Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Nov 1.
Published in final edited form as: Psychol Addict Behav. 2022 Feb 3;36(7):815–823. doi: 10.1037/adb0000818

Proximal Associations among College Students’ Alcohol Use and Cyber Partner Abuse Perpetration

Meagan J Brem 1, Ryan C Shorey 2, Jim McNulty 3, L Christian Elledge 4, Jeff Temple 5, Gregory L Stuart 4
PMCID: PMC9346098  NIHMSID: NIHMS1770529  PMID: 35113587

Abstract

Objective:

Research and theory support alcohol use as a proximal antecedent to in-person partner abuse (PA). However, event-level research has not examined cyber PA thereby limiting our understanding of whether alcohol use proximally relates to cyber PA.

Method:

We collected daily data on alcohol use and cyber PA from college students (N = 236; 73.3% women) for 60 consecutive days. Controlling for cyber PA victimization, we evaluated whether college students who consumed more drinks perpetrated more cyber PA (between-person effects), whether cyber PA was more likely to occur on days in which alcohol use was higher than each individual’s average (within-person effect), and whether within- and between-person associations between alcohol use and cyber PA varied by sex.

Results:

Women were more likely than men to perpetrate cyber PA but there were no sex differences in the association between alcohol use and cyber PA. Multilevel modeling revealed that neither higher average alcohol use, nor drinking more than one usually does on a given day, associated with odds of subsequent cyber PA. Although alcohol use did not associate with odds of subsequent cyber PA, post-hoc analyses revealed that odds of cyber PA increased as alcohol use increased, regardless of whether drinking occurred before or after cyber PA. Thus, alcohol use may have been more likely to occur after cyber PA.

Conclusions:

Results did not support alcohol use as a proximal antecedent to college students’ cyber PA. Future research should investigate of cyber PA as a proximal risk factor for subsequent alcohol use.

Partner abuse (PA) peaks between the ages of 18 and 25 (Johnson et al., 2015), a time in which many young adults are in college. With the advent and ubiquity of smartphones and social media, college couples have virtually constant access to one another thereby creating new contexts for individuals to monitor, harass, threaten, humiliate, and otherwise abuse their partners via technology, otherwise known as cyber PA (Elphinston & Noller, 2011; Borrajo et al., 2015; Leisring & Giumetti, 2014). Approximately 44% of college men and 55% of college women perpetrate at least one form of cyber PA each year which may increase their partners’ likelihood of experiencing deleterious mental health consequences (e.g., depression, suicidal ideation, and risky sexual behaviors; Brem et al., 2019a; van Ouytsel et al., 2016; Wolford-Clevenger et al., 2016a, 2016b). Despite the prevalence and negative experiences associated with cyber PA, theory and research in this area remain underdeveloped.

Substantial evidence suggests that alcohol consumption contributes to in-person PA perpetration at a global and acute level (Foran & O’Leary, 2008; Leonard & Quigley, 2017). Acute alcohol use is purported to influence aggression by focusing cognitive and attentional resources on the most emotionally salient contextual cues. For instance, if relationship conflict or a perceived threat to the relationship occurs while intoxicated, then attentional resources are directed toward that stimulus and psychological state at the expense of prosocial, aggression-inhibiting stimuli (e.g., the consequences of aggression, prosocial conflict resolution strategies), thereby increasing the likelihood of aggression (Giancola et al., 2010; Steele & Josephs, 1990). In support of the notion that alcohol use confers risk for PA, college students’ physical and sexual PA perpetration is more likely to occur on a drinking day relative to a non-drinking day, on a heavy-drinking day, and as the number of drinks consumed increases on a given day, and psychological PA is more likely to occur on a heavy-drinking day relative to a non-heavy-drinking day (Moore et al., 2011; Shorey et al., 2014a, 2014b). The proximal relationship between alcohol and PA may extend to cyber PA.

Investigations of cyber PA risk factors among college students have only recently emerged with mixed findings on the association between alcohol and cyber PA. Notably, only alcohol problem severity (assessed via the Alcohol Use Disorders and Identification Test; Saunders et al., 1993) has been examined in relation to cyber PA perpetration, thereby precluding conclusions regarding the association between alcohol use and cyber PA. Cross-sectional data collected by Brem et al. (2019a) supported a positive association between college students’ alcohol problem severity and cyber PA perpetration whereas Watkins et al. (2020) found that alcohol problem severity positively related to only some forms of cyber PA (e.g., using technology to emotionally harm a partner) but not others (e.g., cyber monitoring, sexual coercion/abuse perpetrated via technology). Similarly, cross-sectional, crowdsourced data collected from adults revealed that alcohol problem severity positively related to some, but not all, forms of cyber PA (Crane et al., 2018; Watkins et al., 2016). In the only longitudinal examination of college students’ alcohol problem severity and cyber PA, Brem et al. (2019b) reported that alcohol problem severity positively associated with cyber PA perpetrated three months later.

Together, these studies support an association between alcohol use and cyber PA; however, the proximal associations between alcohol use and cyber PA remain unknown. It is plausible that the cross-sectional and longitudinal positive associations between alcohol problem severity and cyber PA exist because alcohol use precedes cyber PA. A theory-informed investigation of the potential risk alcohol use poses for subsequent cyber PA will (1) identify whether alcohol use should be targeted in programs aiming to reduce cyber PA, and (2) contribute to a conceptual model of alcohol-facilitated cyber PA. Toward this end, the present study used daily diary data to examine proximal associations among alcohol use and cyber PA perpetration among college students.

Aims and Hypotheses

For the present study, we collected data across 60 days from college students in intimate relationships wherein individuals completed once-daily surveys assessing their alcohol use and cyber PA. Ecological momentary assessment (e.g., daily diary reporting) offers several benefits not offered by experimental or cross-sectional methodologies, including greater ecological validity, reduced error and bias associated with retrospective reporting, and greater understanding of how participants’ experiences and behaviors vary over time (Shiffman et al., 2008). Considering that an individual can both perpetrate and experience victimization in one PA episode (Derrick & Testa, 2017), we tested the proximal association between alcohol use and cyber PA perpetration after statistically controlling for cyber PA victimization in analyses. Specifically, we hypothesized that, controlling for cyber PA victimization, (1) odds of cyber PA perpetration would be greater on days in which alcohol use was higher than each individual’s average (within-person effect), and (2) college students with higher average alcohol consumption would be more likely to perpetrate cyber PA relative to those with lower average alcohol consumption (between-person effect). As an exploratory aim we examined whether the hypothesized within- and between-person associations between alcohol use and cyber PA varied by sex. In the absence of data or theory to support sex differences, no a priori hypotheses were made regarding sex differences.

Method

Participants

Undergraduate students (N = 236) from a large, public, southeastern university participated in the present study. Most participants (75.1%) reported that their sex assigned at birth was female. With regard to gender identity, 67.7% of participants identified as women, 22.5% identified as men, and 0.4% (one person) identified as non-binary; 9.4% of participants did not report their gender identity. Sex assigned at birth was used to examine sex differences in analyses.

To be eligible, participants had to (1) be at least 18 years of age, (2) be in a current dating relationship that lasted for at least one month in duration, (3) have a dating partner who was 18 years of age or older, (4) have consumed alcohol in the past month, and (5) have a minimum of two contact days with their dating partner each week (e.g., face-to-face contact or technology-facilitated contact). Men and women recruited for the present study were not dating each other. Three hundred forty-nine undergraduate students participated in a screener to determine eligibility. Of those screened, 343 (98.28%) were eligible to participate in the study, completed a baseline survey, and were sent daily surveys. Seventy-three percent (n = 251; 72.5% women) of the baseline sample completed at least one daily survey. Fifteen participants (5.98%) were excluded from the final sample due to having less than three daily data points, which was the minimum number of data points needed to conduct analyses. Thus, the final sample consisted of 236 participants.

The average age of participants was 20.53 (SD = 3.31) years. The racial composition was 83.89% white; 5.51% African American or Black; 4.24% Hispanic or Latino(a); 2.97% Asian; 1.69% multiracial; 0.42% American Indian, Native American or Alaskan Native; and 0.42% Middle Eastern or Indian; two participants (0.85%) did not report their race. Most participants (78.81%) identified as heterosexual followed by bisexual (13.56%), gay (1.27%), lesbian (1.27%), queer (1.27%), and pansexual (0.43%); eight participants (3.39%) did not report their sexual orientation. Most participants reported they were dating their partner (92.79%) though some reported they were engaged (4.24%) or married (2.97%). The average relationship length was 18.83 (SD = 18.04) months (Mdn = 12.00 months). Most participants (50.85%) endorsed living with friends or roommates, followed by living with their partner (24.15%), living alone (15.68%), living with family (5.51%), or other living arrangements (3.39%). One participant (0.42%) did not provide details regarding living arrangements. Regarding annual family income, the distribution was as follows: Less than $50,000 (39.83%), $50,000 to $100,000 (22.88%), $100,000 to $150,000 (13.13%), $150,000 to $200,000 (7.20%), and greater than $200,000 (7.63%); 22 participants (9.32%) did not know their annual family income or chose not to report their annual family income. Participants reported spending an average of 4.62 (SD = 2.35) hours on the internet each day.

Procedure

Participants were undergraduate students recruited via a university research participation website of which students were made aware through their psychology courses and by flyers posted across campus. Interested participants were asked to complete a brief screening measure on Qualtrics.com to determine eligibility. Eligible participants were sent an email stating that they were eligible to participate in an optional research study that would require them to complete a baseline assessment of self-report measures followed by daily questionnaires for 60 consecutive days; participants would receive partial course credit (if they were enrolled in an eligible introductory psychology course) or $20.00 monetary compensation after completing the baseline questionnaire. Eligible participants completed the baseline assessment to assess background information, including dating history and demographics. The baseline assessment was completed using Qualtrics.com.

After this baseline assessment, participants were asked to complete a daily questionnaire for 60 consecutive days. Each participant was emailed a link to the 5-minute daily questionnaire each day at 6:00AM. Each questionnaire asked participants to report about their previous day’s behavior, defined for participants as the time they awoke until the time they went to sleep, consistent with prior daily diary work (Shorey et al., 2014a,b). Participants who did not complete their daily survey received two reminder emails each day (i.e., one at 12:00 PM and another at 5:00 PM). Participants received $1.00 per daily survey completed for 60 consecutive days. Participants were compensated an additional $5.00 per week that they completed all 7 surveys. Thus, participants had the opportunity to earn up to $120.00 for participating in the present study.

Measures

Demographics.

Demographic data assessed at baseline included age, sex, gender identity, racial/ethnic identity, sexual orientation, academic level, family income, relationship status, relationship length, current living arrangements, and average length of time spent on the internet each day.

Daily Measures.

Relationship Status.

Participants provided information related to their relationship status each day. Participants were asked if they were in the same relationship that they reported at baseline. All participants were asked to complete all questions regarding their baseline partner as aggression often continues and increases in frequency and severity even after the relationship ends (Anderson & Saunders, 2003).

Partner Contact.

Participants reported the level of contact (i.e., face-to-face, phone, text messaging, and internet) with their partner each day.

Cyber Partner Abuse.

Participants were asked if they perpetrated cyber PA, were victimized by cyber PA, or both perpetrated and were victimized by cyber PA during the previous day using items adapted from the Controlling Partner Inventory (CPI; Burke et al., 2011), the Electronic Aggression Scale (Bennett et al., 2011), and the Facebook Mate-Retention Tactics Inventory (Brem et al., 2015). Each of the scales from which cyber PA acts were chosen evidenced good internal consistency and criterion validity and have been used in multiple studies assessing cyber PA among college-aged adults (Bennett et al., 2011; Brem et al., 2015, 2019a; Burke et al., 2011; Lopes et al., 2017; Ramos et al., 2021). Cyber PA acts were chosen from each of these validated measures to assess various forms of abuse, harassment, humiliation, monitoring, coercing, and threatening that occur in cyber contexts. Each of the six items used to assess daily cyber PA listed several cyber PA tactics. For example, one item asked whether a participant perpetrated any of the following: took a video or photo of a partner to embarrass them; threatened to post inappropriate photos or videos of partner online; posted embarrassing, insulting, or inappropriate material (e.g., posts, photos, or videos) about partner online (e.g., on a social networking site or blog). Other examples of cyber PA acts listed in items included “Sent mean, hurtful, or threatening emails, text messages, or online messages (e.g., on a social networking site or blog) to partner,” “Told partner to block, unfriend, or unfollow someone on social media; threatened partner that something bad would happen if s/he interacted with a certain person online,” and “Snooped through partner’s sent/received call histories, text messages, or online messages.” Responses were dichotomized such that if a participant perpetrated any of the cyber PA perpetration acts listed in a single item, they responded in the affirmative to the item (i.e., scored “1”). If a participant responded “yes” to any of the six cyber PA perpetration items, that participant’s overall cyber PA perpetration score for that day was coded “1;” if no cyber PA perpetration items were endorsed for the day (i.e., a participant responded “no” to all six cyber PA items), that participant’s overall cyber PA perpetration score for the day was coded “0.” Cyber PA victimization was assessed daily to include as a covariate. Participants’ cyber PA victimization score for the day was coded “1” if participants reported any cyber PA victimization (i.e., if they responded affirmatively that their partner had engaged in the cyber PA behaviors toward them, or if they responded affirmatively that the participant and their partner both engaged in the behavior) and “0” if no cyber PA victimization was endorsed.

Additionally, participants were asked to report how many instances of cyber PA perpetration occurred each day. Because cyber PA can occur multiple times throughout the day, participants were also asked if alcohol was consumed before any cyber PA event that occurred on a given day (i.e., scored as yes-no) for the purpose of providing descriptive data (i.e., participants’ response to this item was not used in hypothesis testing).

Alcohol Use.

Each day, all participants were asked if they consumed alcohol during the previous day, and if so, how many standard drinks they consumed. Participants were provided with a definition of a standard drink based on guidelines proposed by the National Institute on Alcohol Abuse and Alcoholism (NIAAA; 2010). If participants reported that they consumed alcohol on a day when cyber PA occurred, they were also asked to report whether they consumed alcohol prior to the initial cyber PA event and, if so, how many drinks they consumed prior to this first cyber PA event. If drinking only occurred after the first cyber PA event, the alcohol use variable was recoded to zero to better approximate drinking that occurred before cyber PA; this was done to prevent confounding drinking that occurred after cyber PA with drinking that occurred prior to cyber PA. If an individual reported drinking before and after cyber PA, only alcohol consumed before the first cyber PA event was included in the count index of number of drinks consumed. Because cyber PA can occur several times throughout the day, we aimed to determine whether alcohol preceded the onset of cyber PA and therefore assessed number of drinks consumed before the first PA event.

To further clarify, if a participant reported consuming three drinks on a given day but did not perpetrate cyber PA, their number of drinks was coded as “3.” If a participant reported consuming 6 drinks on a given day, but only three were consumed prior to the first cyber PA event, that participant’s number of drinks was coded “3.” If a participant reported consuming 6 drinks on a given day, but all were consumed after cyber PA perpetration, then the participant’s number of drinks total was coded “0.” If a participant did not consume any alcohol on a given day their number of drinks for the day was coded “0.” This alcohol coding approach was previously used in daily diary investigations of alcohol and PA (Shorey et al., 2014a, 2014b).

Data Analytic Strategy

Descriptive analyses were completed using SPSS Version 24.0. Hypotheses were tested with multilevel modeling using HLM Version 8.0. Compliance rates were computed by calculating the percentage of missed surveys in relation to total surveys.

Multilevel modeling using fixed slopes specifying a logit link function with a Bernoulli sampling distribution and full maximum likelihood estimation was used for analyses because the multilevel nature of these models accounts for the correlated errors associated with individuals’ repeated reports. Despite some missing data due to attrition and noncompliance, all participants were used in analyses because the multilevel modeling procedure uses all available data and estimates model parameters according to Bayesian rules. All available data were used including days when cyber PA did not occur, days when drinking did not occur, and days when participants did not have direct contact with their partner because cyber PA can occur in the absence of direct partner contact (e.g., monitoring a partner using technology, snooping through a partner’s private online messages).

Two-level models were employed. Alcohol use (a time-varying predictor) was centered on each individual’s overall mean (i.e., person-mean centered) to yield Level 1 within-person effects. We calculated each individual’s overall mean for alcohol use, centered the mean on the grand mean, and included these values in Level 2 to provide between-person effects for alcohol use. This approach allowed us to examine within-person effects controlling for between-person effects to better approximate proximal associations among alcohol use and cyber PA. Cyber PA victimization was included as a level 1 covariate in all analyses. Sex (dummy coded; women = 0, men = 1) was included as a cross-level moderator in all analyses; if no significant cross-level interactions emerged, sex remained in the model as a level 2 covariate. We also computed a sex x overall mean alcohol use variable (grand-mean centered) and included this variable in level 2 to examine sex as a moderator of the between-person effects of alcohol use and cyber PA perpetration. If this interaction term was not significant, it was removed to test the main effects model.

We tested hypotheses by estimating (1) the between-person effects of sex, alcohol use (grand-mean centered), and their two-way interaction on repeated measures of cyber PA, all entered in level 2, and (2) the within-person effects of alcohol use (person-mean centered, entered in level 1) and it’s cross-level interaction with sex on repeated measures of cyber PA perpetration. Dichotomously-scored (i.e., 0 = no victimization, 1 = victimization) cyber PA victimization was included in level 1 to statistically control for cyber PA victimization in analyses given that prior daily diary research indicated that an individual may both perpetrate and be victimized by PA in a PA episode (Derrick & Testa, 2017).

Results

Descriptive Statistics

Compliance rates.

Of the 236 people who completed at least three daily surveys, 14,160 surveys were sent, and 10,149 surveys were completed for an overall compliance rate of 71.67%. Participants completed an average of 42.57 daily surveys (SD = 20.09). Individuals’ number of daily surveys completed was not significantly associated with the average number of drinks an individual consumed each day, but age positively associated with number of daily surveys completed (r = .15, p = .03). Number of daily surveys completed did not significantly relate to frequency of cyber PA perpetration (r = .07, p = .27). Number of daily surveys completed did not significantly differ between men (M = 40.80, SD = 20.11) and women (M = 43.09, SD = 20.12).

Daily data descriptive statistics.

Of all the daily surveys completed (N = 10,149), there were 189 instances of cyber PA perpetration that occurred on 175 days (1.73% of the daily survey data). Cyber PA perpetration was reported by 74 participants (31.3% of participants). Of those who perpetrated cyber PA at least once during the study, there was an average of 2.36 days that included cyber PA perpetration. Frequency of discrete cyber PA acts perpetrated in one day ranged from 0-30. Sixty-one days included cyber PA victimization, 24 of which occurred on a day that also included cyber PA perpetration (i.e., 13.7% of cyber PA perpetration days included cyber PA victimization). Thirty-seven days included cyber PA victimization but not cyber PA perpetration (i.e. .37% of days that did not include cyber PA perpetration).

Most (83.5%) participants had at least one drinking day during their participation. On average, participants reported consuming alcohol on an average of 6.35 days during the study; however, it should be noted that not all participants completed surveys on all 60 days. Participants consumed alcohol on an average of 14.4% of days in which they completed surveys. On drinking days, participants reported drinking an average of 3.36 (SD = 2.79) standard drinks.

Co-occurrence of alcohol use and cyber partner abuse.

Participants reported drinking alcohol on 1,498 days; 1,022 days of alcohol use occurred on days in which the participant had face-to-face contact with their partner (i.e., 17.1% of all partner contact days included alcohol use). Participants consumed an average of .48 drinks (SD = 1.43) before cyber PA perpetration. Of the 189 instances of cyber PA perpetration, alcohol was consumed before the first cyber PA event on 16 days (8.47%). Because cyber PA can occur multiple times throughout the day, participants were also asked if alcohol was consumed before any cyber PA event that occurred on a given day. Alcohol was consumed before at least one cyber PA event on 49 days (25.94%). Alcohol was not consumed at all on 130 days in which cyber PA occurred (77.84% of cyber PA days). When examining only drinking days, the average number of drinks consumed before cyber PA (M = 2.67; SD = 2.38) was comparable to the number of drinks consumed on a non-perpetration drinking day (M = 3.36, SD = 2.77), t(1472) = 1.35, p = .18.

Sex differences.

Examination of sex differences in the primary daily study variables revealed that women reported perpetrating cyber PA on more days (2.00%) than did men (0.89%), χ2(1) = 12.28, p < .001. The total number of drinking days did not differ between men (M = 5.60, SD = 8.01) and women (M = 6.74, SD = 8.28), F(233) = 1.08, p = .32. Men reported drinking a greater number of standard drinks on drinking days (M = 2.94, SD = 0.15) than did women (M = 1.45, SD = 0.04), F(233) = 207.10, p < .001. Examining only days on which cyber PA perpetration occurred, men consumed an average of .73 drinks (SD = 2.34) before cyber PA perpetration, which was not significantly greater than the average number of drinks women consumed before cyber PA perpetration (M = .45; SD = 1.25), F(173) = 3.84, p = .37.

Hypotheses Tests

Results of the initial model (Model 1; see Table 1) revealed that neither the between-person nor within-person association between alcohol use and cyber PA perpetration odds significantly varied as a function of sex. After removing sex x alcohol use interactions for Model 2, results indicated that, controlling for sex and cyber PA victimization, neither the between-person nor within-person association between alcohol use and cyber PA perpetration were significant.1 The main effect of sex emerged as a significant such that men’s odds of cyber PA perpetration were less than women’s odds. Cyber PA victimization positively associated with cyber PA perpetration odds. Table 1 displays parameters for Models 1 and 2.

Table 1.

Parameters of models predicting odds of daily cyber partner abuse perpetration

Model 1
 Model 2
Variable B SE OR 95% CI B SE OR 95% CI
Level 2 (between-person effects)
 M Alcohol use −.02 .23 .98 (.70, 1.36) .02 .11 1.02 (.81, 1.28)
 Sex −.82*** .24 .44 (.28, .70) −.77*** .22 .46 (.30, .71)
 Alcohol x sex .09 .24 1.09 (.70, 1.70)
Level 1 (within-person effects and cross-level interactions)
 CPA vict 2.73*** .58 15.30 (4.95, 47.34) 2.72*** .57 15.20 (5.02, 45.98)
 Alcohol use −.01 .03 .99 (.92, 1.06) −.02 .04 .98 (.91, 1.06)
 Alcohol x sex .05 .06 1.06 (.94, 1.19)
Open in a new tab
***

p < .001

Note. M alcohol use = overall average alcohol use, grand-mean centered. Alcohol use = # drinks consumed prior to cyber partner abuse perpetration, person-mean centered. CPA vict = daily cyber partner abuse victimization, grand-mean centered.

Post-Hoc Analyses

Post-hoc analyses were conducted to highlight the importance of testing temporal ordering when examining alcohol use and cyber PA perpetration. Specifically, we examined whether odds of cyber PA increased as the number of drinks consumed on a given day increased beyond each individual’s average alcohol use, regardless of whether drinking occurred before or after cyber PA perpetration. Controlling for cyber PA victimization (level 1, grand-mean centered) and the between-person effects of overall mean alcohol use (level 2, grand-mean centered), results revealed that the relation between number of drinks consumed (person-mean centered) and odds of cyber PA perpetration did not vary by sex (see Model 3 in Table 2). Removing sex as a cross-level moderator in Model 4 revealed that the odds of cyber PA perpetration increased as the number of drinks consumed on a given day increased beyond each individual’s average (see Table 2). When considered with the null association between drinking prior to cyber PA and cyber PA perpetration, this finding suggests that cyber PA perpetration may positively relate to odds of subsequent drinking. Women’s cyber PA perpetration odds were greater than men’s odds, and cyber PA victimization positively related to odds of cyber PA perpetration.

Table 2.

Parameters of post-hoc models predicting odds of daily cyber partner abuse perpetration

Model 3
 Model 4
Variable B SE OR 95% CI B SE OR 95% CI
Level 2 (between-person effects)
 M Alcohol use .00 .11 1.00 (.81, 1.25) −.07 .11 .93 (.75, 1.15)
 Sex −.79*** .21 .46 (.30, .70) −.77*** .22 .46 (.30, .71)
Level 1
 CPA vict 2.72*** .56 15.25 (5.12, 45.42) 2.72*** .55 15.12 (5.17, 44.21)
 # drinks .08** .03 1.08 (1.02, 1.14) .08** .03 1.08 (1.02, 1.15)
 # drinks x sex .06 .05 1.06 (.96, 1.17)
Open in a new tab
**

p < .01;

***

p < .001

Note. M alcohol use = overall average alcohol use, grand-mean centered.

#

drinks = number of drinks consumed on a given day, including drinks consumed after cyber partner abuse perpetration, person-mean centered. CPA vict = daily cyber partner abuse victimization, grand-mean centered.

Discussion

In the first study to examine the proximal associations between alcohol use and cyber PA perpetration, we failed to support the hypothesis that daily drinking associates with increased likelihood of subsequent cyber PA perpetration (i.e., no significant within-person effect). Similarly, college students with higher average alcohol consumption were not more likely to perpetrate cyber PA than were those with lower average alcohol consumption (i.e., no significant between-person effect). These findings did not vary as a function of participant sex; rather, women’s odds of cyber PA perpetration were greater than men’s odds of cyber PA perpetration. Results also indicated that cyber PA perpetration was more likely to occur on a day that included cyber PA victimization relative to a day in which no cyber PA victimization occurred.

In contrast to data supporting a proximal association between alcohol use and in-person PA perpetration among college students (Moore et al., 2011; Shorey et al., 2014b), and cross-sectional and longitudinal data suggesting that alcohol use positively relates to cyber PA perpetration (Brem et al., 2019a, 2019b; Crane et al., 2018; Watkins et al., 2016, 2020), our results do not indicate that alcohol use is associated with college students’ increased likelihood of subsequent, same-day cyber PA perpetration. Examining aggressogenic traits and affective experiences prior to cyber PA may provide insight into important affective cues that may increase the likelihood of cyber PA after drinking. Indeed, prominent theories of alcohol-related PA (e.g., Finkel & Eckhardt, 2014) which have been extended to cyber PA (e.g., Brem et al., 2019a) purport that alcohol use alone is insufficient in predicting PA perpetration. Rather, the association between alcohol use and cyber PA perpetration may be moderated by contextual instigators (e.g., perceived partner infidelity), dispositional aggressogenic traits (e.g., trait anger, jealousy, or hostility), and couple-level factors (e.g., one’s partner’s alcohol use) which were unexamined in the present study. For instance, alcohol use was unrelated to college women’s psychological and physical PA perpetration on days when angry affect was low, but was positively associated with both forms of PA on days when angry affect was high (Shorey et al., 2014b). Future research should explore whether these contextual moderators extend to cyber PA.

Alternatively, in line with the threshold model for PA (Rosenbaum et al., 1997), it is plausible that alcohol use is not necessary to meet the threshold for perpetrating cyber PA since many cyber PA acts may be considered normative among young adults. In other words, college students may readily perpetrate cyber PA regardless of alcohol use because cyber PA may be seen as a low-risk behavior. In the absence of contextual data for why certain cyber PA acts were performed, it is also possible that some cyber PA behaviors assessed in this study were not perpetrated in an abusive context and therefore would not be more likely to occur after alcohol use. For example, using technology to monitor a partner’s location could be motivated by concern or care for a partner traveling alone at night rather than suspicion that the partner is doing something that would threaten one’s relationship. This line of research would benefit from examining the intent and impact of cyber PA using refined assessments that (1) reduce the likelihood of capturing non-abusive cyber behaviors, and (2) assess a multitude of contextual correlates (e.g., intent and emotionality).

Notably, post-hoc analyses suggested that odds of cyber PA perpetration increased as number of drinks consumed on a given day increased beyond each individual’s average, regardless of whether drinking occurred before or after cyber PA. Our primary analyses were restricted to only examining alcohol use that occurred prior to the onset of cyber PA perpetration, and alcohol use did not associate with odds of subsequent cyber PA perpetration; these post-hoc findings indicate why it is important for investigators to assess the temporal order of alcohol use in relation to cyber PA perpetration to avoid misleading conclusions. In other words, drinking may have occurred after cyber PA perpetration. Indeed, daily diary data revealed that perpetrating in-person PA associated with increased likelihood of alcohol use within the next three hours (Derrick & Testa, 2017). Future research should evaluate the plausibility that cyber PA perpetration may be a risk factor for subsequent drinking among college students.

Although extant data suggests that the prevalence of past-year cyber PA perpetration is comparable across college women and men (Brem et al., 2019a), the present data revealed that women’s odds of daily cyber PA perpetration were greater than men’s odds. It is plausible that cyber PA perpetration is perceived as a more acceptable form of abuse among women relative to men. Alternatively, cyber PA may be a mechanism through which women are better able to assert control within the relationship whereas men may be better able to assert control through in-person PA tactics. Future investigations should evaluate men’s and women’s perceived acceptability of cyber PA and reasons for using cyber PA. Notably, no studies have been published that use couple-level data to investigate actor and partner effects in relation to cyber PA. Couple-level investigations of in-person PA (e.g., Crane & Testa, 2014) and alcohol-related PA (e.g., Testa et al., 2014) suggest that partner-level predictors, which were unexamined in the present study, may better explain men’s PA perpetration. Future research may consider evaluating whether an individual’s cyber PA perpetration associates with their partner’s alcohol use and in-person PA perpetration.

Limitations

Regarding sample-specific factors, the unequal number of men and women within the sample may yield unreliable results with regards to sex differences. The present sample was comprised of primarily white college students, and the percentage of lesbian, gay, bisexual, and pansexual participants in the present sample (17.9%) was greater than that of the general population (estimates range from 4.5% to 11.0%; Gay and Lesbian Alliance Against Defamation, 2019; Newport, 2018). Results may have limited generalizability to other populations. Data collection for this study took place during heightened political and social discussions surrounding violence towards women (e.g., the #MeToo movement), which may have impacted daily reporting. Reporting bias may have impacted endorsement of cyber PA perpetration and victimization. Indeed, college women report higher social desirability scores than do men, and women (but not men) with high social desirability report less PA perpetration and victimization than do women with lower social desirability scores (Bell & Naugle, 2007). Additionally, participants’ M relationship length was 18.83 months, which may mean that the participants in the present study were less representative of college students who are in shorter-term relationships. Approximately 31% of those who completed the baseline survey were not included in the final sample because they completed fewer than three daily surveys. It is plausible that those who were included in the final sample represent a less at-risk population, thereby limiting the generalizability of results.

Regarding data collection methods, participants in the present study were emailed each morning with a link to a survey that assessed their previous day’s experiences. Though this method is presumably less influenced by recall errors than other retrospective methods (e.g., asking participants to report PA perpetration in the past year; Straus et al., 1996), participants’ recall of events, temporal order, and alcohol consumption may not be entirely accurate, particularly after heavy drinking occurred. This study demonstrated that cyber PA can occur as many as 30 times per day, which poses difficulties in assessing the temporal order of alcohol use and cyber PA perpetration. Causal influences of alcohol use on cyber PA perpetration cannot be inferred. Finally, the context of drinking episodes (e.g., affective experiences, location) was not assessed which could elucidate alcohol myopia processes hypothesized to facilitate PA (Steele & Josephs, 1990).

Directions for Future Research

Advanced EMA methods that survey participants several times per day have recently been implemented in PA research (e.g., Testa et al., 2020). Such methods may be applied to cyber PA research elucidate the temporal ordering of PA types (e.g., whether cyber PA precedes in-person psychological, physical, and sexual PA), alcohol use, and other contextual correlates (e.g., affect, drinking setting). Future research should also examine the extent to which cyber PA perpetration increases the likelihood of subsequent drinking, and distal and contextual factors that may moderate this association. Recent investigations of the effects of alcohol intoxication on lab-based PA perpetration (e.g., Eckhardt et al., 2021; Watkins et al., 2015) could be extended to this line of research by developing methods to approximate cyber PA perpetration in a laboratory setting, which would allow investigators to assess the effects of acute intoxication on cyber PA, as well as the effects of cyber PA on alcohol craving. Investigators may also consider recruiting a sample of higher risk populations (e.g., those with a PA perpetration history) and couples to better approximate moderators of, and actor-partner effects within, the alcohol-cyber PA relationship. Notably, there were only 61 days in which cyber PA victimization was reported. This may reflect the difficulty in capturing cyber PA victimization in daily diary studies that do not sample couples. Some cyber PA acts can be performed discretely without a partner’s knowledge (e.g., snooping through emails or social media messages without their consent; sending private photos to others without their consent) which hinders an individual’s capacity to report cyber PA victimization. Collecting partner data may provide more accurate reports of cyber PA perpetration and victimization.

Conclusion

Although cyber PA perpetration was more likely to occur as the number of drinks consumed on a given day increased, alcohol use prior to cyber PA perpetration did not associate with subsequent perpetration. Specifically, odds of cyber PA perpetration after alcohol use were not greater on days in which alcohol use increased beyond one’s average daily alcohol use, and college students with higher average alcohol use were not more likely to perpetrate cyber PA after drinking than were those with lower average alcohol consumption. Post-hoc analyses support the importance of assessing temporal associations when examining alcohol use and cyber PA.

Public Health Significance Statements.

  1. Among college students, neither drinking more alcohol than one usually drinks on a given day, nor higher average alcohol use across all 60 days of the study, associated with odds of subsequent cyber PA perpetration.

  2. Cyber PA perpetration odds increased as the number of drinks consumed increased, regardless of whether drinking occurred before or after cyber PA, suggesting that cyber PA perpetration may confer risk for subsequent alcohol use.

Acknowledgments

This work was supported by a Visionary Grant from the American Psychological Foundation (APF) and by grant F31AA026489 from the National Institute on Alcohol Abuse and Alcoholism (NIAAA) awarded to the first author. The content is solely the responsibility of the authors and does not necessarily represent the official views of the APF, NIAAA, or the National Institutes of Health.

Footnotes

1

Results were not substantially different with cyber PA victimization removed from analyses.

References

  1. Anderson DK, & Saunders DG (2003). Leaving an abusive partner: An empirical review of predictors, the process of leaving, and psychological well-being. Trauma, Violence, and Abuse, 4(2), 163–191. [DOI] [PubMed] [Google Scholar]
  2. Bell KM, & Naugle AE (2007). Effects of social desirability on students’ self-reporting of partner abuse perpetration and victimization. Violence and Victims, 22(2), 243–256. [DOI] [PubMed] [Google Scholar]
  3. Bennett DC, Guran EL, Ramos MC, & Margolin G (2011). College students’ electronic victimization in friendships and dating relationships: Anticipated distress and associations with risky behaviors. Violence and Victims, 26(4), 410–429. [DOI] [PubMed] [Google Scholar]
  4. Borrajo E, Gámez-Guadix M, Pereda N, & Calvete E (2015). The development and validation of the cyber dating abuse questionnaire among young couples. Computers in Human Behavior, 48, 358–365. [Google Scholar]
  5. Brem MJ, Florimbio AR, Grigorian H, Elmquist J, Wolford-Clevenger C, & Stuart GL (2018). College-based dating violence prevention strategies In Temple JR & Wolfe DA (Eds.), Adolescent Dating Violence: Theory, Research, & Prevention. Cambridge, MA: Academic Press. [Google Scholar]
  6. Brem MJ, Romero G, Garner AR, Grigorian H, & Stuart GL (2019a). Alcohol problems, romantic jealousy, and cyber dating abuse perpetration among men and women: Towards a conceptual model. Journal of Interpersonal Violence. Advance online publication. doi: 10.1177/0886260519873333. [DOI] [PubMed] [Google Scholar]
  7. Brem MJ, Spiller LC, & Vandehey MA (2015). Online mate-retention tactics on Facebook are associated with relationship aggression. Journal of Interpersonal Violence, 30(16), 2831–2850. [DOI] [PubMed] [Google Scholar]
  8. Brem MJ, Stuart GL, Cornelius TL, & Shorey RC (2019b). A longitudinal examination of alcohol problems and cyber, psychological, and physical dating abuse: The moderating role of emotion regulation. Journal of Interpersonal Violence. Advance online publication. doi: 10.1177/0886260519876029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Burke SC, Wallen M, Vail-Smith K, & Knox D (2011). Using technology to control intimate partners: An exploratory study of college undergraduates. Computers in Human Behavior, 27, 1162–1167. [Google Scholar]
  10. Crane CA, & Eckhardt C (2013). Negative affect, alcohol consumption, and female-to-male intimate partner violence: A daily diary investigation. Partner Abuse, 4(3), 332–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Crane CA, & Testa M (2014). Daily associations among anger experience and intimate partner aggression within aggressive and nonaggressive community couples. Emotion, 14(5), 985–994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Crane CA, Umehira N, Berbary C, & Easton CJ (2018). Problematic alcohol use as a risk factor for cyber aggression within romantic relationships. The American Journal on Addictions, 27, 400–406. [DOI] [PubMed] [Google Scholar]
  13. Derrick JL, & Testa M (2017). Temporal effects of perpetrating or receiving intimate partner aggression on alcohol consumption: A daily diary study of community couples. Journal of Studies on Alcohol and Drugs, 78(2), 213–221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Duval A, Lanning BA, & Patterson MS (2020). A systematic review of dating violence risk factors among undergraduate college students. Trauma, Violence, and Abuse, 21(3), 567–585. doi: 10.1177/1524838018782207. [DOI] [PubMed] [Google Scholar]
  15. Eckhardt CI, Parrott DJ, Swartout KM, Leone RM, Purvis DM, Massa AA, & Sprunger JG (2021). Cognitive and affective mediators of alcohol-facilitated intimate-partner aggression. Clinical Psychological Science, 1–18. doi: 10.1177/2167702620966293 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Elmquist J, Hamel J, Febres J, Zapor H, Wolford-Clevenger C, Brem M, Shorey RC, & Stuart GL (2016a). Motivations for psychological aggression among dating college students. Partner Abuse, 7(2), 157–168. [Google Scholar]
  17. Elmquist J, Wolford-Clevenger C, Zapor H, Febres J, Shorey RC, Hamel J, & Stuart GL (2016b). A gender comparison of motivations for physical dating violence among college students. Journal of Interpersonal Violence, 31(1), 186–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Elphinston RA, & Noller P (2011). Time to face it! Facebook intrusion and the implications for romantic jealousy and relationship satisfaction. Cyberpsychology, Behavior, and Social Networking, 14, 631–635. [DOI] [PubMed] [Google Scholar]
  19. Flynn A, & Graham K (2010). “Why did it happen?” A review and conceptual framework for research on perpetrators’ and victims’ explanations for intimate partner violence. Aggression and Violent Behavior, 15, 239–251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Foran HM, & O’Leary KD (2008). Problem drinking, jealousy, and anger control: Variables predicting physical aggression against a partner. Journal of Family Violence, 23, 141–148. [Google Scholar]
  21. Gay and Lesbian Alliance Against Defamation. (2019). Accelerating acceptance 2019 Executive Summary: A Survey of American Acceptance and Attitudes Towards LGBTQ Americans. Retrieved from https://www.glaad.org/sites/default/files/Accelerating%20Acceptance%202019.pdf.
  22. Giancola PR, Josephs RA, Parrott DJ, & Duke AA (2010). Alcohol myopia revisited: Clarifying aggression and other acts of disinhibition through a distorted lens. Perspectives on Psychological Science, 5(3), 265–278. doi: 10.1177/1745691610369467 [DOI] [PubMed] [Google Scholar]
  23. Johnson WL, Giordano PC, manning WD, & Longmore MA (2015). The age-IPV curve: Changes in intimate partner violence perpetration during adolescence and young adulthood. Journal of Youth and Adolescence, 44(3), 708–726. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Johnson PO, & Neyman J (1936). Tests of certain linear hypotheses and their applications to some educational problems. Statistical Research Memoirs, 1, 57–93. [Google Scholar]
  25. Leisring PA, & Giumetti GW (2014). Sticks and stones may break my bones, but abusive text messages also hurt: Development and validation of the Cyber Psychological Abuse scale. Partner Abuse, 5, 323–341. doi: 10.1891/1946-6560.5.3.323 [DOI] [Google Scholar]
  26. Leonard KE (1993). Drinking patterns and intoxication in marital violence: Review, critique, and future directions for research. In U.S. Department of Health and Human Services (Ed.), Research monograph 24: Alcohol and interpersonal violence: Fostering multidisciplinary perspectives (NIH Publication No. 93–3496, pp. 253–280). Rockville, MD: National Institutes of Health. [Google Scholar]
  27. Leonard KE, & Quigley BM (2017). Thirty years of research show alcohol to be a cause of intimate partner violence: Future research needs to identify who to treat and how to treat them. Drug and Alcohol Review, 36, 7–9. doi: 10.1111/dar.12434. [DOI] [PubMed] [Google Scholar]
  28. Moore TM, Elkins SR, McNulty JK, Kivisto AJ, & Handsel VA (2011). Alcohol use and intimate partner violence perpetration among college students: Assessing the temporal association using electronic diary technology. Psychology of Violence, 1, 315–328. doi: 10.1037/a0025077. [DOI] [Google Scholar]
  29. Muscanell NL, Guadagno RE, Rice L, & Murphy S (2013). Don’t it make my brown eyes green? An analysis of Facebook use and romantic jealousy. Cyberpsychology, Behavior, and Social Networking, 16(4), 237–242. [DOI] [PubMed] [Google Scholar]
  30. National Institute on Alcohol Abuse and Alcoholism. (2010). Rethinking Drinking: Alcohol and Your Health. NIH Pub. No. 13-3770. Rockville, MD: National Institutes of Health. [Google Scholar]
  31. National Institute on Alcohol Abuse and Alcoholism (2015). Drinking Levels Defined. Retrieved from http://www.niaaa.nih.gov/alcohol-health/overview-alcohol-consumption/moderate-binge-drinking.
  32. Nemeth JM, Bonomi AE, Lee MA, & Ludwin JM (2012). Sexual infidelity as trigger for intimate partner violence. Journal of Women’s Health, 21(9), 942–949. doi: 10.1089/jwh.2011.3328. [DOI] [PubMed] [Google Scholar]
  33. Newport F (2018). In U.S., Estimate of LGBT Population Rises to 4.5%. New York: Gallup Press. Retrieved from https://news.gallup.com/poll/234863/estimate-lgbt-population-rises.aspx. [Google Scholar]
  34. Pfeiffer SM, & Wong PT (1989). Multidimensional jealousy. Journal of Social and Personal Relationships, 6, 181–196. [Google Scholar]
  35. Preacher KJ, Curran PJ, & Bauer DJ (2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics, 31(4), 437–448. [Google Scholar]
  36. Rodriguez LM, DiBello AM, & Neighbors C (2015). Positive and negative jealousy in the association between problem drinking and IPV perpetration. Journal of Family Violence, 30, 987–997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Rosenbaum A, Geffner R, & Benjamin M (1997). A biopsychosocial model for understanding relationship aggression. Journal of Aggression, Maltreatment, and Trauma, 1, 57–79. [Google Scholar]
  38. Shiffman S, Stone AA, & Huffard MR (2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4, 1–32. [DOI] [PubMed] [Google Scholar]
  39. Shorey RC, Stuart GL, McNulty JK, & Moore TM (2014a). Acute alcohol use temporally increases the odds of male perpetrated dating violence: A 90-day diary analysis. Addictive Behaviors, 39(1), 365–368. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Shorey RC, Stuart GL, Moore TM, & McNulty JK (2014b). The temporal relationship between alcohol, marijuana, angry affect, and dating violence perpetration: A daily diary study with female college students. Psychology of Addictive Behaviors, 28(2), 516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Singh V, Lee S, Epstein-Ngo Q, Carter P, Cunningham R, Walsh T, Tolman R (2015). Men who perpetrate physical and technology-delivered intimate partner violence: Correlates with substance use and beliefs about children. Injury Prevention, 21, Supplement 2: A5. 1–A5. [Google Scholar]
  42. Stappenbeck CA, & Fromme K (2014). The effects of alcohol, emotion regulation, and emotional arousal on the dating aggression intentions of men and women. Psychology of Addictive Behaviors, 28(1), 10–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Stappenbeck CA, Gulati NK, & Fromme K (2016). Daily association between alcohol consumption and dating violence perpetration among men and women: Effects of self-regulation. Journal of Studies on Alcohol and Drugs, 150–159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Steele CM, & Josephs RA (1990). Alcohol myopia: Its prized and dangerous effects. American Psychologist, 45(8), 921. [DOI] [PubMed] [Google Scholar]
  45. Straus MA, Hamby SL, Boney-McCoy S, & Sugarman DB (1996). The Revised Conflict Tactics Scales (CTS-2): Development and preliminary psychometric data. Journal of Family Issues, 17, 283–316. doi: 10.1177/019251396017003001 [DOI] [Google Scholar]
  46. Straus MA, Hamby SL, & Warren WL (2003). The Conflict Tactic Scales handbook. Los Angeles, CA: Western Psychological Services. [Google Scholar]
  47. Taft CT, O’Farrell TJ, Torres SE, Panuzio J, Monson CM, Murphy M, & Murphy CM (2006). Examining the correlates of psychological aggression among a community sample of couples. Journal of Family Psychology, 20, 581–588. doi: 10.1037/0893-3200.20.4.581. [DOI] [PubMed] [Google Scholar]
  48. Testa M, Crane CA, Quigley BM, Levitt A, & Leonard KE (2014). Effects of administered alcohol on intimate partner interactions in a conflict resolution paradigm. Journal of Studies on Alcohol and Drugs, 75, 249–258. 10.15288/jsad.2014.75.249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Testa M, & Derrick JL (2014). A daily process examination of the temporal association between alcohol use and verbal and physical aggression in community couples. Psychology of Addictive Behaviors, 28(1), 127–138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Testa M, Wang W, Derrick JL, Crane C, Leonard KE, Collins RL, … & Muraven M. (2020). Does state self-control depletion predict relationship functioning and partner aggression? An ecological momentary assessment study of community couples. Aggressive Behavior, 46(6), 547–558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Tolman RM (1989). The development of a measure of psychological maltreatment of women by their male partners. Violence and Victims, 4, 159–177. [PubMed] [Google Scholar]
  52. Van Ouytsel J, Ponnet K, Malrave M, & Temple JR (2016). Adolescent cyber dating abuse victimization and its associations with substance use, and sexual behaviors. Public Health, 135, 147–151. [DOI] [PubMed] [Google Scholar]
  53. Watkins LE, DiLillo D, & Maldonado RC (2015). The interactive effects of emotion regulation and alcohol intoxication on lab-based intimate partner aggression. Psychology of Addictive Behaviors, 29(3), 653–663. doi: 10.1037/adb0000074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Watson D, Clark LA, & Tellegen A (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. doi: 10.1037/0022-3514.54.6.1063 [DOI] [PubMed] [Google Scholar]
  55. Whitaker MP (2013). Motivational attributions about intimate partner violence among male and female perpetrators. Journal of Interpersonal Violence, 29, 517–535. doi: 10.1177/0886260513505211. [DOI] [PubMed] [Google Scholar]
  56. White GL (1981). A model of romantic jealousy. Motivation and Emotion, 5, 295–309. [Google Scholar]
  57. Wolford-Clevenger C, Elmquist J, Brem MJ, Zapor H, & Stuart GL (2016a). Dating violence victimization, interpersonal needs, and suicidal ideation among college students. Crisis, 37(1). 51–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wolford-Clevenger C, Zapor H, Brasfield H, Febres J, Elmquist J, Brem M, Shorey RC, & Stuart GL (2016b). An examination of the Partner Cyber Abuse Questionnaire in a college student sample. Psychology of Violence, 6(1), 156–162. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES