The Relationship Behavior Survey: A Comprehensive Measure of Psychological Intimate Partner Violence for Adolescents

Abstract

Commonly used scales of psychological intimate partner violence (ψIPV) for adolescents may not include sufficient items to measure adequately different forms of ψ aggressive behaviors. They may also characterize as harmful ψ aggressive behaviors occurring in non-conflictual or joking contexts. The current study examined a new scale, the Relationship Behavior Survey (RBS), which was designed to measure three different forms of ψIPV (denigrating, controlling, and intrusive behaviors) and the appraisals of the perpetrator’s intent. Factor structure was examined, and incremental validity was tested by examining the relation of the RBS to general aggressive tendencies and physical IPV after accounting for the emotional abuse and threatening behavior subscales of the Conflict in Adolescent Dating Relationships Inventory (CADRI). Criterion validity and gender differences were also examined. Data were collected using an online survey from a national sample of 1,100 13 to 17-year-olds in the United States (51% identified as female, 80% as White) in a dating relationship. Confirmatory factor analysis supported one ψIPV construct for males and females. The RBS demonstrated incremental validity; criterion validity was supported for controlling intent for males and females and mistrust intent for females. Joking intent was the most reported intent for males and correlated unexpectedly with physical IPV. The RBS captures a larger range of ψIPV behaviors than the CADRI. The addition of intent appraisals provides important information to differentiate forms of ψIPV, particularly in relation to acts motivated by coercion and mistrust.

Keywords

dating violence youth violence stalking sexual harassment psychological aggression measurement adolescent relationships

Psychological intimate partner violence (ψIPV)—including insults, threats, monitoring, and controlling behaviors—is a serious concern in adolescent populations (e.g., Taylor & Mumford, 2016). Rates of ψIPV have been found to be over 60% in nationally representative samples of 12 to 18-year-olds in the United States (Borrajo et al., 2015; Taylor & Mumford, 2016), and rates as high as 75% to 90% have also been reported for both in-person and electronic ψIPV (Borrajo et al., 2015; Carney & Barner, 2012). Presently, ψIPV is commonly assessed with self-report measures of the frequency of specific behaviors with a partner (e.g., SafeDates, Foshee et al., 1998; Conflict in Adolescent Dating Relationships Inventory [CADRI], Wolfe et al., 2001). However, the measures that are commonly used with adolescents do not appear to cover the multidimensional nature and variety of ψIPV behaviors that have been identified in research with adults (Follingstad, 2011; Follingstad et al., 2015; Murphy & Hoover, 1999). Additionally, the measures that are commonly used with adolescents do not attempt to assess the context in which the behaviors occurred. Certain contextual variables, such as the intent or motive of behaviors, may be important to understand and account for behavior that may be joking or accidental (i.e., the perpetrator did not realize their action was unwanted).

The current study was designed to test a self-report survey of ψIPV perpetration in adolescence, the Relationship Behavior Survey (RBS). The RBS assesses (1) a larger number and variety of ψIPV behaviors than other commonly used scales of ψIPV in adolescence; and (2) appraisals of intent for engaging in different forms of ψIPV behaviors. The CADRI was used to test incremental validity as it is considered to be a “gold standard” scale for measuring IPV in adolescent relationships (Exner-Cortens et al., 2016).

ΨIPV as a Multidimensional and Multimodal Construct

Scholars (e.g., Follingstad et al., 2015; Murphy & Hoover, 1999) have developed multidimensional scales of ψIPV in samples of college-age adults, differentiating among denigration, jealous, controlling, and intimidation behaviors. Factor analyses have supported different subtypes of ψIPV, and tests of criterion validity have shown both overlapping and distinct correlates. For example, using a college student sample, Murphy and Hoover (1999) found that all forms of perpetrating ψIPV were associated with vindictive, domineering, and intrusive personality, whereas denigration and controlling behaviors showed the strongest association with physical IPV (Murphy & Hoover, 1999). Theorists suggest that controlling behavior (e.g., restricting partner’s freedom) is a particularly serious form of ψIPV (Johnson, 2011), and its perpetration is relatively stable over time (Miller et al., 2013). Furthermore, research has shown that adolescent relationships characterized by a high degree of controlling behavior involve more frequent acts of physical IPV and fear (Messinger et al., 2014; Reidy et al., 2016). Yet, measures designed for use with adolescents, such as SafeDates (Foshee et al., 1998) and the CADRI (Wolfe et al., 2001) include items reflecting distinct types of ψIPV (e.g., controlling behavior, degradation, jealousy), but do not identify them as different types of dimensions of ψIPV. Moreover, other measures of ψIPV combine behaviors and appraisals of intent into some items in non-systematic ways (e.g., destroyed personal things to intimidate, tried to make personal choices to control partner) and include items that are not appropriate for adolescents (e.g., implied having an affair to make worried).

Although research on ψIPV via electronic communication has grown recently (Borrajo et al., 2015; Mehari & Farrell, 2018; Muñoz-Ferńandez & Sánchez-Jiménez, 2020), existing measures do not include both in-person behavior and electronic communication. The latter has become a mainstay in adolescent relationships; 95% use the Internet and 85% use social media (Lenhart et al., 2011). This increased access to each other makes adolescents more susceptible to ψIPV such as denigration and intrusive behaviors (Southworth et al., 2007). Denigration is the act of harming a partner’s reputation or self-worth through ridicule, and can occur in-person or online. Intrusive behavior involves an intentional invasion of privacy, such as by hacking into a partner’s social media account. Research has shown that 50% to 75% of adolescents have either initiated or been victims of ψIPV through electronic communication (Borrajo et al., 2015). One key difference between online and in-person denigration is that the online form of denigration is more likely to be observed by others when it occurs on social media. Invasion of privacy can also occur online or in-person, such as by going through a person’s mobile phone or personal belongings. It is important for ψIPV scales to include in-person and electronic forms as this will allow for a more comprehensive understanding of ψIPV (Mehari & Farrell, 2018; Muñoz-Ferńandez & Sánchez-Jiménez, 2020).

Context Matters

Lack of context information contributes to difficulties in interpreting when certain behaviors may be problematic. It may also inflate estimates of ψIPV by including developmentally typical adolescent dating behavior. Refining the measurement of an individual’s intent for engaging in the behavior is one aspect of context that may improve the study of risk factors for ψIPV and guide the development of new prevention targets. One type of intent that has been studied recently is whether acts such as pushing or shoving are intended as playful or joking, and if they are, how to account for these behaviors in studies of risk factors and correlates (Ackerman, 2016; Hamby, 2016; Sargent et al., 2020). This is particularly relevant since many adolescents report engaging in physical IPV in a playful or joking context (Fernández-Gonzalez et al., 2013; Foshee et al., 2007; Gonzalez-Mendez & Hernandez-Cabrera, 2009; Jouriles et al., 2009). However, this intention has not been examined concerning ψIPV.

It is also unclear if behaviors that are intended as playful or joking are distinct from those occurring with other intentions, such as to express mistrust or anger, to control one’s partner, to retaliate against one’s partner, or to feel a loss of control over one’s emotions. Mistrust may be linked to feelings of insecurity and fear of rejection by a romantic interest, which may be heightened during adolescence, and may specifically relate to controlling behaviors or invasions of a partner’s privacy (Cascardi & Avery-Leaf, 2015; Furman et al., 2007). Mistrust, jealousy, control, and retaliation are correlates of (Copp & Johnson, 2015; Johnson et al., 2015) and reasons for (Bair-Merritt et al., 2010; Hughes et al., 2016; Langhinrichsen-Rohling et al., 2012; Leisring, 2013) physical IPV, but they have not been examined regarding ψIPV. Assessment of intentions for ψIPV may be particularly important in adolescence when name-calling or other inappropriate behavior may be expressed in a joking or playful manner, or to gain closeness (e.g., Arriaga, 2002). Joking intentions may be developmentally relevant as adolescents learn ways to express affection, closeness, and intimacy. Lack of information about intentions also limits the investigation of developmentally typical interactions among adolescents in other ways. For instance, the act of checking in with a partner throughout the day, both by text or in-person, may be intended to show care and concern for one’s partner, and may similarly be experienced as such. Measurement of these context factors (i.e., joking, showing care) can help to distinguish whether acts like checking in are harmful or not.

Another important aspect of context that has not been measured is if ψIPV acts persist after one’s partner has asked them to stop. Such acts may be experienced as intrusive. To date, items have not been constructed to evaluate whether a behavior continued even after their partner asked them to stop. This qualifier may be especially helpful for acts that occur accidentally or without intentions to harm, especially for behaviors that may have more innocuous or unclear motives, such as playful name-calling or repeatedly checking in by text messages or social media. Adding the qualifier that behavior continued even after knowing their partner wanted it to stop provides a clear and explicit way to assess if the behavior was unwanted. This distinction may help to differentiate seemingly aggressive acts intended to gain closeness to a partner or as a means of emotional expression with a positive impact, from those with harmful intent. Including information about intentions for specific acts of ψIPV and adding a qualifier to some acts (e.g., repeatedly checking in with a partner via text), specifying if the act was explicitly unwanted, may increase the precision of ψIPV measurement.

Relationship of Gender With ψIPV

Male and female adolescents have reported similar levels of ψIPV in their relationships (Orpinas et al., 2012; Taylor & Mumford, 2016), although the form of ψIPV varies by gender (Cascardi & Avery-Leaf, 2016). For example, females have reported more insults, sulking, spiteful behavior, and rumor spreading than males (Cascardi & Avery-Leaf, 2016; Taylor & Mumford, 2016), whereas reports of controlling behavior have been similar for males and females (e.g., Cascardi & Avery-Leaf, 2016). Results on gender differences in intentions of IPV have been mixed (Brooks-Russell et al., 2013; Flynn & Graham, 2010). A measure that assesses the intentions of ψIPV systematically may allow for a closer examination of gender differences.

The Current Study

The current study was designed to test the dimensional structure, incremental validity, and criterion validity of a novel self-report scale (RBS) of in-person and electronic ψIPV behaviors. The RBS assesses a variety of ψIPV behaviors, which were hypothesized to measure multiple dimensions of ψIPV. The RBS was also designed to measure the perceived intent driving these acts. The first study aim was to test the factor structure and gender invariance of the RBS using confirmatory factor analysis. We hypothesized a three-factor solution based on the existing research (e.g., Borrajo et al., 2015; Follingstad et al., 2015; Murphy & Hoover, 1999): denigration, controlling, and intrusive behaviors. A distinction between controlling and intrusive behaviors is that the former may be intended to restrict a partner’s activities or autonomy, whereas the latter refers to invading a partner’s privacy and surveillance. We expected the three factors to be correlated and tested three models to account for the correlation among factors: a correlated factor model, a higher-order factor model, and a bifactor model.

The second aim was to test the incremental validity of the RBS, examining if the RBS related to general aggressive tendencies and physical IPV, after accounting for the emotional abuse and threatening behavior subscales of the CADRI. We hypothesized that the RBS would improve the prediction of criterion variables. The third aim was to assess the intent driving the ψIPV acts, including to joke, show care, express mistrust, display anger, and control. As intentions may be more readily accessible to those who engage in these acts, we focused on perpetrators’ intent. Due to inconclusive research on intentions associated with ψIPV, we examined intentions by gender and the variety of ψIPV acts without specific hypotheses. The fourth and final aim was to test the criterion validity of intent appraisals among those reporting any ψIPV with general aggressive tendencies and physical IPV. We chose general aggressive tendencies assessed by the Aggression Questionnaire (Buss & Perry, 1992) and physical IPV assessed by the CADRI, given the importance of testing the association of ψIPV with other aggressive acts. We hypothesized that joking and show care intentions would be negatively associated with aggressive tendencies and physical IPV whereas mistrust, anger, and control would be positively associated with these criterion variables.

Method

Procedures

Data were collected from a United States (U.S.) online panel sample of youth aged 13 to 17 years through Qualtrics’ panel sample (Qualtrics.com). Qualtrics’ staff sent an invitational email to all parents in their panel sample who had children in the target age range (N = 4,982). The vendor managed data collection and incentives, which were nominal tokens (value $1 USD) redeemable for products through Qualtrics. Of the email invites, 88% (n = 4,384) of parents consented to their children’s participation in a quiet, private location after viewing sample items. Parents were not asked if their children had dated; youth provided this information. The final sample consisted of 1,100 youth (25% of consented sample) who reported having been in a current or recent dating relationship within the prior 6 months and gave assent. This proportion of youth in a current or recent dating relationship is consistent with the proportion of 13 to 17-year-olds in other research studies who reported being in a current or recent dating relationship (20%; Lenhart et al., 2015). The Institutional Review Board approved the protocol. The presentation order of the RBS and CADRI was counterbalanced; the RBS was presented first to one-half of participants and the CADRI was presented first to one-half of participants.

Participants

Youth in the panel sample identified as White (80.2%, n = 882), Black (10.9%, n = 120), Multi-ethnic/racial (3.6%, n = 40), or Asian/Pacific Islander (2.2%, n = 24); 3% (n = 34) did not answer. In addition, 8.1% (n = 89) identified as Latinx regardless of racial identity. The sample represented each region of the United States: Northeast (21.9%, n = 241), Southeast (29.7%, n = 327), Midwest (25.2%, n = 277), Southwest (11.1%, n = 122), and West/Pacific Northwest (11.8%, n = 130); n = 3 did not answer. Most of the sample was 16 to 17 years (57.4%, n = 631); 19.6% (n = 216) was 15 years, and 23% (n = 253) was 13 to 14 years. Fifty-one percent of the sample identified as female (n = 567), 46.6% (n = 513) as male, and 1.8% (n = 20) as non-binary or gender non-conforming. Most identified as heterosexual (87.9%, n = 967), 7.5% (n = 83) as bisexual, and 3.2% (n = 35) as homosexual; 1.4% (n = 15) did not answer. Their romantic relationships tended to be brief (< 3 months) (33.8%, n = 372); yet 20.1% (n = 221) reported relationships lasting 1 year or more. Due to the small number of those identifying as non-binary or gender non-conforming, only those identifying as male (n = 513) or female (n = 567) were included in the current study.

Parents of youth in the panel were more educated and included a higher percentage of very high-income earners (personal communication with Qualtrics, November 9, 2018) than the general population: 96% of the parent sample graduated from high school or had some higher education compared to 88% in the 2015 U.S. census, and 14% of the parent sample earned more than $250K compared to 6.1% earning more than $200K in the 2015 U.S. census.

Measures

The 40-item RBS was designed to assess three types of ψIPV: denigration, controlling, and intrusive behaviors. The RBS was developed using a multi-step process (Cascardi et al., 2018). First, existing scales of in-person and electronic ψIPV were reviewed to generate a sample of potential items (see Supplemental Table 1 for a list of scales). Second, focus groups with adolescents in grades 7 to 8 (n = 38) and 9 to 12 (n = 33) were conducted to inform the initial item pool and intentions. This entailed asking open-ended questions about the reasons for engaging in specific acts of aggression. Third, we generated a list of 125 items which were rated by adolescents in grades 7 to 9 (n = 53) and 9 to 12 (n = 56) on a 4-point scale of acceptability. This rating scale was chosen to ensure the RBS represented a range of behaviors that varied in intensity. Results from this step reduced the item pool to 80 items. Fourth, the 80 items were reviewed by a panel of scholars and practitioners with expertise in adolescent relationships and IPV (n = 15). They rated the items based on ease of comprehensiveness, relevance, and likelihood of improving measurement. In the final step before testing the RBS, we conducted cognitive interviews to test item interpretation and comprehension with adolescents in grades 7 to 8 (n = 22) and 9 to 12 (n = 17). We probed ambiguous or difficult concepts (e.g., check-in via text or phone) to confirm item wording was understood as intended and modified items to increase clarity and interpretation when they were not.

A description of the three subtypes of RBS behavior follows. There are 7-items for denigration behavior, for example, “I called my partner mean names even after they said to stop,” “I made fun of my partner even after they said to stop;” 7-items for controlling behavior, for example, “I told partner whom to spend time with,” “I demanded to know where partner was;” and 6-items measuring intrusive behavior, for example, “I checked my partner’s phone or social media even after they said to stop,” “I monitored my partner’s posts on social media.” Respondents indicated how often they engaged in (self-report) or experienced (report about partner) each of the 20 behaviors with their current or recent partner in the past 6 months on a 5-point scale: “not at all,” “once or twice,” “a few times a month,” “a few times a week,” or “daily/multiple times a day.” Variety scores were computed, which is a count of the number of items endorsed more than “not at all.” Since this study focused on intent, only perpetration items were used.

Participants who endorsed one or more items on the denigration ψIPV subtype completed a checklist of 9-intentions, marking as many as apply. Intention checklist options are to joke, hurt, retaliate, control, show care, put down, express anger, mistrust, and other. The survey was personalized; participants were reminded of the behaviors they endorsed before answering the intent checklist. This procedure was repeated for controlling and intrusive ψIPV subtypes with the same intention checklist options. Intentions were aggregated across subtypes; each intention was coded “yes, reported at least once” or “no, not reported.”

The emotional abuse, threatening behavior (threatening), and physical IPV subscales from the Conflict in Adolescent Dating Relationships Inventory (CADRI; Wolfe et al., 2001) were used to examine incremental and criterion validity. There are 10 items on the emotional abuse subscale, including insults, jealous accusations, and controlling actions (e.g., checking a partner’s whereabouts), 4 items on the threatening subscale (e.g., threaten to hurt partner), and 4 items on the physical IPV subscale (e.g., push, shove, or shook partner). Participants indicated the frequency of each behavior in the previous 6 months on a four-point Likert scale from “never happened” to “often happened.” Intentions are not assessed on the CADRI. The CADRI has good psychometric properties among adolescents (Wolfe et al., 2001) and emerging adults (Caiozzo et al., 2016; Cascardi et al., 2019). Internal consistency in the current sample was α = .92 for emotional abuse, α = .83 for threatening, and α = .92 for physical IPV. Variety scores were computed, that is, a count of items endorsed more than “never happened.” For physical IPV, square root transformation was used to normalize the right skew distribution (zero-inflated).

General aggressive tendencies were measured with the Aggression Questionnaire (AQ) (Buss & Perry, 1992). The AQ consists of 29-items and measures four types of aggressive tendencies: verbal aggression (5 items; e.g., “I find myself disagreeing with others”), physical aggression (9 items; e.g., “I will resort to violence to protect my rights”), anger arousal (7 items; e.g., “I have trouble controlling my temper”), and hostility (8 items; e.g., “I know that ‘friends’ talk behind my back”). Items are rated on a 7-point Likert scale from “extremely uncharacteristic of me” to “extremely characteristic of me.” The AQ has demonstrated adequate levels of internal consistency and construct validity in community samples (Bernstein & Gesn, 1997; Buss & Perry, 1992). In the current study, internal consistency for the AQ was α = .97.

Data Analyses

We tested the measurement model of the RBS in MPlus7 (Muthén & Muthén, 2017). Due to violations of normality, we used weighted least squares confirmatory factor analysis for categorical data to test the fit of the measurement model of the RBS. Four models were compared: one factor model, a 3-factor correlated model (denigration, controlling, and intrusive), a second order factor model (all lower order factors loading on a common factor), and a bifactor model (a nested factor model). In a bifactor model, each item loads on a common factor and residual variance loads on orthogonal specific factors (Brunner et al., 2012; Dueber, 2020). One advantage of the bifactor model is it tests for one overarching construct of ψIPV while simultaneously measuring distinct, uncorrelated ψIPV subscales. Each model was assessed in a multigroup (i.e., males and females) analysis using the chi-square test of model fit (Bentler & Bonett, 1980), the Comparative Fit Index ([CFI] Bentler, 1990; Hu & Bentler, 1999), and Tucker-Lewis Index ([TLI]; Tucker & Lewis, 1973). These indices estimate the proportionate improvement in fit by comparing the hypothesized model to the baseline model, typically the null model (Hu & Bentler, 1999). The fit estimates are scaled to an approximate range of 0 and 1. By convention, values of .90 or higher are considered acceptable values of fit and values above .95 indicate good fit (Bentler & Bonett, 1980). Root Mean Square Error Adjustment (RMSEA) values less than .08 indicate a reasonable error of approximation (Schreiber et al., 2006). Gender invariance was tested for configural, metric, and scalar models.

Hierarchical multiple regression analyses were computed to test incremental validity of the RBS, after accounting for the CADRI emotional abuse and threatening subscales, using aggressive tendencies and physical IPV as criterion variables. CADRI emotional abuse and threatening subscales were entered in the first step, ψIPV RBS in the second step, and the interaction of ψIPV RBS by sex in the third step. Intentions for ψIPV were calculated as an unduplicated frequency of endorsement. Intent appraisals were coded as “yes, intent reported at least once” or “no, intent not reported.” Z-tests for the difference between proportions and t-tests were used to examine gender differences on the ψIPV RBS scores and intentions. Finally, to examine criterion validity for ψIPV intent appraisals, point biserial bivariate correlations were computed, separately for those identifying as male or female. Only those who reported ψIPV on the RBS provided intent appraisals, so the sample was limited to those reporting one or more acts of ψIPV (n = 741, 68.6%).

Results

Sixty-three percent of males and 73.7% of females endorsed at least one item on the RBS. The mean number of items endorsed was 6.59 (6.46) for males and 5.78 (5.25) for females. Although the rate of ψIPV was significantly higher for females (X², df = 1, 14.47, p < .001), there was no significant difference in the number of RBS items endorsed, t(739) 1.88, p = .06. Item endorsement on the RBS ranged from 10.6% to 42.7%. “Screamed at partner,” “made partner show text messages,” “monitored social media posts,” and “checked phone or social media even after being told to stop” were among the more commonly reported behaviors. Females were more likely to report “screamed at partner” (z = 2.53, p < .05), “demanded to know where partner was” (z = 1.81, p < .05), and “monitored social media posts” (z = 2.08, p < .05) than males; males were more likely to report “did not allow partner to wear what they wanted” (z = 3.95, p < .001).

The rate of item endorsement on all RBS items is shown in Table 2.

Test of RBS Factor Structure (Aim 1)

As shown in Table 1, the one-factor, three-factor correlated, and bifactor models each fit the data well. However, the factor loading among the three-factors was nearly perfect, λ = .938–1.00. For this reason, the higher-order factor model was not positive definite, indicating linear dependence among the factors, and therefore could not be tested. The results of the bifactor model indicated one common factor and no separate subscales (see Table 2). Factor loadings on the single, common factor were uniformly high (0.745 to 0.980) and statistically significant, whereas loadings on the specific factors were low and not statistically significant (<0.50, ranging −0.006 to 0.293). In further support of a unidimensional bifactor model, Relative Omega, a statistic measuring the % of variance that is independent of the common factor, was less than 1% (Relative Omega = .009) (Dueber, 2017). The hypothesized three-factor solution was not supported; a bifactor model with one common factor and no distinct subscales as well as a one-factor model were the best fit model. In tests of gender invariance, the three-factor correlated model was not positive definite and thus gender invariance could not be calculated for this model. Tests of configural, metric, and scalar gender invariance models supported a one-factor model as the best fit for males and females. For remaining analyses, a single factor model of ψIPV was used.

Table 1.

Confirmatory Factor Analysis Model Fit: RBS ψIPV.

	Chi-Square	CLI	TLI	RMSEA
One factor model		0.994	0.994	0.049 (0.044–0.053)
Total	df = 418, 949.41
Male	df = 418, 368.72
Female	df = 418, 580.69
3-factor correlated model		0.994	0.995	0.047 (0.043–0.051)
Total	df = 408, 888.16
Male	df = 408, 360.83
Female	df = 408, 527.32
Second order model	not positive definite
Bifactor model		0.994	0.994	0.051 (0.047–0.055)
Total	df = 396, 949.25
Male	df = 396, 368.04
Female	df = 396, 581.21
Test of Gender Invariance+
Configural	df = 304, 866.651	0.993	0.992	0.060 (0.055–0.064)
Metric	df = 322, 905.394	0.993	0.992	0.058 (0.053–0.062)
Scalar	df = 378, 883.785	0.994	0.994	0.050 (0.046–0.054)
Metric versus Configural	df = 18, 20.524, p = .304
Scalar versus Configural	df = 74, 82.582, p = .231
Scalar versus Metric	df = 56, 62.927, p = .245

Notes. + results reported when item “did not allow partner to wear what they wanted” removed. When this item was included, configural and scalar invariance were supported but metric invariance (i.e., factor loading equivalence between males and females) was not.

Table 2.

Bifactor Model Results: Confirmatory Factor Analysis of RBS.

	Male (n = 513)			Female (n = 567)
	%	λ common factor	λ specific factor	%	λ common factor	λ specific factor
Denigrating
Screamed at partner	34.1	.882	.293	41.6^a	.913	.254
Called partner mean names after said stop	18.5	.940	.092	18.9	.879	.086
Threatened to tell secrets	14.4	.967	.137	11.5	.899	.079
Made fun of partner even after said stop	23.2	.898	.141	21.5	.835	.131
Shared private pictures of partner	15.2	.951	.062	12.7	.838	.055
Made partner send private photos	15.6	.940	.092	10.8	.879	.086
Threatened to hurt	13.6	.967	.085	11.6	.899	.079
Controlling
Not allow partner to do what they wanted	19.7	.942	.073	23.1	.803	.063
Said who partner can talk to	20.7	.950	.105	21	.886	.098
Not allow partner to wear what they wanted	19.1	.962	−.006	10.6^a	.884	−.006
Told partner whom to spend time with	26.3	.882	.111	28.6	.777	.098
Made partner show text messages	28.8	.886	.144	32.8	.839	.137
Made partner give passwords	18.3	.951	.079	18	.858	.071
Demanded to know where partner was	24.4	.927	.125	29.3^a	.864	.098
Intrusive
Checked in, texted, called after told stop	20.7	.933	.062	19.9	.823	.054
Hacked into social media account	13.5	.980	.047	11.3	.917	.044
Listened to phone calls, read emails, or went through belongings	20.1	.946	.108	24	.870	.099
Followed or spied on	22.6	.949	.081	23.8	.866	.074
Monitored social media posts	36.5	.859	.145	42.7^a	.745	.126
Checked phone or social media after told stop	27.1	.922	.086	24.9	.830	.077

Note. Superscripts indicate significant gender differences in item endorsement.

Test of Incremental Validity (Aim 2)

Endorsement on the CADRI emotional abuse or threatening subscales (76.2% of males and 82.3% of females) was significantly higher than endorsement rates of abuse on the RBS (63.0% of males and 73.7% of females), z = 4.52, p < .001 and z = 3.50, p < .001, respectively. We hypothesized that ψIPV scores on the RBS would have incremental validity, after accounting for the emotional abuse and threatening CADRI subscale scores. As shown in Table 3, after CADRI subscale scores were entered into the regression equation, ψIPV scores on the RBS improved the prediction of aggressive tendencies and physical IPV, ΔF(1,1077) = 28.44, p < .001 and ΔF(1,1080) = 101.63, p < .001, respectively. In addition, there was a significant ψIPV by gender interaction such that males who engaged in ψIPV demonstrated significantly more aggressive tendencies than females, ΔF(1,1076) = 6.88, p < .001.

Table 3.

Test of RBS ψIPV Incremental Validity in the Full Sample (N = 1,080).

	Aggressive Tendencies					Physical IPV (sq root)
	B	t	R ²	r	F	B	t	R ²	r	F
CADRI-Emotional AbuseCADRI-Threatening	0.3340.323	9.69* 9.39*	0.369		F (2,1078) = 316.34*	0.1170.716	4.53 * 27.69*	0.645		F (2,1081) = 983.92*
CADRI-Emotional AbuseCADRI-ThreateningψIPV RBS	0.2630.2160.210	7.20* 5.45* 5.33*	0.384		ΔF (1,1077) = 28.44*	0.0210.5650.288	.808, ns19.59* 10.01*	0.675		ΔF (1,1080) = 101.63*
CADRI-Emotional AbuseCADRI-ThreateningψIPV RBSψIPV RBS by Sex	0.2740.2050.255−0.077	7.49* 5.15* 5.95* −2.63**	0.388	.565* .562* .559* .284*	ΔF (1,1076) = 6.88*	0.0170.5690.2690.033	.640, ns19.67* 8.64* 1.55, ns	0.675	0.629* 0.799* 0.762* 0.422*	ΔF (1,1079) = 2.40

p < .001. **p < .01.

Intent Appraisals (Aim 3)

The most reported intentions for males were to joke (44.6%), show care (34.7%), control (26%), anger (25.1%), and mistrust (24.5%), and for females they were to show care (43.8%), joke (37.6%), anger (37.6%), mistrust (29.2%), and control (22%). Males were significantly more likely than females to report joking (44.6% vs. 37.6%, z = 2.33, p < .01) and put partner down (8.7% vs. 4.8%, z = 2.57, p < .01) as their intent appraisals, whereas females were significantly more likely to report show care (43.8% vs. 34.7%, z = 3.06, p < .01) and anger (37.6% vs. 25.1%, z = 4.14, p < .001) than males. Descriptive analyses showed that for males and females, endorsement of control, hurt, and put partner down intent appraisals were associated with a larger number of ψIPV acts than other intent appraisals (see Table 4).

Table 4.

Intentions for ψIPV by Gender^a

	Male % (n)	Mean # items (SD)	Female % (n)	Mean # items (SD)	z-test for proportions
Joke	44.6% (144)	8.86 (7.54)	37.6% (157)**	6.59 (5.90)	2.33**
Hurt	6.5% (21)	18.62 (3.22)	4.1% (17)	13.41 (7.08)⁺	1.77
Control	26% (84)	13.74 (6.61)⁺	22% (92)	11.72 (5.20)⁺	1.54
Retaliate	8.7% (28)	9.71 (7.15)	9.1% (38)	9.32 (5.53)	−0.23
Show care	34.7% (112)	7.76 (6.77)	43.8% (183)**	6.54 (5.48)	3.06**
Put down	8.7% (28)	15.46 (5.94)⁺	4.8% (20)**	11.75 (5.60)⁺	2.57**
Anger	25.1% (81)	5.28 (4.46)	37.6% (157)***	6.02 (4.54)	4.14***
Mistrust	24.5% (79)	6.45 (5.16)	29.2% (122)	7.71 (4.54)	1.74
Other	13.3% (43)	4.31 (4.93)	14.4% (60)	5.32 (4.53)	0.52

p < .01.

***

p < .001.

Intent associated with significantly more RBS items than intent without ⁺, p < .05.

Categories are not mutually exclusive.

Test of Criterion Validity for Intent Appraisals (Aim 4)

The test of criterion validity only included those who reported any ψIPV, since those who did not report ψIPV did not have data on intent. We hypothesized that to joke and show care intent would be negatively associated with aggressive tendencies and physical IPV whereas mistrust, anger, and control would be positively associated with these variables. However, as shown in Table 5, for males, to joke and to control were positively and significantly associated with aggressive tendencies (r = .128, p < .001 and r = .415, p < .001) and physical IPV (r = .238, p < .001 and r = .496, p<.001). For females, to control and mistrust were positively associated with aggressive tendencies (r = .296, p < .001 and r = .118, p < .05) and physical IPV (r = .438, p < .001 and r = .122, p < .05). In addition, to joke was significantly associated with physical IPV for females (r = .116, p < .05). Of note, to joke was also negatively correlated with mistrust and anger for males and females, as well as negatively correlated with control for females.

Table 5.

Test of Criterion Validity: Aggressive Tendencies and Physical IPV among Those Reporting Any ψIPV on RBS (n = 741).

	1	2	3	4	5	6	7	8
(1) Aggressive tendencies	1	0.567**	0.128*	0.415**	0.081	0.036	0.018	0.638**
(2) Physical IPV (Sq root)	0.491**	1	0.238**	0.496**	0.083	−0.124	−0.174	0.783**
(3) Joke	0.012	0.116*	1	−0.006	0.106	−0.148*	−0.232**	0.266**
(4) Control	0.296**	0.438**	−0.15**	1	−0.046	0.040	−0.034	0.605**
(5) Show care	−0.007	−0.006	0.054	0.021	1	−0.233**	−0.001	0.103
(6) Mistrust	0.118*	0.122*	−0.128**	0.218**	−0.28**	1	0.136*	−0.029
(7) Anger	0.003	−0.015	−0.255**	0.053	−0.084	0.176**	1	−0.127*
(8) RBS	0.440**	0.658**	0.089	0.565**	0.075	0.221**	0.025	1

Note. Males above the diagonal, females below the diagonal.

Report of ψIPV on the RBS is controlled; the sample only includes those who reported at least one act of ψIPV. Intent is coded as 1 = yes, reported at least once or 0 = no. Positive correlation with criterion variables = intent associated with more acts of ψIPV and negative correlation with criterion variables = intent associated with fewer acts of ψIPV. Positive correlation between intent = intentions co-occur and negative correlation between intent = intentions do not co-occur.

p < .05. **p < .001.

The unexpected finding that joking intent was associated with a higher level of aggressive tendencies for males and with physical IPV for males and females warranted further examination. We reasoned that the co-occurrence of joking intent with any other intent appraisals may change its meaning. We tested this idea by computing partial correlations controlling for the total number of non-joking intent appraisals. This allowed us to isolate joking intent versus all other intent appraisals. The number of intent appraisals reduced the association between joking intent and aggressive tendencies for males and females (r_partial = .015, p = .79 and r_partial = −.02, p = .68, respectively) and with physical IPV for females but not males (r_partial = .154, p < .01 and r_partial = .07, p = .14, respectively). Thus, when joking intent was the only reported intent, the association between joking intent and aggressive behaviors was not significant.

Discussion

The primary aim of the present study was to test a survey of ψIPV perpetration in adolescence, the RBS. Contrary to our hypothesis, tests of confirmatory factor models did not support denigrating, controlling, and intrusive behaviors as three distinct subtypes of ψIPV. Instead, a single common factor was supported, reflecting one underlying construct of ψIPV for both in-person and electronic forms for males and females. The pattern of findings suggests denigration, controlling, and intrusive behaviors are all part of the same unidimensional construct. Adolescent relationships differ from young adult relationships in several ways, which may explain, in part, the different factor structure across age groups. For adolescents, who are starting to explore intimacy and romance in relationships that tend to be brief and less committed than adult relationships, the subtypes of their ψIPV behaviors may not be fully differentiated. The lack of gender differences in factor structure is like previous work on adolescent ψIPV (Orpinas et al., 2012; Taylor & Mumford, 2016).

Consistent with our hypothesis, the RBS improved prediction of aggressive tendencies and physical IPV for males and females, after accounting for the CADRI emotional abuse and threatening behavior subscales (Wolfe et al., 2001). That is, the addition of more varied RBS items alone, without consideration of intent, demonstrated incremental validity. Even though the RBS did not conform to the expected dimensions, the larger number of items covering a wider range of behaviors represent a measurement improvement over the range of items measured on the CADRI. There were several meaningful enhancements to the RBS in relation to the types of items and their wording. The RBS includes a larger assortment of items for controlling acts, both in-person and electronic communication, digital surveillance, and intrusions on a partner’s privacy. Additionally, the RBS added qualifiers to specify if certain common behaviors, such as texting a partner, persist after a request to stop. These qualifiers help to focus on unwanted acts, which are not mutually agreeable between partners. The range of items and qualifiers for unwanted acts are not included in nor redundant with the CADRI or similar measures of IPV for this age group (e.g., Safe Dates, Foshee et al., 1998; CTS-2, Straus et al., 1996). The RBS is a unidimensional scale measuring a larger variety of ψIPV behaviors than current measures, and it holds promise for refining the study of risk factors and correlates to strengthen ψIPV prevention.

A novel aspect of the RBS is the assessment of intent for engaging in ψIPV, which aims to improve the interpretation of ψIPV behaviors. To joke was the most reported intent for males and second most reported intent for females. These findings extend previous research on physical IPV (Fernández-Gonzalez et al., 2013; Foshee et al., 2007; Gonzalez-Mendez & Hernandez-Cabrera, 2009; Jouriles et al., 2009), demonstrating that joking intent is not limited to physically aggressive acts against a partner, and includes ψIPV behaviors, such as name-calling, threats to hurt, and telling a partner what to wear. Prior research has found that adolescents may use joking as a way of expressing affection, possibly for lack of maturity and inexperience with intimacy (Baxter, 1992; Capaldi & Crosby, 1997; Ryan & Mohr, 2005). However, other research has shown that aggression with joking intent is correlated with psychological distress, anxiety about the relationship, and depressive symptoms, particularly for those on the receiving end of these acts (Jouriles et al., 2009; Sargent et al., 2020).

The hypothesis that joking intent would be associated with a lower level of general aggressive tendencies and less physical IPV was not supported. In contrast, joking intent was associated with a higher level of aggressive tendencies for males and with physical IPV for males and females. Males’ greater socialization in aggressive behavior may lead them to discount the severity or impact of their psychologically aggressive behavior toward a partner, and thus consider it to be joking or playful. It is also plausible that joking intent is a defensive re-evaluation of their behavior afterwards to minimize its severity (e.g., Arriaga, 2002; Jouriles et al., 2009). The current findings suggest joking intent appraisals are complex, with multiple meanings. While males appraised their behavior as only joking, future research could clarify if they genuinely intended the act as a joke or as a playful gesture. By systematically measuring joking intent, we can improve our understanding of ψIPV behaviors occurring for this reason. We can evaluate if joking intent portends more serious IPV longitudinally, and make more accurate decisions about when and whether to define ψIPV with joking intent as problematic behavior at a specific point in time. The post-hoc partial correlation analysis lends some support to this idea. Once other intentions were statistically controlled, joking intent was not correlated with aggressive behaviors. From a developmental perspective, one may start out with joking intent as a way of getting close to someone, but if that does not work, the joking may turn aggressive.

The findings in relation to controlling and mistrust intent appraisals show promise for future research and practical application. About one-quarter of males and females reported controlling intent appraisal and about one-third reported mistrust intent. Both intents were associated with general aggressive tendencies and physical IPV for males and females. There has been a longstanding challenge among IPV scholars to measure coercive control, and the RBS provides a novel method for identifying adolescents who engage in ψIPV with coercive motives. Mistrust may both reflect and foster jealousy, erode trust, and increase anxiety in a relationship. It is plausible that adolescents who acknowledge mistrust motives may be at greater risk for stably maladaptive relationships including engaging in IPV across time and relationships.

Anger and showing care intent were among the most reported motives for perpetrating ψIPV. However, neither of these intents were correlated with general aggressive tendencies or physical IPV for males or females. In some instances, ψIPV behaviors may promote closeness, particularly if both partners appreciate the intended care and concern, whereas anger intent may reflect difficulty managing emotions or skills to effectively express. The finding regarding anger intent is somewhat surprising, since anger correlates with IPV at a moderate level in meta-analyses with adult samples (Birkley & Eckhardt, 2015). However, one possible explanation for this discrepancy is anger is more strongly related to moderate-severe physical IPV, and the current study measured ψIPV. Anger may be overshadowed by control and mistrust, intentions that tended to co-occur with anger.

Limitations

This study is not without limitations. This study relied on an online survey where outside influences could not be monitored or controlled. Although parents were advised to provide quiet and private space for survey completion, there is no guarantee that this occurred. However, all online surveys are subject to similar concerns. Replication with an in-person survey will strengthen results. It should also be acknowledged that many adolescents may not be self-aware about the intentions for their behaviors. Additionally, if they are self-aware, they may not recall the reasons for them or mischaracterize the intent. Another limitation is the generalizability of findings. Although the study used a large, self-selected sample of adolescents, and all U.S. regions were included, the parents who participated in the panel were more educated and had higher incomes, and adolescents who participated were more likely to identify as White and non-Latinx than the U.S. population. Finally, the sample did not include enough adolescents who identified as gender or sexual minorities to examine if the RBS results were similar for them. Future research testing invariance in minority samples is necessary.

Future Research Directions

The RBS is a new scale for assessing ψIPV behavior in adolescents that measures the intent for engaging in different forms of ψIPV behaviors, including joking. Intentions improved criterion validity with general aggressive tendencies and physical IPV, and there may be other correlates where different intentions affect results, such as relationship quality, attachment insecurity (in relation to mistrust), and the impact of ψIPV behaviors on a partner. Given the frequency with which joking intent was reported, it is particularly important for future research to examine the impact of these acts on the victim. Acts perpetrated as a joke may not be experienced as such by the recipient. Research on dyads can provide more complete information on both intent and impact. In this way, the impact on the victim can be evaluated for perpetrators who report different intentions. For example, it is important to know if joking, showing care, or controlling intent is experienced as positive or negative by the victim.

Prevention and Clinical Implications

The RBS measures a larger variety of ψIPV than existing scales, such as the CADRI, yet it yields a lower endorsement rate, suggesting it may capture more serious forms of behavior. This may be helpful in differentiating adolescents who engage in developmentally typical behavior from those at risk for patterns of harm in romantic relationships. A key finding relates to intent appraisals. There are distinct intentions for ψIPV acts that may be meaningfully related to outcomes for the relationship and each partner. Another important finding is intentions for ψIPV behaviors related to multiple motives, including joking. These distinctions between ψIPV intentions can inform prevention targets to foster intimacy, ranging from psychoeducation; alternatives to name calling, surveillance, or controlling behavior (even if intended as joking or showing care); and improved strategies for conflict resolution and emotion regulation.

In summary, adolescents’ reported perpetration of ψIPV on the RBS was unidimensional, but multidetermined, with both positive (e.g., show care) and negative (e.g., control, mistrust) intent. The RBS improved the prediction of aggressive tendencies and physical IPV after accounting for the CADRI emotional abuse and threatening behavior subscales. Importantly, the findings suggest that aggressive acts with joking intent should not be excluded from research on ψIPV, as has been suggested by some scholars (e.g., Hamby, 2016). Rather, continued examination of self-reported intentions for aggressive acts will increase researchers’ ability to isolate behaviors which may be developmentally typical, benign, positive, or harmful.

Supplemental Material

sj-docx-1-jiv-10.1177_08862605221140044 – Supplemental material for The Relationship Behavior Survey: A Comprehensive Measure of Psychological Intimate Partner Violence for Adolescents

Supplemental material, sj-docx-1-jiv-10.1177_08862605221140044 for The Relationship Behavior Survey: A Comprehensive Measure of Psychological Intimate Partner Violence for Adolescents by Michele Cascardi, Raghad Hassabelnaby, Hannah Schorpp, Amy M. Smith Slep, Ernest N. Jouriles and K. Daniel O’Leary in Journal of Interpersonal Violence

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interests with respect to the authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a federal grant from the National Institute of Justice (Grant # 2016-MU-CX-0007) and an internal research award from William Paterson University.

ORCID iDs

Michele Cascardi

Ernest N. Jouriles

Author Biographies

Michele Cascardi, PhD, is Professor of Psychology and Dean’s Fellow at William Paterson University. Her research interests include intimate partner and relationship violence, interventions for preventing violence, the impact of childhood victimization, and variables related to the development of post-traumatic stress disorder. She has developed and evaluated school-based programs to prevent teen dating violence and promote healthy dating relationships.

Raghad Hassabelnaby is a doctoral student in clinical psychology (PsyD) at William Paterson University of New Jersey. She holds a master’s degree in clinical and counseling psychology. Raghad was a recipient of the Community Service Grant Award from the New Jersey Psychological Association. Her clinical work and interests include mood and trauma disorders with children and adolescents.

Hannah Schorpp is a doctoral student in clinical psychology (PsyD). She holds a master’s degree in forensic psychology. Hannah’s clinical work focuses on rehabilitation psychology, working with individuals recovering from issues such as brain injuries, spinal cord injuries, and limb loss. Her research interests include intimate partner violence and sexual assault, with a goal of improving intervention strategies.

Amy M. Smith Slep, PhD, is a professor and the co-director of the Family Translational Research Group at New York University. Dr. Slep’s work on family maltreatment and prevention science has been funded by over 70 federal grants and contracts from funders such as the Institute for EducationalSciences, CDC, NIH, NIJ, DoD, U.S. Air Force, and the Department of Agriculture. She has authored over 150 publications.

Ernest N. Jouriles, PhD, is the Dale McKissick Endowed Professor of Psychology at Southern Methodist University. His research interests include sexual and relationship violence among adolescents and young adults, children’s exposure to interparent conflict and violence, variables associated with psychological adjustment among adolescent sexual violence survivors, and interventions for preventing violence and assisting individuals adversely affected by violence.

Daniel O’Leary, PhD, is Distinguished Professor of Psychology at Stony Brook University. He was chairman of the Psychology Department, and the Director of Clinical Training at Stony Brook University. He was also President of the Association of Behavior and Cognitive Therapy and the Association of Family and Conciliation Courts. His awards include Fellow of the American Association for the Advancement of Science, 2019.

References

Ackerman

J. M.

(2016). Over-reporting intimate partner violence in Australian survey research. British Journal of Criminology, 56, 646–667.

Arriaga

X. B.

(2002). Joking violence among highly committed individuals. Journal of Interpersonal Violence, 17, 591–610.

Bair-Merritt

M. H.

Shea Crowne

Thompson

D. A.

Sibinga

Trent

Campbell

(2010). Why do women use intimate partner violence? A systematic review of women’s motivations. Trauma, Violence, & Abuse, 11, 178–189.

Baxter

L. A.

(1992). Forms and functions of intimate play in personal relationships. Human Communication Research, 18, 336–363.

Bentler

P. M.

(1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246.

Bentler

P. M.

Bonett

D. G.

(1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88, 588–606.

Bernstein

I. H.

Gesn

P. R.

(1997). On the dimensionality of the Buss/Perry aggression questionnaire. Behaviour Research and Therapy, 35, 563–568.

Birkley

E. L.

Eckhardt

C. I.

(2015). Anger, hostility, internalizing negative emotions, and intimate partner violence perpetration: A meta-analytic review. Clinical Psychology Review, 37, 40–56.

Borrajo

Gámez-Guadix

Pereda

Calvete

(2015). The development and validation of the cyber dating abuse questionnaire among young couples. Computers in Human Behavior, 48, 358–365.

10.

Brooks-Russell

Foshee

V. A.

Ennett

S. T.

(2013). Predictors of latent trajectory classes of physical dating violence victimization. Journal of Youth and Adolescence, 42, 566–580.

11.

Brunner

Nagy

Wilhelm

(2012). A tutorial on hierarchically structured constructs. Journal of Personality, 80(4), 796–846.

12.

Buss

A. H.

Perry

(1992). The aggression questionnaire. Journal of Personality and Social Psychology, 63, 452–459.

13.

Caiozzo

C. N.

Houston

Grych

(2016). Predicting aggression in late adolescent romantic relationships: A short-term longitudinal study. Journal of Adolescence, 53, 237–248.

14.

Capaldi

D. M.

Crosby

(1997). Observed and reported psychological and physical aggression in young, at-risk couples. Social Development, 6, 184–206.

15.

Carney

M. M.

Barner

J. R.

(2012). Prevalence of partner abuse: Rates of emotional abuse and control. Partner Abuse, 3, 286–335.

16.

Cascardi

Avery-Leaf

(2015). Gender differences in dating aggression and victimization among low-income, urban middle school students. Partner Abuse, 6(4), 383–402.

17.

Cascardi

Kammen

Pearsall

Navarro

Connolly

(July, 2018). A conceptual typology for measuring adolescent relationship abuse: Developmental and contextual considerations. Poster presented at the International Family Violence and Child Victimization Research Conference, Portsmouth, NJ.

18.

Cascardi

Blank

Dodani

(2019). Comparison of the CADRI and CTS2 for measuring psychological and physical dating violence perpetration and victimization. Journal of Interpersonal Violence, 34, 3466–3491.

19.

Copp

J. E.

Johnson

W. L.

(2015). Patterns, precursors, and consequences of teen dating violence: Analyzing gendered and generic pathway. Final Report to the National Institute of Justice, Washington, DC.

20.

Dueber

D. M.

(2017). Bifactor Indices Calculator: A Microsoft Excel-based tool to calculate various indices relevant to bifactor CFA models. https://doi.org/10.13023/edp.tool.01 [Also available at http://sites.education.uky.edu/apslab/resources/]

21.

Dueber

D. M.

(2020). A bifactor approach to dimensionality assessment. Theses and Dissertations–Education Sciences, 59. https://uknowledge.uky.edu/edsc_etds/59

22.

Exner-Cortens

Gill

Eckenrode

(2016). Measurement of adolescent dating violence: A comprehensive review (Part 1, behaviors). Aggression and Violent Behavior, 27, 64–78.

23.

Fernández-González

O’Leary

K. D.

Muñoz-Rivas

M. J.

(2013). We are not joking: Need for controls in reports of dating violence. Journal of Interpersonal Violence, 28, 602–620.

24.

Flynn

Graham

(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.

25.

Follingstad

D. R.

(2011). A measure of severe psychological abuse normed on a nationally representative sample of adults. Journal of Interpersonal Violence, 26, 1194–1214.

26.

Follingstad

D. R.

Coker

A. L.

Lee

Williams

C. M.

Bush

H. M.

Mendiondo

M. M.

(2015). Validity and psychometric properties of the measure of psychologically abusive behaviors among young women and women in distressed relationships. Violence Against Women, 21, 875–896.

27.

Foshee

V. A.

Bauman

K. E.

Arriaga

X. B.

Helms

R. W.

Koch

G. G.

Linder

G. F.

(1998). An evaluation of Safe Dates, an adolescent dating violence prevention program. American Journal of Public Health, 88, 45–50.

28.

Foshee

V. A.

Bauman

K. E.

Linder

Rice

Wilcher

(2007). Typologies of adolescent dating violence: Identifying typologies of adolescent dating violence perpetration. Journal of Interpersonal Violence, 22, 498–519.

29.

Furman

M. J.

Low

S. M.

(2007). The rocky road of adolescent romantic experience: Dating and adjustment. In Engels

R. C. M. E.

Kerr

Stattin

(Eds.), Friends, Lovers and Groups: Key Relationships in Adolescence (pp. 61–80). John Wiley & Sons Ltd.

30.

Gonzalez-Mendez

Hernandez-Cabrera

J. A.

(2009). Play context, commitment, and dating violence: A structural equation model. Journal of Interpersonal Violence, 24, 1518–1535.

31.

Hamby

(2016). Self-report measures that do not produce gender parity in intimate partner violence: A multi-study investigation. Psychology of Violence, 6, 323–336.

32.

L. T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55.

33.

Hughes

H. M.

Massura

C. E.

Anukem

O. V.

Cattage

J. S.

(2016). Women college students’ reasons for engaging in psychological dating aggression: A qualitative examination. Journal of Family Violence, 31, 239–249.

34.

Johnson

M. P.

(2011). Gender and types of intimate partner violence: A response to an anti-feminist literature review. Aggression and Violent Behavior, 16, 289–296.

35.

Johnson

W. L.

Manning

W. D.

Giordano

P. C.

Longmore

M. A.

(2015). Relationship context and intimate partner violence from adolescence to young adulthood. Journal of Adolescent Health, 57, 631–636.

36.

Jouriles

E. N.

Garrido

Rosenfield

McDonald

(2009). Experiences of psychological and physical aggression in adolescent romantic relationships: Links to psychological distress. Child Abuse & Neglect, 33, 451–460.

37.

Langhinrichsen-Rohling

McCullars

Misra

T. A.

(2012). Motivations for men and women’s intimate partner violence perpetration: A comprehensive review. Partner Abuse, 3, 429–468.

38.

Leisring

P. A.

(2013). Physical and emotional abuse in romantic relationships: Motivation for perpetration among college women. Journal of Interpersonal Violence, 28, 1437–1454.

39.

Lenhart

Madden

Smith

Purcell

Zickuhr

Rainie

(2011). Teens, kindness and cruelty on social network sites: How American teens navigate the new world of “Digital Citizenship.” Pew Internet & American Life Project.

40.

Lenhart

Anderson

Smith

(2015). Chapter 1: Basics of Teen Romatic Relationships. In Teens, Technology, and Romantic Relationships. Pew Research Center. https://www.pewresearch.org/internet/2015/10/01/basics-of-teen-romantic-relationships.

41.

Mehari

K. R.

Farrell

A. D.

(2018). Where does cyberbullying fit? A comparison of competing models of adolescent aggression. Psychology of violence, 8, 31–42.

42.

Messinger

A. M.

Fry

D. A.

Rickert

V. I.

Catallozzi

Davidson

L. L.

(2014). Extending Johnson’s intimate partner violence typology: Lessons from an adolescent sample. Violence Against Women, 20, 948–971.

43.

Miller

Williams

Cutbush

Gibbs

Clinton-Sherrod

Jones

(2013). Dating violence, bullying, and sexual harassment: Longitudinal profiles and transitions over time. Journal of Youth and Adolescence, 42, 607–618.

44.

Muñoz-Fernández

Sánchez-Jiménez

(2020). Cyber-aggression and psychological aggression in adolescent couples: A short-term longitudinal study on prevalence and common and differential predictors. Computers in Human Behavior, 104, 106191.

45.

Murphy

C. M.

Hoover

S. A.

(1999). Measuring emotional abuse in dating relationships as a multifactorial construct. Violence and Victims, 14, 39–53.

46.

Muthén

L.K.

Muthén

B.O.

(2017). Mplus User’s Guide (7th ed.). Los Angeles, CA: Muthén & Muthén.

47.

Orpinas

Nahapetyan

Song

McNicholas

Reeves

P. M.

(2012). Psychological dating violence perpetration and victimization: Trajectories from middle to high school. Aggressive Behavior, 38, 510–520.

48.

Reidy

D. E.

Ball

Houry

Holland

K. M.

Valle

L. A.

Kearns

M. C.

Marshall

K. J.

Rosenbluth

(2016). In search of teen dating violence typologies. Journal of Adolescent Health, 58, 202–207.

49.

Ryan

K. M.

Mohr

(2005). Gender differences in playful aggression during courtship in college students. Sex Roles, 53, 591–601.

50.

Sargent

K. S.

Yule

Bridges

Jouriles

E. N.

Grych

J. H.

(2020). Do instructions intended to reduce false positives improve the measurement of physical partner violence victimization among adolescents and young adults? Psychology of Violence, 10, 152–161.

51.

Schreiber

J. B.

Nora

Stage

F. K.

Barlow

E. A.

King

(2006). Reporting structural equation modeling and confirmatory factor analysis results: A review. The Journal of Educational Research, 99, 323–338.

52.

Southworth

Finn

Dawson

Fraser

Tucker

(2007) Intimate partner violence, technology, and stalking. Violence Against Women, 13(8), 842–856.

53.

Straus

M. A.

Hamby

S. L.

Boney-McCoy

S. U. E.

Sugarman

D. B.

(1996). The revised conflict tactics scales (CTS2) development and preliminary psychometric data. Journal of Family Issues, 17(3), 283–316.

54.

Taylor

B. G.

Mumford

E. A.

(2016). A national descriptive portrait of adolescent relationship abuse: Results from the National Survey on Teen Relationships and Intimate Violence. Journal of Interpersonal Violence, 31, 963–988.

55.

Tucker

L. R.

Lewis

(1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38(1), 1–10 analysis.

56.

Wolfe

D. A.

Scott

Reitzel-Jaffe

Wekerle

Grasley

Straatman

A. L.

(2001). Development and validation of the conflict in adolescent dating relationships inventory. Psychological Assessment, 13, 277–293.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB