Although research on dating violence perpetrated against adolescent girls has become more prevalent, there continues to be a great need for better understanding of both its correlates and consequences (Fernet et al., 2019; Satyanarayana et al., 2015). Based on a nationally representative sample of American high schoolers, the Centers for Disease Control and Prevention (CDC, 2019) has provided evidence that adolescent girls are more likely than boys to be victims of physical dating violence (9.3% vs. 7.0%) as well as sexual dating violence (12.6% vs. 3.8%). Both rates have been rising since 2017 (physical dating violence victimization from 9.1 to 9.3%, and sexual dating violence victimization from 10.7 to 12.6%), and both rates remain high among adolescent girls (CDC, 2019). These trends are alarming, as dating violence victimization has been shown to be associated with both negative short- and long-term developmental adjustment measures, including internalizing problems, such as depression and anxiety (e.g., Ackard et al., 2007; Exner-Cortens et al., 2013). A recent review also concludes that girls are at greater risk for experiencing more adverse consequences than boys (Taquette & Monteiro, 2019). To better address and prevent victimization and to potentially mitigate the negative effects of being victimized, a better understanding of underlying mechanisms of how victimization experiences are linked with internalizing problems is needed, something already recognized by Banyard and Cross (2008) over a decade ago, and as more recently echoed by Sianko et al. (2019). In addition, despite an improved awareness of dating violence against adolescent girls, studies focused on adolescent girls in rural areas remain scarce (e.g., Burton et al., 2016; Taylor & Xia, 2020). Given that girls and women residing in rural contexts may experience more dating violence, have fewer resources, and suffer greater negative consequences in comparison to those from urban or suburban areas (Martz et al., 2016; Peek-Asa et al., 2011), it seems paramount to focus more on this particular population. Thus, to address these gaps and limitations in both understanding and the extant literature, the current investigation studied dating violence victimization among rural adolescent girls in the American rural South. More specifically, it tested the relationships between victimization experiences and measures of internalizing problems as well as whether and how maternal peer approval conditioned these links. In the following section, the relevant literature informing the present study is reviewed.

The Relationship Between Dating Violence Victimization and Internalizing Problems

Dating violence victimization comes in several forms. For research purposes, victimization is often distinguished, namely physical and psychological. Physical dating violence is frequently operationalized as having been beaten or hit by a dating partner (e.g., Banyard et al., 2008), and it can also include forced sexual acts (e.g., Foshee et al., 2013). Alternatively, psychological dating violence is marked by insults and psychological control on the part of the perpetrator (e.g., Foshee et al., 2013; Romito et al., 2013).

The evidence shows that both types of dating violence are linked with a number of negative adjustment outcomes, including internalizing problems (e.g., Ackard et al., 2007; Markz et al., 2016; Taquette & Monteiro, 2019); these include depression, anxiety, and low self-esteem. An association between dating violence victimization and depressive symptoms has been established in cross-sectional studies carried out in the United States as well as abroad (Brar et al., 2020; Martz et al., 2016; Romito et al., 2013; Van Ouytsel et al., 2017). Longitudinal work focusing on the same link has also substantiated the relationship between dating violence victimization and depression (Ackard et al., 2007; Exner-Cortens et al., 2013). This was further supported by work carried out on minority youth; based on a sample of African American and Hispanic girls, threatening behaviors by their partner was associated with greater depressive symptoms (Teitelman et al., 2011). Dating or intimate partner violence has also been linked to anxiety/panic attacks (Garthe et al., 2021; Romito et al., 2013), posttraumatic stress (Callahan et al., 2003), a greater likelihood of suicidal ideations (Marzt et al., 2016; Romito et al., 2013), and lower levels of self-esteem (Van Ouytsel et al., 2017). Comparatively few studies have focused on the relationships between intimate partner violence and measures of anxiety or low self-esteem. Nevertheless, the existing evidence supports that intimate partner or physical dating violence among teens is associated with a variety of internalizing problems, which in turn have been shown to be linked to poorer school outcomes (e.g., Banyard & Cross, 2008); this evidence supports the idea that internalizing symptoms themselves potentially adversely affect or at least are associated with a whole host of negative adjustment measures for teenagers. This evidence calls for continued research to develop a better understanding of these associations between dating violence victimization and internalizing problems among adolescents.

The Role of Parental Peer Approval

One important sphere of influence that has often been studied related to dating violence victimization risks and consequences includes the role played by parental socialization (e.g., Earnest & Brady, 2016; East et al., 2010; Howard et al., 2003; Lantrip, 2015). Positive parental behaviors, such as supportive parenting, parental care and knowledge, play important roles in delaying the sexual debut (i.e., onset of sexual intercourse), in promoting healthier sexual behaviors and decision-making, in lowering the risk of dating violence victimization, and in buffering the negative impacts of dating violence victimization among adolescents (Earnest & Brady, 2016; Howard et al., 2003; Mumford et al., 2016; Rodgers & McGuire, 2012; Smith & Donnelly, 2001). Meta-analytical research has provided evidence that both parental support and monitoring have served as protective factors for teen dating violence victimization (Hébert et al., 2019). Maternal parenting seems to be particularly salient based on existing evidence. For example, a higher quality mother-daughter dyadic relationship which includes improved communication about dating relationships and sexual behaviors, might instill healthier beliefs about dating in general, and also lower risks of premature sexual behaviors as well as other forms of dating violence victimization (Lantrip, 2015; Ombayo et al., 2019). Greater maternal monitoring has also been shown to be associated with a lower level of risk for experiencing dating aggression and dating victimization (East et al., 2010; Pflieger & Vazsonyi, 2006). These findings provide important implications on the importance of maternal parenting as a protective mechanism in trying to better understand the precise mechanisms underlying dating aggression risk, dating violence, as well as dating violence victimization.

Interventions focused on altering parenting behaviors and involvement have shown promising preventive effects on reducing adolescent girls’ victimization. Such efforts have focused on improving positive parenting and decreasing coercive parenting behaviors, a known risk factor of dating violence victimization (Lundgren & Amin, 2015); they have also focused on decreasing the likelihood of associations with aggressive dating partners, that in turn lowered the risk for victimization (Ehrensaft et al., 2018). Finally, interventions have focused on altering and improving parental response efficacy when dealing with and confronting dating abuse as well as coercive communication, both designed to improve adolescent knowledge and awareness about abuse relationships (Foshee et al., 2012). In conclusion, these intervention efforts highlight the key role by parents and parenting behaviors in being protective mechanisms for adolescent girls related to dating violence victimization.

However, although positive parenting should have a protective effect on the risk for and consequences of dating violence, research has provided the paradoxical evidence that high parental autonomy support may be associated with negative adjustment outcomes among adolescents. For example, low maternal responsiveness coupled with high maternal autonomy granting behaviors might in fact be associated with greater sexual risk taking (Lanza et al., 2013). Similarly, autonomy supporting behaviors have been found to predict greater risk for dating violence victimization among adolescent girls two years later (Niolon et al., 2015). The associated sexual risk taking could be due to low monitoring as a result of high autonomy support, which gives adolescents more freedom and opportunities to engage in sexual behavior (Sneed et al., 2009), and early sexual behavior may escalate into risky sex (Lanza et al., 2013). Regarding the association between maternal autonomy supporting behaviors and dating violence victimization among adolescent girls, this could possibly be because of conflicts with the dating partner who demands dependence while girls strive for more autonomy/independence as supported by the mother (Niolon et al., 2015). Thus, high parental autonomy support behaviors might in fact not be protective, but rather potentiate dating violence victimization risk among girls. The current study was particularly interested in further testing and better understanding whether autonomy granting behaviors moderated the link between dating violence victimization and risk for internalizing problems.

Parental attitudes toward peers is an important part of autonomy granting behaviors (Lanza et al., 2013; Niolon et al., 2015), whether about friendships or dating relationships. Little research has considered how parental attitudes about peer relationships influence adolescent dating violence experiences or their adjustment following victimization. The existing evidence indicates that maternal restrictions on peer relationships in fact increase the risk of deviant peer associations among adolescent girls (Updegraff et al., 2010). On the other hand, parental involvement in peer relationships, which includes supporting peer-oriented activities, mediating or providing advice on peer relationships, or granting more autonomy, is associated with more positive peer relationships and less conflict with peers (Mounts, 2004; Updegraff et al., 2001). Thus, parents play an important role in how peer relationships develop as well as how influential they are on adolescent development. The current study sought to develop a more nuanced understanding of and further test the paradoxical effects found in previous work on the importance of autonomy granting behaviors, particularly related to violence victimization and its consequences. It sought to further test and better understand how maternal peer approval was associated with adjustments in adolescent girls who experienced dating violence victimization.

The Current Study

The current study aimed to test the relationships between dating violence victimization and internalizing problems (i.e., depression, anxiety, and low self-esteem) and the extent to which maternal peer approval conditions these links in a sample of adolescent girls located in the rural South of the United States. The following hypotheses were developed based on a review of the literature:

H1

General dating violence victimization (GDVV) would be positively associated with each internalizing problem (i.e., depression, anxiety, and low self-esteem).

H2

Maternal peer approval would moderate the link between GDVV and each internalizing problem (i.e., depression, anxiety, and low self-esteem).

H3

Sexual dating violence victimization (SDVV) would be positively associated with each internalizing problem (i.e., depression, anxiety, and low self-esteem).

H4

Maternal peer approval would moderate the link between SDVV and each internalizing problem (i.e., depression, anxiety, and low self-esteem).

Method

Participants

Data of the current study came from a larger study on 943 adolescents between 10 and 21 years of age (M = 14.80 years, SD = 1.85; 52.8% female) from two middle schools and two high schools in American rural South in 2015. The overall sample included 77.1% European American, 9.1% African American, 2.4% Native American, 8.1% Latino youth, and 3.1% youth who belong to other racial/ethnic groups. The rural county where the sample came from had a low population density (69.0 persons per square mile), and households in that area had relatively low median household incomes ($40,933 versus $53,046 at the national median level); 15.8% of its population were found to be college graduates, smaller than the national figure (22.8%); and 17.2% of the county population lived below the poverty line (U.S. Census Bureau, 2015). The voluntary and anonymous study was approved by a University Institutional Review Board. Consent was obtained from both the students and their parents before the students filled out the surveys. Data in the present study were collected using both paper and pencil and online surveys.

For the current study, only the 335 girls from the two high schools, where data of dating violence victimization were collected, were included (Mage = 15.77 years, SD = 1.12). The sample included 80.3% European American, 6.7% African American, 1.5% Native American, 8.2% Latino youth, and 3.3% youth who belong to other racial/ethnic groups. About 19.1% of the sample took the paper and pencil survey, while 80.9% took the online survey. To address this difference, main study analyses controlled for any effects by data collection method.

Measures

Demographics

Participants were asked to indicate their age, which was recorded as birth month and year. Seven options of home situations were provided for reporting family structure: biological parents (1), biological mother only (2), biological father only (3), biological mother and stepfather (4), biological father and stepmother (5), biological parent and significant other (6), and other (7). Responses were then recoded into a dichotomous variable indicating whether they lived in a two-parent (biological parents or other types of parents) household (1) or other types of household (0). Socioeconomic status (SES) was computed by averaging three items. They included the standardized mean of maternal and paternal education, an ordered categorical variable with six response options: does not apply (1), finished elementary or junior high school/middle school (through 8th grade) (2), finished high school (through 12th grade) (3), finished some college or technical school (4), has a college degree (4 years) (5), and has a graduate degree (advanced degree, e.g., masters or doctorate) (6), and the standardized mean self-reported annual family income, an ordered categorical variable with five response options: $20,000 or less (1), $20,000 to $35,000 (2), $35,000 to $60,000 (3), $60,000 to $100,000 (4), and $100,000 or more (5). Participants also indicated their race/ethnicity by choosing one of the following options: European American (1), African American (2), Latino/Hispanic (3), Asian American (4), American Indian (5), Pacific Islander (6), Arab (7), and other (8). Responses were recoded into a dichotomous variable, where 1 represented European American, while 0 represented other racial/ethnic groups.

Survey Administration Method

A paper and pencil survey method was implemented at one high school and an online survey method at the other. The paper and pencil method was coded as 0 and the online survey method as 1; 64 girls took the paper and pencil survey, while 271 girls completed the online survey.

General Dating Violence Victimization (GDVV)

General dating violence victimization (GDVV) was measured by an item that summarizes eight types of dating violence victimization from the original measure of Victimization in Dating Relationships by Foshee et al. (1996). Participants responded to the question “How many times has any person with whom you have been on a date ever pushed, slapped, punched, or beat you up, thrown an object at you or hit you with an object, or forced you to have sex or do sexual things?” with 5-point Likert-type response options: never (1), occasionally (2), sometimes (3), often (4), and very often (5). A higher response score indicated a higher frequency of experiencing general dating violence victimization.

Sexual Dating Violence Victimization (SDVV)

Sexual dating violence victimization (SDVV) was measured by an item that summarizes two types of sexual violence victimization from the original measure of Victimization in Dating Relationships by Foshee et al. (1996). Participants responded to the question “How many times has any person with whom you have been on a date ever forced you to have sex or do sexual things?” with 5-point Likert-type response options: never (1), occasionally (2), sometimes (3), often (4), and very often (5). A higher response score indicated a higher frequency of experiencing sexual dating violence victimization.

Depression

The 4-item depression subscale of the Weinberger Adjustment Inventory–Short Form (Weinberger, 1990) was used to assess participants’ depression. Participants were asked to indicate their degree of agreement with three statements, including that they felt sad or unhappy, felt lonely, and felt so down and unhappy that nothing made them feel much better, with 5-point Likert-type response options: false (1), somewhat false (2), not sure (3), somewhat true (4), and true (5); or almost never (1), not too often (2), sometimes (3), often (4), and almost always (5). Response scores were averaged across the three items, and a higher score indicated a higher level of depression. The scale was internally consistent (α = 0.87).

Anxiety

The 4-item anxiety subscale of the Weinberger Adjustment Inventory–Short Form (Weinberger, 1990) was used to assess participants’ anxiety. Participants were asked to indicate their degree of agreement with four statements, such as that they worried too much about things that were not important, and that they felt afraid something terrible might happen to them or somebody they cared about, with 5-point Likert-type response options: false (1), somewhat false (2), not sure (3), somewhat true (4), and true (5); or almost never (1), not too often (2), sometimes (3), often (4), and almost always (5). Response scores were averaged across the four items, and a higher score indicated a higher level of anxiety. The scale was internally consistent (α = 0.82).

Low Self-Esteem

The 3-item low self-esteem subscale of the Weinberger Adjustment Inventory–Short Form (Weinberger, 1990) was used to assess participants’ self-esteem. Participants were asked to indicate their degree of agreement with 3 statements, including how true it was that they did not like themselves very much, felt so sad that they wished they were somebody else, and were not very sure of themselves, with 5-point Likert-type response options: false (1), somewhat false (2), not sure (3), somewhat true (4), and true (5). Response scores were averaged across the three items, and a higher score indicated a lower level of self-esteem. The scale was internally consistent (α = 0.90).

Maternal Peer Approval

The 3-item maternal peer approval subscale of the maternal version of Adolescent Family Process Measure (Vazsonyi et al., 2003) was used to assess adolescents’ reported frequency of maternal approval of their friends, boyfriends, and going out with friends. Participants responded with 5-point Likert-type response options: never (1), occasionally (2), sometimes (3), often (4), and very often (5). Response scores were averaged across the three items with a higher score indicated a higher level of maternal peer approval. The scale was internally consistent (α = 0.76).

Plan of Analysis

First, normality (i.e., based on skewness and kurtosis statistics) and missingness in the data (using Little’s MCAR test for missing completely at random and sensitivity analysis for missing at random) were examined. In addition, as there were two survey administration methods used, to test whether there were any differences between the two, an analysis of covariance (ANCOVA) was completed with the main focal variables (GDVV, SDVV, maternal peer approval, depression, anxiety, and self-esteem) as the dependent measures, the method of data collection as the independent variable, and age as the covariate. In cases where the results were statistically significant for a dependent variable, the survey administration method was added as a control variable for all relevant analyses.

Second, descriptive statistics and correlations among study variables were computed in SPSS 26. Lastly, moderation analyses were conducted using the PROCESS macro, version 3 (Hayes, 2017) in SPSS 26, testing the following links: GDVV-depression, GDVV-anxiety, GDVV-self-esteem, SDVV-depression, SDVV-anxiety, and SDVV-self-esteem, with maternal peer approval as the moderator for all links. Variables included in the interactions were centered for ease of interpretation and to reduce multicollinearity.

Results

In a first step, normality and missingness of data as well as whether the two different survey administration modes impacted participant responses were examined. There was some evidence of non-normality of independent and dependent variables; assumptions of homoscedasticity of residuals and homogeneity of residual variance of estimations were examined and acceptable, and therefore, no transformations of study variables were conducted. The percentage of missingness on study variables ranged from 0.6 to 11%; results of Little’s MCAR test indicated that the data were missing completely at random, justifying the use of list-wise deletion implemented in regression analyses in SPSS; to examine the study power after list-wise deletion, G*Power was used to do a post hoc power analysis, and results indicated that the current study had sufficient power (i.e., > 0.80). Results of an ANCOVA provided evidence that responses on the three internalizing problem measures differed by the two survey administration modes, and therefore survey administration method was added as a control variable in subsequent analyses.

Demographics and Correlations

Next, descriptive statistics (see Table 1) and correlations (see Table 2) among study variables were computed. Among the 299 girls (89.3% of total sample) who provided data on the frequency of general dating violence victimization, 48 (16.0%) selected response options other than never. Among the 298 girls (89.0% of total sample) who provided data on the frequency of sexual dating violence victimization, 50 (16.6%) selected response options other than never. Based on correlations, general dating violence victimization (GDVV) was positively associated with self-esteem, indicating that more victimization was related to a lower self-esteem; sexual violence victimization (SDVV) was positively associated with depression and self-esteem, indicating that more victimization was related to more depression and a lower self-esteem. GDVV was not associated with depression, and neither GDVV nor SDVV was associated with anxiety. Maternal peer approval was negatively associated with both types of dating violence victimization, depression, anxiety, and self-esteem, indicating that more maternal peer approval was related to less GDVV, less SDVV, lower levels of depression and anxiety, and a higher self-esteem.

Table 1 Descriptive statistics on respondent demographics
Full size table
Table 2 Descriptive statistics of and correlations among study variables
Full size table

In addition, older girls seemed to have experienced GDVV more often, reported a higher level of depression, a lower level of self-esteem, and more maternal peer approval. Girls from a two-parent family reported a lower level of depression and anxiety, a higher level of self-esteem, and a higher level of maternal peer approval; similar patterns of correlations were found between girls with a higher SES and other measures. European American girls seemed to have a lower level of depression and a higher level of maternal peer approval, as compared to girls from other racial/ethnic groups.

Moderation Analyses

Finally, a series of moderation analyses were conducted to test the moderation effects of maternal peer approval on the links between the two dating violence victimization measures and internalizing problems among adolescent girls. As based on correlation analysis, GDVV was not correlated with depression or anxiety, and SDVV was not correlated with anxiety, moderation analysis was only conducted on these links: GDVV-self-esteem, SDVV-depression, and SDVV-self-esteem.

Results of moderation analyses are presented in Table 3. In the model with GDVV as the predictor and low self-esteem as the outcome, moderation effects by maternal peer approval were statistically significant: b = 0.17, se = 0.09, p = 0.048, 95% CI [0.002, 0.347], and so were the main effects by GDVV: b = 0.23, se = 0.11, p = 0.034, 95% CI [0.02, 0.44], and by maternal peer approval: b = − 0.37, se = 0.09, p < 0.001, 95% CI [− 0.54, − 0.19], after accounting for other variables in the model; statistically significant control variables included family structure and survey administration methods, with girls from two-parent families or girls who took the online survey reporting higher levels of self-esteem, after accounting for other variables in the model. The significant main effects by GDVV on low self-esteem partly supported Hypothesis 1, and the significant moderation effects by maternal peer approval partly supported Hypothesis 2. The model explained 16% of variance in self-esteem. For ease of interpretation of the moderation effects, a plot was developed at three levels of maternal peer approval (-1 SD, mean, and + 1 SD; see Fig. 1a). The figure shows that the relationship between GDVV and self-esteem differed across the three levels of maternal peer approval: At -1 SD of maternal peer approval, no relationship was found; at the mean, a positive relationship was observed between victimization and low self-esteem; and finally, at + 1 SD of maternal peer approval, the positive relationship appeared stronger. The observed pattern across the three different levels of maternal peer approval provided evidence that girls who experienced GDVV and reported higher levels of maternal peer approval were at greater risk for low self-esteem.

Table 3 Results of moderation analyses
Full size table
Fig. 1
figure 1

a Moderation effects by maternal peer approval on the general dating violence victimization-low self-esteem link. b Moderation effects by maternal peer approval on the sexual dating violence victimization-depression link. c Moderation effects by maternal peer approval on the sexual dating violence victimization-low self-esteem link

Full size image

In the model with SDVV as the predictor and depression as the outcome, moderation effects by maternal peer approval were statistically significant: b = 0.20, se = 0.09, p = 0.020, 95% CI [0.03, 0.37], and so were the main effects by maternal peer approval: b = − 0.31, se = 0.09, p < 0.001, 95% CI [− 0.47, − 0.15], after accounting for other variables in the model, although main effects by SDVV were not statistically significant. The significant moderation effects by maternal peer approval partly supported Hypothesis 4. None of the control variables reached significance. The model explained 14% of variance in low self-esteem. Again, for ease of interpretation, a plot was developed across three levels (− 1 SD, mean, and + 1 SD) of maternal peer approval (see Fig. 1b). The figure shows that the relationship between SDVV and depression again differed across the three levels of maternal peer approval: At − 1 SD, no relationship was observed; at the mean, a positive relationship was found, and at + 1 SD, a stronger positive relationship was observed. The patterns of the slopes across the three levels of peer approval provided evidence that girls with SDVV experiences and higher levels of maternal peer approval were at greater risk for depression.

In the model with SDVV as the predictor and low self-esteem as the outcome, moderation effects by maternal peer approval were statistically significant: b = 0.26, se = 0.10, p = 0.008, 95% CI [0.07, 0.45], and so were the main effects by SDVV: b = 0.24, se = 0.11, p = 0.026, 95% CI [0.03, 0.45], and by maternal peer approval: b = − 0.37, se = 0.09, p < 0.001, 95% CI [− 0.55, − 0.19], after accounting for other variables in the model. The significant main effects by SDVV on low self-esteem partly supported Hypothesis 3, and the significant moderation effects by maternal peer approval partly supported Hypothesis 4. Statistically significant control variables included family structure and survey administration methods, with girls from two-parent families or girls who took the online survey reporting higher levels of self-esteem, after accounting for other variables in the model. The model explained 17% of variance in self-esteem. Again, a plot was developed across the three levels (− 1 SD, mean, and + 1 SD) of maternal peer approval (see Fig. 1c). The figure shows that the relationship between SDVV and self-esteem differed across the three levels of maternal peer approval: At − 1 SD of maternal peer approval, no relationship was observed; at the mean, a positive relationship was found between victimization and low self-esteem, while at + 1 SD, a stronger positive relationship was observed. This showed that girls with SDVV experiences and higher levels of maternal peer approval were at greater risk for low self-esteem.

Discussion

The current study sought to examine the relationships between general dating violence victimization/sexual dating violence victimization and internalizing problems as well as the role of maternal peer approval in the relationships among a sample of rural adolescent girls. Research results and findings are summarized and discussed as follows.

Main Effects by Victimization and Moderation Effects by Maternal Peer Approval

The results partly supported all four hypotheses. Specifically, consistent with Hypothesis 1, general dating violence victimization (GDVV) positively predicted low self-esteem, indicating that more dating violence victimization was associated with lower levels of self-esteem; as consistent with Hypothesis 2, maternal peer approval moderated the association between GDVV and low self-esteem, although the main effects by GDVV did not achieve statistical significance; as consistent with Hypothesis 4, maternal peer approval moderated the association between sexual dating violence victimization (SDVV) and depression; as consistent with Hypothesis 3, SDVV positively predicted low self-esteem, indicating that more sexual dating violence victimization was associated with lower self-esteem; and as consistent with Hypothesis 4, maternal peer approval moderated the association between sexual dating violence victimization (SDVV) and low self-esteem. The hypothesized links between GDVV or SDVV and anxiety were not supported by study results.

These results indicate that there were relationships between GDVV or SDVV and internalizing problems including depression and low self-esteem, and that these relationships were conditioned by levels of maternal peer approval. A closer look at the patterns of the moderation effects shows that adolescent girls who had GDVV or SDVV experiences and reported greater peer approval were at relatively greater risk to report depressive symptoms or to report low self-esteem. These results suggest that high parental autonomy support is not only related to more dating violence victimization among adolescent girls as found by Niolon et al. (2015), but it may also put girls who experienced dating violence victimization at elevated risk for internalizing problems. Therefore, parental autonomy support behaviors, which include peer approval and trusting daughters in their peer related activities, might conversely not prevent negative outcomes, such as internalizing problems after experiencing GDVV or SDVV. This study finding implicates that in terms of the approval of peer relationships or activities, parental approval that provides high levels of freedom may place daughters at greater risk for dating violence victimization and associated internalizing problems.

In fact, in addition to dating violence victimization and associated internalizing problems, high parental autonomy has also been found to be related to other negative adjustment outcomes. For example, interactions with mothers who provided autonomy support were associated with decreased social skills as well as increased delinquency (Allen et al., 2002); maternal autonomy granting has also been found to be associated with lower mother-adolescent relationship quality, especially for lower SES families (McElhaney & Allen, 2001). Findings from the current study further support previous research in that parental autonomy support may not be protective of poor adjustment outcomes; they also extend findings which have shown that parental autonomy support is associated with adolescent adjustment, following a negative event, such as dating violence victimization. Therefore, the potential impact by and effects by parental autonomy support on adolescent adjustment are mixed.

Study findings concerning the observed negative correlates, including depression and low self-esteem, of experiencing dating violence are consistent with previous research. For example, more frequent experiences of both general and sexual dating violence victimization were associated with lower levels of self-esteem, consistent with findings by Van Ouytsel et al. (2017) based on a Belgian sample of adolescents. This evidence shows that experiencing dating violence victimization likely impacts a victim’s self-image. More frequent experiences of sexual dating violence victimization were also associated with greater levels of depressive symptoms, consistent with previous research (e.g., Ackard et al., 2007; Brar et al., 2020; Markz et al., 2016; Romito et al., 2013), documenting how victimization experiences impact a girl’s affect and mood, not a surprising finding. Unexpectedly, this link was not found between the general dating violence victimization and depressive symptoms in the current study. In addition, no associations were found between either general or sexual dating violence victimization and anxiety, inconsistent with previous research (Garthe et al., 2021; Romito et al., 2013). A number of explanations could underlie these null findings, including the specified timeframe of measurement. In fact, the two dating violence victimization variables were assessed as life-time experiences, while internalizing problems asked about the present status. More specifically, the anxiety measure used by Romito et al. (2013), for instance, focused on an anxiety crisis or panic attacks, generally more severe symptoms than what was measured in the present study. On the other hand, Garthe et al. (2021) combined both perpetration and victimization of dating violence in one variable/class, which was then used to predict anxiety symptoms, thus potentially obscuring the link between victimization and anxiety symptoms. In addition, anxiety was assessed as a broad construct in the present study, potentially also a reason why no associations were observed between dating violence and anxiety. Nevertheless, current study findings are generally consistent with previous research in describing associations between dating violence victimization and internalizing problems, despite the null finding with anxiety. As it is known that internalizing problems are in turn associated with a host of adjustment problems, including poor academic performance or suicidal behaviors (e.g., Banyard & Cross, 2008; Soto-Sanz et al., 2019), the current study findings provide important support for the necessity of additional research as well as resulting prevention and intervention efforts directed toward dating violence victimization among adolescent girls in rural areas.

Study Limitations and Research Implications

First, data were collected from only two schools in a county located in the American rural South, and therefore study findings may not be generalized beyond the context of these schools; however, the findings are consistent with those from previous work. Future research should capitalize on larger and more representative sample.

Second, as cross-sectional data were used, results can only be interpreted as being correlational in nature, rather than implying causality. Therefore, even though GDVV and SDVV were used as the predictors, the limitation of the data used only permit a conclusion that internalizing problems are important correlates. It is also conceivable that the tested relationship works in reverse, where girls who developed internalizing symptoms early might be at elevated risk for dating violence victimization (e.g., Foshee et al., 2004). Future research should use longitudinal data to better study the long-term correlates of dating violence victimization.

Third, the current study did not focus on correlates of mother’s low or high peer approval/autonomy granting regarding peer relationships. In fact, underlying motivations or reasons for parental autonomy granting behaviors might have contributed to the current paradoxical findings. Parents might grant high autonomy to their children to help them become more independent or simply because they are not able to set sufficient limits due to emotional unavailability (e.g., suffering from depression themselves) or even due to being neglectful, for instance. Research has shown that parental emotional problems or neglectful parenting are associated with a variety of adolescent adjustment problems (e.g., Christ et al., 2017; Monti & Rudolph, 2017). In addition, high parental autonomy support, no matter for what reason, could be stressful to adolescents, as although adolescents strive for more independence while growing up, emotional care and continued closeness from parents remains paramount (Christ et al., 2017). This might explain why high autonomy support is, associated with some negative experiences among adolescent girls, such as greater sexual risk taking (Lanza et al., 2013; Niolon et al., 2015) or even more dating violence victimization (Sneed et al., 2009). Therefore, future research needs to study a variety of parental behaviors, including autonomy support as well as emotional responsiveness (Lanza et al., 2013), to better understand the links and underlying relationships between dating violence victimization and internalizing problems. It should be noted that in the present study, the interpretation of maternal peer approval effects on internalizing behaviors need to be considered together with victimization experiences; neither effect should be interpreted in isolation.

Last but not the least, the current study is one of very few that tested moderation effects by maternal peer approval, and thus, more research is needed to replicate study findings. In addition, using a direct measure of maternal approval of dating relationships might be better suited to assess how mothers’ support or lack of support for dating relationships impacts their daughters’ behaviors and risk for dating violence victimization. A measure of dating violence victimization that includes multiple items would capture victimization experiences more accurately. Moreover, future research should also test other competing explanatory mechanisms of the victimization-internalizing links, such as potential moderation effects by other parental behaviors (Banyard & Cross, 2008) or coping strategies (Coffey et al., 1996), to better understand the links and inform interventions.

Conclusion and Prevention Implications

The current study tested the relationships between two dating violence victimization items and internalizing problems including depression, anxiety, and low self-esteem, and the role maternal peer approval played in these relationships. Study findings show that both dating violence victimization measures were associated with internalizing problems, and that the associations differed across levels of maternal peer approval. They also show the salience of both victimization measures and maternal peer approval in understanding rates of depression and low self-esteem among adolescent girls.

Girls who experienced dating violence were in fact at elevated risk for internalizing problems if they also indicated greater maternal peer approval; this would suggest that higher levels of peer approval, indicating possibly greater freedom in dating for daughters, might be detrimental. Study findings have important implications for interventions that target parenting in preventing dating violence victimization or negative effects of dating violence victimization among adolescent girls in rural developmental contexts. The finding that dating violence victimization was associated with internalizing problems also has implications for early intervention efforts with individuals who have experienced dating violence. More specifically, interventions could include providing parents with guidance and support about better ways to interact or communicate with their adolescent daughters, and to improve openness and communication, that in turn would increase awareness of their daughters’ relationship status. This might also focus on supervision and monitoring strategies surrounding peer and dating relationships. Furthermore, schools should make every effort to ensure safe dating and encourage girls to immediately speak up and seek help in case of threats or experiences of victimization. Following victimization experiences, a whole host of measures by parents, schools, and mental health professionals need to be put into action to prevent or at least alleviate potential internalizing problems. With greater education and awareness through prevention efforts by parents, by teachers, as well as by clinicians, the potentially deleterious effects from experiencing dating violence victimization might be prevented or at least lessened.