Youth intimate partner violence: prevalence, characteristics, associated factors and arenas of violence

ABSTRACT

Internationally, it is established that many youth experience violence within their romantic intimate relationships (youth intimate partner violence, Youth IPV). In Sweden, however, despite a well-developed research field on domestic violence and violence against children, knowledge about the prevalence of Youth IPV remains lacking. Therefore, taking its starting point in a mixed-method study on youth IPV in Sweden, this study presents incidence rates of Youth IPV victimization in a sample of Swedish youth, as well as associated factors to such victimization. Arenas of violence – i.e. settings where the violence takes place – are also addressed. In total, 59.7% (n = 526) of study respondents reported having experienced some form of IPV either once or multiple times. The results suggest gender differences: girls reported higher rates of victimization, especially for multiple experiences of violence. Markers for IPV victimization are presented. It is shown that Youth IPV happens in arenas to which adults have access. Viable markers for victimization are suggested; furthermore, the belief that ‘only at-risk youth experience IPV’ is challenged. It is critical that Youth IPV be considered a social problem in Sweden, deserving of specific attention and interventions. Youth IPV does not happen behind closed doors, but under the same roof as adults who have the obligation to protect them (i.e. parents and school staff) and this offers unique opportunities to respond and help. Further research should investigate the support offered to IPV-exposed youth, thereby shifting the focus to how this social problem is tackled.

KEYWORDS:

• youth intimate partner violence
• dating violence

Introduction

In recent decades, the knowledge base on dating violence/youth intimate partner violence (IPV)¹ has expanded, establishing that many youth experience violence within their romantic intimate relationships. Such violence is shown to be physical, sexual, and/or psychological, and to cause serious impact (Stonard et al. 2014; Romito, Beltramini, and Escribà-Agüir 2013; Barter et al. 2017).² Thus, the previous understanding of IPV as primarily an adult problem has been challenged (Barter 2011), and youth IPV has received increasing attention within social work practice and research.

The research field on dating violence/youth IPV was launched in North America; however, in recent years, scholars across Europe have conducted studies underlining that youth IPV is a social problem that requires societal attention (Korkmaz 2017). Correspondingly, a report from the Swedish National Council for Crime Prevention indicates that youth IPV is a substantial issue among Swedish young people (Brå 2018). Nevertheless, in Sweden, despite a well-developed research field on adult IPV and violence against children, knowledge about prevalence of and associated factors to youth IPV remains lacking. Therefore, taking its starting point in a mixed-method study on youth IPV in Sweden, this study draws upon quantitative data and aims to present incidence rates and characteristics of youth IPV victimization among a sample of Swedish youth, as well as associated factors to such victimization. Arenas of violence – i.e., where the violence take place – will also be addressed, placing youth IPV in a physical social context. Moreover, this study follows the extensive five-country EU-funded European study, ‘Safeguarding teenagers’ intimate relationships’ (STIR), which is its methodological forerunner. The rationale for the STIR project came from previous research on teenage partner violence among British teenagers, thus drawing upon existing empirical findings (Barter 2009; Barter et al. 2009). Overall, the STIR study aimed to further explore partner violence among teenagers, as well as investigate the role of digital means.

In 2017, the STIR study, conducted in Bulgaria, Cyprus, England, Italy, and Norway, presented prevalence rates from each of these countries (Barter et al. 2017). In Norway, 42.9% of participating youth reported that they had experienced at least one type of IPV, once or on several occasions. By type, digital violence was most prevalent (reported by 29.1% of the Norwegian sample), psychological violence was second most prevalent (25.9%), sexual violence third (18.8%), and physical violence the least frequent (12.8%) (Hellevik and Øverlien 2016). Overall, the STIR project shows gender differences both in victimization rates (prevalence rates were significantly associated with gender in England and Norway) and in the subjective impact of violent experiences. Girls reported more violence and significantly greater negative impact than boys, leading the researchers to conclude that IPV experiences may be differentiated by gender (Barter et al. 2017). Consistent with other studies (e.g. Gadd et al. 2015), the STIR findings unequivocally show that gender is a variable that needs to be acknowledged when investigating youth IPV.

Building on qualitative findings from the Swedish research project (Korkmaz and Överlien 2020; Överlien, Hellevik, and Korkmaz 2019), as well as the STIR study, this article poses the following research questions:

1. How prevalent is IPV victimization in a sample of Swedish youth?

2. What are the characteristics and determinants of youth IPV victimization?

3. To what degree is youth IPV victimization gendered?

4. In which arenas³ do youth IPV take place?

Prevalence of violence in youth romantic relationships

Many studies have brought forward quantitative evidence on youth experiences of intimate partner violence (IPV), presenting incidence rates of both victimization and perpetration. Overall, challenges emerge when comparing IPV prevalence rates across countries and between studies (Barter and Stanley 2016; Stanley et al. 2015; Hamby et al. 2012; Nocentini et al. 2011). Methodologies (including terminology used, definition of violence, samples) as well as ‘cultural factors’⁴ may influence the results. Further, youth may differ in how they relate the operationalizations used in research to their own experiences (Karlsson et al. 2018), contributing to the challenge. Thus, comparisons between studies should be made with caution. It is important to note, however, that prevalence rates may indeed also vary due to actual differences in experiences.

In regard to reported prevalence rates, studies show great variation (Stonard et al. 2014; Wincentak, Connolly, and Card 2017; Korkmaz 2017). For example, focusing just on research conducted in Europe, the prevalence of general experience of IPV (experiencing at least one type of violence) is found to vary from 23.1% in a Belgian study (Van Ouytsel, Ponnet, and Walrave 2017) to 45% in an English study (Fox, Hale, and Gadd 2014). Regarding physical violence victimization, studies also reveal great variation: from 13.4% in Portugal to 48.8% in a study conducted in French-speaking Switzerland (Hamby et al. 2012). Moreover, in an international meta-analytic review that included countries outside of Europe, (Wincentak, Connolly, and Card 2017) report that the prevalence of sexual dating violence among adolescents varied from less than 1% to 54%, depending upon the study (31 studies were included); their own meta-analysis indicates a rate of 9%. Hence, studies show a notable difference in reported prevalence rates; nevertheless, methodological factors aside, the body of quantitative evidence shows that youth IPV is a social problem requiring societal attention.

Methods

Youth intimate partner violence – prevalence, contexts, and youth voices

This article represents a sub-study within a larger mixed methods research project on youth IPV in Sweden, approved by the ethical review board in Stockholm in May 2016.⁵ Overall, the research project aims to enhance our understanding of IPV-exposed youth and their experiences, where qualitative analysis of interview data allows for an in-depth exploration while survey data gives an idea of the extent and characteristics of the problem. This specific article draws upon the survey data gathered within the research project, carried out during 2017 and 2018.

Sample

All high schools (n = 31) in one specific county in Sweden were invited to participate, as a way to recruit high school students (age 15–19) for the study. The county has both rural and urban areas, and a varied demographic and socioeconomic composition. In 2018, the proportion of social assistance recipients in the county was 4.3% (national average = 4.0%). The same year, the share of foreign-born individuals was 18% (national average = 19.1%). County income levels as well as the tax base strength are virtually the same as the national averages.

Of the invited schools, nine agreed to participate and allowed the researchers to visit to provide information about the study and survey. The participating schools had differing profiles, and included schools offering vocational programmes as well schools offering higher education preparatory programmes, with some schools located in rural areas while others had an urban catchment area. The level of participation varied among schools. In some schools all students had the opportunity to participate. In others, mainly the larger schools, students received information about the study and the opportunity to participate depending on teacher willingness. In total, 1004 high school students were invited to participate in the study. Of these, 956 chose to complete the survey, giving a response rate of 95.2%.

Ethics

The questionnaires were distributed by the researchers, allowing the students to ask questions about the project. The students received oral as well as written information about the study, and it was stressed that participation was voluntary and not part of their ordinary schoolwork.⁶ It was made clear that the questionnaire was anonymous, and the students who undertook it were reassured that their answers would remain confidential. Students were asked to respect one another and not to look at the person sitting next to them. Respondents who finished the survey early were invited to take a pop culture quiz while the others finished, as a way to keep them occupied and not make it obvious who needed more time to fill out the survey, since needing more time could indicate having experience of IPV. Before visiting each school, contact was made with the school counsellor to make sure that he or she would be available on the day of the school visit. While in the classrooms, participants were informed that they could seek support if needed at the counsellor’s office, and they were also informed about other support services.

Research measures – the survey

This study was granted permission⁷ to use the same survey instrument as did the STIR study (Barter et al. 2015). The STIR questionnaire was translated into Swedish and some questions were added: for example, about where the violence took place, in an effort to address in which arenas youth IPV happens. The translated questionnaire was reviewed by a youth advisory group, which was asked to comment on the wording and terms used, ensuring that it was appropriate for their age group.

The questionnaire consisted of eight different sections and a total of 114 items, of which fifteen examined the experience of violence specifically (see Table 1). The violence items involved questions regarding different types of violence, divided into different sections. The questionnaire used Likert scales, where respondents indicated if something had happened ‘never’, ‘once’, ‘a few times’, or ‘often’. Further, questions on where the violence took place (e.g. at school, at home, at the partner’s home) were included after the sections regarding violence perpetrated face-to-face (excluding psychological violence perpetrated online), establishing in which arenas the violence took place. To clarify, as an example, after the questions regarding violence perpetrated face-to-face, the respondent was asked, if he or she had experienced face-to-face violence, to indicate where it took place. The alternatives were: at your house; at your partner’s house; at school; at a leisure club; at a friend’s house; at a club; or other. Multiple answers were possible. Moreover, also included were background questions regarding with whom the respondents resided, whether or not they felt they were generally doing well in school, the ages of their partners, experiences with being bullied, experiences of domestic violence, and whether they and their parents were born in Sweden or abroad.

Table 1. Survey items regarding experiences with IPV

Type of Violence
Psychological violence	Have any of your partners ever …	put you down in nasty ways? shouted at you/screamed in your face/called you names? said negative things about your appearance/body?said negative things about your /friends/family? threatened to hurt you physically?
Psychological violence perpetrated online	Have any of your partners ever …	put you down or sent you nasty messages online or via mobile phone? posted nasty messages about you online that others could see? sent you threatening messages online or by mobile phone? used mobile phones or social networking sites to try to control whom you can be friends with or where you can go? constantly checked up on what you have been doing/who you have been seeing, for example, by sending you messages or checking your social networking page all the time? used mobile phones or social networking sites to stop your friends liking you, for example, pretending to be you and sending nasty messages to your friends?
Physical violence	Have any of your partners ever …	used physical force such as slapping, pushing, hitting or holding you down? used more severe physical force such as punching, strangling, beating you up, hitting you with an object?
Sexual violence	Have any of your partners ever …	pressured you into kissing, intimate touching or something else? physically forced you into kissing, intimate touching or something else? pressured you into having sexual intercourse? physically forced you into having sexual intercourse?
Severe sexual violence	Have any of your partners ever …	pressured you into having sexual intercourse? physically forced you into having sexual intercourse?

Statistical analyses

All statistical analyses were performed using SPSS (version 26). Consistent with the STIR study, the violence items were categorized into four overall types of violence: psychological (5 items), psychological perpetrated online (6 items), physical (2 items), and sexual (4 items) (see Table 1). Moreover, severe sexual violence was singled out, consisting of two items regarding penetration (‘Have any of your partners ever 1. pressured you into having sexual intercourse? 2. physically forced you into having sexual intercourse?’). Severe sexual violence was distinguished so that separate analysis on sexual violence involving penetration would be possible. As a first step, for every form of IPV, experiences of each violence type were calculated, called ‘general experience’, presented in Table 2 where the IPV experiences are accounted for

Table 2. Victimization Rates Proportion of respondents who report experiences of IPV victimization within each index of violence, presenting experiences in total, as well as isolated and repeated experiences respectively. Sample sizes vary and are accounted for in each column. Numbers are given in percentages and divided by gender. ‘Severe sexual violence’ refers to such experiences within the sexual violence index.

Type of Violence	General Experiences Total (n = 548–680)	Girls (n = 299–357)	Boys (n = 241–318)	Isolated Experiences Total (n = 548–680)	Girls (n = 299–357)	Boys (n = 241–318)	Multiple Experiences Total (n = 548–680)	Girls (n = 299–357)	Boys (n = 241–318)
Psychological Violence Face-to-Face	40.2%	45.5% **	33.6% **	28%	28.5%	27.6%	12.2%	17% ***	6% ***
Psychological Violence Online	49.6%	56.1% ***	41.6% ***	39.1%	42.1%	34.6%	10.5%	14% *	7% *
Physical Violence	16.7%	19.5% *	12.8% *	12.1%	13.5%	9.8%	4.6%	6% *	3% *
Sexual Violence	29%	44.4%***	9.9%***	21.5%	33%***	7.4%***	7.5%	11.4%**	2.5% **
Severe Sexual Violence	18.1%	28.8% ***	4.4%***	12.2%	20%***	2.4%***	5.9%	8.8% **	2% **

*p <.05 **p < 0.01 *** p <.001

To further investigate the experiences of IPV and make it possible to indicate whether they were of isolated or multiple nature, an index for each type was made (‘never’ = 1, ‘once’ = 2, ‘a few times’ = 3, and ‘often’ = 4). Within each IPV index, cut-off points were set to mark isolated experience as well as multiple experience. Regarding isolated experience, the cut-off point was set at the number of items in each form of IPV respectively + 1 (excluding respondents who answered ‘never’ to each item), through the number of items x 2. For example, for psychological violence face-to-face (5 items), the cut-off point for isolated experience was set at 6 through 10 points. A score in this low range would indicate that a respondent first and foremost had experienced isolated incidences of IPV (e.g. by answering ‘once’ to all five items, and thus receiving 10 points), or just a few items within the IPV index (e.g. by answering ‘often’ to one item, but ‘never’ to the other ones, and thus receiving 4 + 4 points). A score of 11–20 would indicate that a respondent had experienced multiple incidences of IPV.

Moreover, prevalence rates were calculated for each type of violence. Descriptive statistics are reported in numbers and percentages and with Chi-square results when appropriate.

Places where the violence took place (e.g. at home, at school, or at a club) are called as arenas of violence, and presented for the categories ‘physical violence’ and ‘psychological violence perpetrated face-to-face’, ‘sexual violence’, and ‘severe sexual violence’. For each type of violence, it was calculated how many times every arena respectively was reported, which is presented as a percentage. Thus, the unit of analysis is arenas rather than incidents of violence.

To estimate associations between individual background factors and youth IPV victimization, we used a multiple logistic regression analysis consisting of a model for each type of violence. Based on findings from previous research (e.g. Hellevik and Øverlien 2016; Ruel et al. 2020; Hellevik 2017; Walters and Espelage 2018), seven independent variables were included in the analysis as individual background factors: female gender, having experience of domestic violence, experience of being bullied, generally having older partners, academic achievement (self-assessed), living with two parents, and foreign background. Respondent were considered to have a ‘foreign background’ if they reported being born abroad, and/or if both their parents were born outside of Sweden. First, a simple linear regression for each independent variable was performed with each type of violence as the dependent variable, showing crude odds ratios. Thereafter, a multiple logistic regression analysis⁸ was executed. Results are presented with odds ratios and confidence intervals for each examined variable.

Procedure

In the first section of the questionnaire, respondents were asked to answer whether or not they had any experience with intimate relationships. This ‘gateway question’ was worded broadly to include as many variations of youth intimate relationships as possible. When handing out the survey, the researchers emphasized that ‘relationship’ also referred to shorter and less serious relationships, such as ‘hooking up’, on-and-off relationships’, ‘friends with benefits’, ‘dating’, or any other relationship they would define as romantic.

Sample descriptives

A sample of 956 students chose to participate: 483 girls (50.7%) and 457 boys (48%) (12 students identified as ‘other’ and 4 did not answer the question). Due to the very small numbers reporting that they identified as neither boy nor girl, the results are based on girls’ and boys’ experiences. Of the students participating, 74.5% (n = 956) reported that they had experience of being in a relationship. The findings presented below relate to this subsample of youth with such experience, and who hence were invited to complete the whole survey. Nevertheless, the sample size varies from one section to the next, as well as within sections, due to differences in the number of respondents for each separate item. Thus, we present separate sample sizes within the tables as well as throughout the text.

Of the subsample of respondents who had been in a relationship, 51% were girls and 47% boys. The mean age was 17.2 years (SD = 0.97). Around a third of the youth stated that they generally had older partners; of these, 82% were girls. Heterosexual relationships were the most common type: 91% of the girls and 94% of the boys reported only having relationships with the opposite sex. A majority of the youth lived with parents/custodians (single parent 18.7%; two parents 72.3%), whereas 9% had other living arrangements: e.g. with a partner, on their own, or at a residential treatment facility. Of the subsample, 15.4% were considered to have a ‘foreign background’; i.e. they reported being born abroad, and/or that their parents were born outside of Sweden.

Results

In the following, characteristics of youth intimate partner violence (IPV) will be unpacked by presenting rates of psychological violence perpetrated face-to-face, psychological violence perpetrated online, physical violence, and sexual violence, respectively. Additionally, ‘severe sexual violence’, referring to violence involving penetration, was singled out within the sexual violence index to investigate experiences of such sexual violence more thoroughly. Arenas of violence will also be presented, highlighting where these different types of violence take place. Moreover, in order to discuss how individual background factors – such as gender or exposure to domestic violence – might be associated with IPV victimization, results from a multivariate logistic regression model, performed for each type of violence, are presented.

The extent and characteristics of IPV victimization

In total, 59.7% (n = 526) of the respondents had experienced some form of intimate partner violence (IPV) on one or multiple occasions. Psychological violence perpetrated online was most prevalent, with almost half of the sample reporting experiences of such violence, while physical violence was the least common type of violence reported, but still with a 16.7% prevalence rate (see Table 2). Of the respondents who had experienced some form of IPV (n = 314), 36.3% had experienced one type of violence, 28.4% had experienced two types, 22.6% had experienced three types, and 12.7% reported having experienced all four types of violence. Overall, the rates suggest that a great proportion of the respondents had been subjected to IPV, and that many of them experienced more than one type of violence.

To unfold how characteristics of victimization might differ, we examined both isolated experiences and multiple experiences, as well as girls’ and boys’ experiences separately. Although the rates of multiple experiences are lower than the rates of isolated experience, a substantial number of the respondents had experienced IPV on multiple occasions.

Focusing on gender, girls reported having experienced violence more extensively in all domains. This gender difference holds true especially in relation to sexual violence, where a substantially higher proportion of girls reported victimization compared with boys; this is true for girls reporting isolated experiences as well as multiple experiences. Boys seem foremost to experience psychological violence, in particular psychological violence online; again, this holds true for boys reporting isolated experiences as well as multiple experiences.

Arenas of violence

Overall, the results show that a substantial number of youth experience violence within their romantic relationships. To put these experiences in a social context, and to further our understanding of the characteristics of IPV victimization, we focus on where the violence took place, i.e. on different arenas of violence. This advances our knowledge of where youth are subjected to violence by a partner, following on from a contextual approach to youth IPV (Överlien, Hellevik, and Korkmaz 2019). In the following, five different arenas are presented: the most reported ones, as well as the arenas that previous research shows can typify the social settings of youth everyday life (e.g. the school arena, at a friend’s house, at a club) (Korkmaz and Överlien 2020). The remaining arenas, such as ‘outside’ and ‘at a leisure club’, are broadly categorized as ‘other arenas’ due to the small numbers reported. It should be noted that the unit of analysis is arenas rather than incidents of violence.

The results show that youth first and foremost experience IPV at home, either in their own home or their partner’s home, independent of the type of violence. The partner’s house stands out as the most prevalent reported arena, especially in regards to severe sexual violence.

In Table 3, it is shown that psychological violence face-to-face and physical violence were the most prevalent type of violence taking place at school. Noteworthy, sexual violence, including severe sexual violence, is also reported to take place at school premises. Youth also seem to experience violence in the arenas of a friend’s house and a club/pub, but these arenas showed the lowest rates.

Table 3. Proportions of Arenas Where Violence Took Place (n = 241–678).

Type of Violence | Arenas	At the youth’s house	At the partner’s house	At school	At a friend’s house	At a club/pub	Other Arena	Total
Violence Face-to-Face (n = 345)	26,7	36,5	13,6	2,6	7,6	13	100
Sexual Violence (n = 678)	31,3	37,6	5,2	4,6	6,3	15	100
Severe Sexual Violence (n = 241)	33	46	5	2,4	3,7	9,9	100

n = arenas reported where violence took place, answers reported in percentage. ‘Violence Face-to-Face’ refers to Psychological Violence Face-to-Face, and Physical Violence. ‘Severe Sexual Violence’ refers to sexual violence including penetration, n = 241, of which also are included in sexual violence n = 678. ‘Other Arena’ refers to e.g. outside, at a leisure club.

Overall, the results show that youth experience violence in several different arenas: arenas that are typical settings for many youth. Arguably, young victims of IPV have few chances for a respite from violence, since it seems to occur in arenas where youth should be protected.

Associated factors to youth IPV

Table 4 shows the results of the multivariate logistic regression analyses where background factors have been analysed in relation to each type of IPV, presenting five separate models (psychological violence face-to-face; psychological violence online; physical; sexual; severe sexual) and factors found to be associated with IPV victimization.

Table 4. Odds ratios for associated background characteristics with experiencing different forms of IPV.

	Psychological Violence Face-to-Face (n = 612)		Psychological Violence Online(n = 609)		Physical Violence (n = 515)		Sexual Violence (n = 503)		Severe Sexual Violence (n = 518)
Background Characteristics	Crude	Adjusted	Crude	Adjusted	Crude	Adjusted	Crude	Adjusted	Crude	Adjusted
Female gender	1.66 (1.22–2.28)***	1.19 (0.76–1.65)	1.77 (1.31–2.40)***	1.49 (1.02–2.16)*	1.64 (1.08–2.50) *	0.99 (0.58–1.68)	6.72 (4.21–10.71)***	5.20 (3.04–8.86)***	7.50 (4.07–13.81)***	5.11 (2.56–10.19)***
Low academic achievements	1.43 (0.90–2.25)	1.36 (0.82–2.27)	1.75 (1.10–2.81)*	1.79 (1.07–2.99)*	1.55 (0.89–2.72)	1.42 (0.77–2.59)	1.30 (0.75–2.26)	1.15 (0.60–2.20)	1.09 (0.57–2.09)	0.94 (0.45–1.98)
Having been bullied	2.02 (1.47–2.80)***	1.77 (1.24–2.53)**	1.74 (1.28–2.38) ***	1.56 (1.11–2.19)**	2.36 (1.51–3.68)***	2.02 (1.23–3.30)**	1.53 (1.05–2.25)*	1.13 (0.73–1.76)	1.99 (1.25–3.14)**	1.43 (0.85–2.39)
Generally having older partners	2.890 (2.08–4.02)***	2.44 (1.64–3.63)***	2.10 (1.51–2.91)***	1.63 (1.10–2.42)*	2.31 (1.52–3.49) ***	1.92 (1.14–3.22)*	3.46 (2.34–5.12)***	1.83 (1.15–2.90)**	3.97 (2.52–6.23)***	2.14 (1.25–3.61)**
Experiencing Domestic Violence	1.76 (1.28–2.43)**	1.35 (0.94–1.93)	1.64 (1.20–2.24)**	1.34 (0.94–1.89)	2.53 (1.62–3.93) ***	2.05 (1.26–3.35)**	2.28 (1.55–3.35)***	1.81 (1.16–2.82)**	2.00 (1.28–3.13)**	1.43 (0.85–2.41)
Living with two parents	0.40 (0.29–0.57)***	0.574 (0.389–0.847)**	0.56 (0.40–0.79)**	0.758 (0.514–1.118)	0.41 (0.27–0.63) ***	0.67 (0.41–1.09)	0.53 (0.36–0.80)**	0.68 (0.41–1.11)	0.48 (0.30–075) **	0.57 (0.33–0.98)*
Foreign background	1.16 (0.76–1.76)	1.11 (0.70–1.76)	1.26 (0.83–1.93)	1.37 (0.87–2.17)	1.06 (0.61–1.84)	0.98 (0.54–1.79)	1.33 (0.80–2.22)	1.39 (0.78–2.48)	0.58 (0.28–1.14)	0.49 (0.23–1.03)

*p <.05 **p < 0.01 *** p <.001

In our analyses, girls are found to be at a much higher risk (OR = 5.195) of being victims of sexual violence generally, as well as severe sexual violence (OR = 5.110), compared to boys, adjusted for the included background variables. In two other models, however, gender does not have explanatory value. This has been examined more thoroughly through pre-analyses where variables have been excluded from the models. Notably, in the model related to psychological violence (perpetrated face-to-face), when the variable of generally having an older partner was excluded, gender became a marker for such victimization. Hence, when controlling for generally having an older partner, gender is not significant, suggesting that gender does not have an independent explanatory value. On the other hand, having an older partner is arguably a gendered factor (cf. Spierings 2012; Vézina and Hébert 2007), which is shown e.g. in the present sample (of the youth who stated that they generally had older partners, 82% were girls).

Self-reported low academic achievement is significantly associated only with experiencing psychological violence online. This indicates that doing poorly in school does not generally increase the odds of being a victim of IPV.

Further, experiencing multiple victimization of different kinds (polyvictimization) – experiencing domestic violence, or having been bullied – seems to be a marker for IPV victimization. Hence, such adverse experiences may increase the risk of being subjected to violence by a partner.

Looking at family characteristics, living with two parents is found to decrease the risk (OR = 0.574) for experiencing psychological violence face-to-face, as well as for severe sexual violence (OR = 0.569). Neither parental background – whether the parents were born in a foreign country or in Sweden – nor whether or not youth were born in Sweden is found to affect the risk of being subjected to any type of IPV.

Discussion

In this study, we wanted to present incidence rates for intimate partner violence (IPV) in a Swedish youth sample. Additionally, we aimed to unpack characteristics of youth IPV victimization, focusing on prevalence of isolated as well as multiple experiences, on exploring gender differences in this respect, on the background factors associated with IPV, and on the different ‘arenas’ in which IPV takes place.

Associative factors – challenging the belief in ‘at-risk-youth’

Our results indicate that many youth are subjected to violence within intimate romantic relationships. When unpacking the associated factors to such victimization (Hellevik 2017), we found that a number of individual background variables are markers for IPV victimization, including polyvictimization and generally having older partners. In contrast, we also found that self-reported low academic achievement, as well as whether or not parents were born in a foreign country, do not seem to affect the risk of becoming a victim of IPV to any great extent.

Overall, our findings indicate that viable markers for IPV victimization exist, suggesting that some youth might be more at risk for experiencing such violence. However, the results also challenge any belief that ‘only at-risk youth experience IPV’, since self-reported low academic achievement is significantly associated only with psychological violence online. Thus, generally in our models, self-assessed low academic achievement does not seem to affect the risk of being subjected to IPV, and thereby it is fair to question academic success as a protective factor in this respect. This corresponds with findings from one qualitative study on responses to youth IPV (Korkmaz and Överlien 2020), where we showed that if anything, academic success possibly works as a factor that hinders youth from receiving support.

Capturing a gendered phenomenon – the importance of frequency and severity

How we measure violence has implications for how we understand it. Hence, great responsibility lies with the researcher and his or her definitions and operationalization of questions. In a number of publications, Sylvia Walby and colleagues have been critical of survey methodology, arguing that the questions posed underestimate gendered patterns of violence and abuse (cf. Walby and Towers 2017; Walby, Towers, and Francis 2016). In a re-analysis of the latest Norwegian prevalence study on intimate partner violence by Thoresen and Hjemdal (2014), Bjørnholt and Hejmdal (2018) show how women were exposed to a higher number of severe incidents, to repeated violence, and to multiple forms of severe violence, compared to the male respondents. The re-analysis also shows that sexual violence from a partner is almost exclusively experienced by women. These findings are in line with other prevalence studies, such as the Swedish National Council for Crime Prevention’s study on violence in close relationships. They conclude that although violence rates are similar between women and men, it is more common for women to be subjected to more frequent and severe violence and, as a consequence, to have greater need for help, primarily in the form of medical assistance (Brå 2014).

In our study, focusing on the total reported experiences of IPV, more female than male respondents report victimization. Nevertheless, when unpacking the characteristics, the difference between the genders becomes clearer. For isolated experiences, the gender difference is not that prominent, but when it comes to multiple experiences of violence, it is evident that more girls reported having experienced violence on several occasions, compared to boys. The gender difference in victimization holds especially true in relation to sexual violence and severe sexual violence, which, as argued by among others Bjørnholt and Hejmdal (2018), underlines the importance of including sexual violence when measuring intimate partner violence, something we will discuss more thoroughly below. Furthermore, Bjørnholt and Hjemdal also suggest adding the fear of being injured or killed, as this greatly changes the gender ratio. Although we did not include fear in our survey, our qualitative analysis of interviews with victimized girls, reported in Överlien, Hellevik, and Korkmaz (2019), showed that fear was one of the themes dominant in almost all of the young informants’ narratives. More severe and frequent violence and abuse is also likely to result in more fear and greater harm. Hence, including survey questions of frequency, severity, and the consequences of the victimization, such as experiencing fear and having to seek out help, all contribute to uncovering the gendered dimension of violence and abuse, refuting the understanding of intimate partner violence as gender-symmetrical.

Unpacking intimate partner sexual violence – the necessity of combining quantitative and qualitative methods

Within the relatively small body of research on intimate partner sexual violence (IPSV), most studies have focused on adults, leading scholars to express an omnipresent need for further research on young people’s experiences of such violence (Fernet et al. 2019). In this study, following their lead, we chose to explore the phenomenon of IPSV more thoroughly by singling out severe IPSV, referring to violence including penetration. More specifically, the youth in our study were asked to indicate whether they had been pressured or physically forced into intercourse, resulting in notably high incidence rates: 28.8% for girls and 4.4% for boys. Nevertheless, IPSV is arguably a tricky phenomenon to unpack, permeated by societal norms and beliefs. Gendered norms can constitute a barrier for girls to identify sexual violent acts as violence (Korkmaz forthcoming), and this may partly also hold true for boys.

Moreover, previous research shows that when sexual violence includes low invasiveness or force, participants tend to minimize or not recognize the behaviour as such (Fernet et al. 2019). Here, a youth’s general inexperience of sexual relationships may also affect the perception of IPSV (Överlien, Hellevik, and Korkmaz 2019; Korkmaz forthcoming). Qualitative data have shown that girls express feeling like they owe sexual activities to their partner, and giving consent despite not really wanting to (Fernet et al. 2019). Correspondingly, in one qualitative study, we showed that female youth might ‘agree’ to have sex as a way to protect themselves from further harm (Korkmaz and Överlien 2020). Overall, these factors put incidence rates of IPSV in a different light. Would a girl who ‘agreed’ to have sex as a strategy to avoid further abuse answer ‘yes’ to a question about whether she had been pressured or forced into intercourse? Relatedly, (Logan, Walker, and Cole 2015) argue that ‘it is likely that women reporting forced sex are the “tip of the iceberg” in terms of the extent and depth of sexual violence perpetrated by partners’ (p. 126). Therefore, even though we agree on the necessity of including items on sexual violence when measuring intimate partner violence, we see a need to use mixed methods to truly show the characteristics of IPSV and make it possible to unpack this severe type of IPV.

Youth IPV taking place in the proximity of adults?

In this study, we show how youth IPV takes place in different arenas. Arguably, the notion of arenas of violence follows on from research that takes a contextual approach to youth IPV (Överlien, Hellevik, and Korkmaz 2019; Fernet et al. 2019; Logan, Walker, and Cole 2015), where the situational setting, as well as youth social networks and their responses, have been in focus. In two studies analysing interview data (Korkmaz and Överlien 2020; Överlien, Hellevik, and Korkmaz 2019), we showed that parents – both the youth’s and the partner’s – are present in youth’s narratives about IPV, and are a possible resource for help, depending on their responses (Korkmaz and Överlien 2020). In this study, correspondingly, we show that most respondents reside with their parents, and that the most commonly reported arenas of violence are ‘partner’s home’ and ‘youth’s own home’: arenas to which parents have access.

Moreover, it is important to highlight that schools also appear as an arena of violence. Previous scholarly work has underlined schools as important when it comes to safeguarding youth wellbeing through challenging the normalization of IPV, as well as addressing instigation and offering support to victims (Barter 2014; Storer, Casey, and Herrenkohl 2017). In our study, consistent with one previous qualitative study (Korkmaz and Överlien 2020), we show that violence recurrently takes place on school grounds. In sum, these findings underline youth IPV as a school problem, requiring attention from and preventive efforts by schools. The arenas ‘at a friend’s house’ and ‘at a club/pub’ show the lowest incidence rates.

The early North-American literature on adult IPV, such as the landmark study ‘Behind Closed Doors: Violence in the American Family’ by Straus, Gelles, and Steinmetz (1982), payed specific attention to the fact that IPV was a type of crime, or phenomenon, that took place in private, without any witnesses except for the husband and wife and their children (although the attention to the children was greatly underplayed in the early literature). As such, it was a social problem difficult to expose, and challenging for professionals to intervene in, unless the victim chose to disclose. Our study shows that youth IPV, in contrast, take place under the same roof as others. Although not in plain sight, as the door to the teenagers’ room, or to the school washroom, is most probably closed, adults with the obligation to protect young people from harm (such as parents and school staff) are close by. Hence, in contrast to adult IPV, the violence and abuse in teenagers’ lives take place, where others have the chance to detect it and intervene.

Comparing prevalence rates: the examples of Norway and Sweden

In the following, we will discuss our results in relation to the Norwegian part of the STIR study (Hellevik and Øverlien 2016). As pointed out, challenges emerge when comparing IPV prevalence rates. Nevertheless, since we used the same survey design, we find this discussion fruitful. Further, Sweden and Norway are, arguably, culturally similar countries in regard to questions about IPV. Both countries have a history of advocating for gender equality and addressing IPV as a social problem. In the Global Gender Gap report for 2020, which measures equality in the areas of economics, politics, education, and health, Norway finishes in second place while Sweden finishes close behind in fourth (World Economic Forum 2020).

Findings from the two studies show some similar tendencies. In regards to gender, in both samples, girls reported significantly higher rates of victimization compared to boys (Barter et al. 2017). Further, in both Norway and Sweden, psychological violence perpetrated online is the most prevalent type reported, followed by psychological violence face-to-face, sexual violence, and lastly physical violence as the least frequent type. The Swedish study, however, presents noticeably higher prevalence rates. This could be a result of several factors. First, sample variations may affect the findings. For example, the mean age in Norway was 15.2 years, in Sweden 17.2. In addition, in regard to violence perpetrated online, the evolution of the digital revolution may have influenced the results. In Norway, data collection was conducted in 2013–2014, in Sweden in 2017–2018. During this time gap of three years, digital tools and social media became even more widespread, as, for example, Snapchat became very popular among youth (Internetstiftelsen 2019). Thus, arguably, the digital arena for Swedish youth in 2017–2018 is different, compared to the one for Norwegian youth in 2013–2014. This may further support the notion that the digital revolution indeed has introduced new arenas for violence to be perpetrated in (Stonard et al. 2015; Hellevik 2019). Further, in Norway, the data was gathered before the #MeToo movement, while in Sweden, the data collection took place before and after the movement (in 2017–2018). We believe this deserves specific attention. The #MeToo movement involved many youth in Sweden,⁹ amplifying the voices of those who had experienced sexual violence. It is not possible to establish whether the #MeToo movement affected the incidence rates in Sweden – if it made Swedish youth more aware of acts defined as sexual violence compared to Norwegian youth – but this is plausible. However, in our Swedish study, statistical analyses do not suggest that date of participation (‘before’ and ‘after’#MeToo movement) affected incidence rates.

Limitations

This study builds on a growing body of evidence about youth IPV and contributes with important knowledge. In Sweden, it is the first large study that specifically focuses on violence in youth intimate relationships. Nonetheless, the findings should be interpreted in light of certain study limitations. We recruited schools in a specific county, one with socio-economic levels similar to the national averages. However, our sample is not a nationally representative sample and the findings should thereby be generalized with caution. Furthermore, our recruitment was dependent on the willingness of schools to participate. As such, not all schools in the county are included. Moreover, our sample only consists of youth who attend school; it excludes dropouts. However, we did get access to a school that offers programmes for youth who have not completed elementary school and thereby cannot begin high school.

Further, since we measured experiences of IPV by having the respondents indicate whether or not they had experienced specific behaviours, the situational setting was not addressed. This could have led to under-reporting if respondents felt their experiences are not included in the questionnaire, or over-reporting if respondents reported an experience, despite it not being violence, because it fit the item asked about (e.g., ‘play fighting’). Finally, in regard to bullying, the informants were simply asked, ‘Have you ever been bullied?’ Thus, we do not know what they considered to be bullying behaviour. Nevertheless, methodologically, it is a strength that we as researchers had the opportunity to distribute the questionnaire ourselves. This allowed us to answer questions about wording and meaning, as well as stress a broad conceptualization of intimate relationships, including different forms of such.

Conclusion

It is critical that youth IPV be considered as a social problem in Sweden, and not disregarded or minimized. Youth IPV deserves specific attention and interventions from service providers, researchers and policy makers as well as from the adults in youths’ lives. Importantly, this study shows that youth IPV does not take place ‘behind closed doors’, but in arenas in which adults dwell. This is valuable knowledge for all persons in young people’s networks, including social workers, since it offers a unique opportunity for the adult world to respond and help youth in the midst of their violent romantic intimate relationships, and also prevent violence from happening.

One of the advantages with social work practice, is the often daily interaction with people in need. When meeting young victims of IPV, it is important that social workers acknowledge and confirm the victimization that many times can be repeated and severe, and offer proper help and support.

Further, social workers, both practitioners and researchers, can play a pivotal role in raising awareness and calling for legislation and other measures that will make it possible to protect and support youth victims along the same lines as adults. This work will need to acknowledge that preventive efforts, as well as supportive interventions, must be developed on the grounds that boys and girls show different patterns of victimization, and that it is not only ‘at-risk’ youth who are prone to being subjected to IPV. Social work researchers must further investigate what support IPV-exposed youth are offered, and what the youth themselves wish for and find helpful, and thereby shift the focus to how this social problem is tackled.

Disclosure statement

No potential conflict of interest was reported by the authors.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by the The Swedish Crime Victim Authority.

Notes

1. There is a lack of common terminology for violence within youth intimate relationships. In North America, the term ‘dating violence’ is commonly used. In Europe it is used to some extent, but it has received some criticism as not applicable outside of North America. This article uses youth intimate partner violence to address and discuss conditions of youthhood, following the WHO definition of ‘youth’ as individuals age 15 to 24.

2. Following the Centre for Disease Control and Prevention (CDC), we define intimate partner violence (IPV) as: ‘physical violence, sexual violence, stalking and psychological aggression (including coercive tactics) by a current or former intimate partner (i.e. spouse, boyfriend/ girlfriend, dating partner, or ongoing sexual partner)’ (Breiding, Basile, and Smith et al. 2015, 11).

3. ‘Arenas’ refer to the physical places where violence occurs, such as home, school or a club.

4. For example, the notion of cultural factors have been addressed as potentially affecting what expressions of violence are culturally accepted or not, e.g. mild expressions of violence may be more tolerated and not defined as violence in some countries, but unacceptable in others (Viejo 2014). The societal view on punishments of violence perpetration etc. can also contribute to this notion, which may affect reported prevalence rates (Lysova and Douglas 2008)

5. The project is funded by the Swedish Crime Victim Compensation and Support Authority.

6. In Sweden, research participants who are at least 15 years of age, and who understand the research objectives, can give their informed consent without parental involvement (The Act concerning the Ethical Review of Research Involving Humans, SFS 2003:460). Since all informants were 15 years of age or older, no parents had to give their consent.

7. We would like to thank Dr Christine Barter, principal investigator in the STIR study.

8. Multicollinearity was controlled using Pearson’s r. The strongest correlation between independent variables in the models is between female gender and generally having older partners (r = 0.430). Following Tabachnick and Fidell (2001), the boundary for multicollinearity was set to about >0.700.

9. See, for example, https://www.svt.se/nyheter/inrikes/tystiklassen-nu-gor-eleverna-uppror-mot-sexuella-trakasserier-i-skolan