support and emotion regulation were not found for electronic dating perpetration, despite direct associations between dating partner perpetration and risky family environments, social support, and emotion regulation. Thus, variables that serve as protective factors against other forms of electronic aggression do not play a similar role with respect to dating partner perpetration. Interestingly, in another comparison of electronic victimization from friends and dating partners, Bennett et al. found that higher levels of upset were associated with electronic victimization from dating partners compared to friends. Electronic perpetration thus may be particularly disturbing when it occurs in the context of an intimate relationship, especially if the behaviors violate trust or expose a partner’s vulnerabilities.

Limitations and Future Directions

This study contains several limitations. Although we collected frequencies of electronic aggression separately for dating relationships and friendships, we did not differentiate between relationship type when inquiring about motivations. Future studies should investigate whether certain motivations are differentially linked to electronic aggression in friendships or dating relationships. Although we focused on electronic aggression with known others to test intergenerational hypotheses, a topic deserving of future investigation is the extent to which emerging adults use electronic aggression to make fun of acquaintances or to anonymously commit electronic “pranks” that ultimately have serious consequences (Dilmac, 2009; Held, 2011). We studied emerging adults in college, a group with frequent use of electronic communication. Future research that examines electronic aggression among emerging adults outside of a 4-year university can help present a more generalizable picture of the role that electronic aggression plays among this heterogeneous group.

With all data collected in this study via self-report, which potentially inflates associations due to shared method and reporter variance, alternative methodologies are worthy of future consideration. A promising next step would be to collect data from two members of a dyad to assess whether there is correspondence in the labeling of electronically aggressive events and in the motivations attributed to such aggression. Dyadic assessments on a more frequent basis, such as through ecological momentary assessment (Shiftman, Stone, & Hufford, 2008) could improve assessment by minimizing recall problems and by providing an index of perpetrator-victim agreement about specific events within a specific time frame. Ecological momentary assessment may be particularly useful for measuring communication through electronic means, such as text messaging, which tends to occur in rapid bursts throughout the day. Finally, direct examination of participants’ text messages and social networking website profiles (Underwood, Rosen, More, Ehrenreich, & Gentsch, 2012) could provide more ecologically valid data.

Summary

As emerging adults continue to function in a digital age, it is necessary to understand maladaptive online patterns in relationships in order to identify factors that may prevent or assist salient developmental tasks during this stage. Findings suggest that aggression through electronic media is quite common for emerging adults and is not a phenomenon unique to adolescents. Risk and protective factors associated with emerging adults’ electronic aggression are not dissimilar from those associated with face-to-face aggression. Yet, the interface between online and offline communications with emerging adults has yet to be investigated. Some electronic aggression may simply be an extension of offline relationship tensions. However, it is also possible that electronic aggression may instigate new or additional relationship problems for emerging adults, which may have long-term implications for their transition to adulthood.

Biographies

Author Biographies

Ilana Kellerman is a doctoral student in clinical psychology at the University of Southern California. Her research focuses on adolescent and emerging adult relationships and mental health.

Gayla Margolin is a professor of psychology and pediatrics at the University of Southern California. Her National Institutes of Health (NIH)-Eunice Kennedy Shriver National Institute of Child Health and Human Development (NIH-NICHD)-funded work examines intergenerational continuity of aggression from childhood experiences to adolescents and young adults’ friendships and dating relationships.

Larissa A. Borofsky is a doctoral student in clinical psychology at the University of Southern California. Her research uses functional magnetic resonance imaging and an experimental paradigm to explore emotion regulation in violence exposed youth. She is a National Science Foundation Graduate Research fellow.

Brian R. Baucom is an assistant professor of psychology at the University of Utah. His research focuses on behavioral and emotional processes during couple interaction and on couple therapy.

Esti Iturralde is a doctoral student in clinical psychology at the University of Southern California where she investigates adolescent friendship and its role in health behavior. She is an NIH-NICHD National Research Service Award fellow.