Adolescent Female Offenders’ Subjective Experiences of How Peers Influence Norm-Breaking Behavior

Delinquent peers have a strong influence on adolescent delinquent behavior. However, few studies have investigated adolescents’, and in particular young females’, own perspectives of the role of peers on their delinquent behavior. The purpose of the present study was to explore how young female offenders described their delinquent behavior and more specifically the role they assign to peer relations in committing or avoiding delinquent acts. Nine female adolescents, sentenced to youth service, were interviewed, and the data was analyzed using the Consensual Qualitative Research (CQR) method. The results showed that committing crimes and taking drugs with peers were portrayed as a way for the female delinquents to socialize. Delinquent and pro-social activities with peers appear to serve similar developmental functions in the sense that it is described to fulfill the same developmental needs. The young offenders also described collectively created pressures and norms in the peer group as the main contributing factor to their norm-breaking behavior, where they described being both recipients and producers of influence in the group. Another important finding was that the female offenders showed an awareness of the importance of pro-social peers and the need to eliminate delinquent friends from their peer network in order to help them refrain from deviant behavior. Implications for prevention and intervention are discussed.


Introduction
Historically, girls who committed crimes were either ignored or regarded as social isolates (Campbell, 1980). Early indepth-case analyses of incarcerated females explained young females' delinquency as a way to cope with their inability to form healthy connections with peers (Konopka, 1966). Similarly, Wattenberg (1956), and Reckless (1957) portrayed delinquent females as loners and misfits, incapable of forming friendships and female offenders were described to, for the most part, engage in delinquent acts by themselves. It has also been assumed that the peer group is of less relevance for female offending, based on the assumption that girls put more emphasis on family rather than friends (Campbell, 1990). More contemporary work indicate, however, that the connections between girls, delinquency and peers are more complex than previously assumed (Giordano, 2009). These studies demonstrate, in contrast to early notions, that delinquent females indeed participate in peer groups and that peer relationships are important for both delinquent and nondelinquent females (e.g., Cairns, Cairns, Neckerman, Gest, & Gariepy, 1988;Demuth, 2004;Giordano, Cernkovich, & Pugh, 1986;Haynie, 2001Haynie, , 2002Vásquez & Zimmerman, 2014). For example, it has been demonstrated that, similar to other adolescents, delinquent females establish meaningful, high-quality friendships and often identify at least one close friend whom they trust (Cairns & Cairns, 1994). Also, a recent study showed that delinquent and non-delinquent adolescent do not differ in how they report and perceive attachment to their peers (Vásquez & Zimmerman, 2014). Pleydon and Schner (2001) studied the quality of female delinquent friendships, by applying self-reported questionnaires, and found that delinquent adolescent females have relationships that are similar to those of non-delinquents, with regard to companionship, conflict, security, and closeness with their 1 3 best female friend. However, female delinquents were found to have less communication and to experience more peer pressure within their peer group compared to non-delinquent females (Pleydon & Schner, 2001). Solomon (2006), using data from a youth detention sample and a matched control group, reported similar findings where she suggested that friends in general seemed to serve the same emotional purpose in the lives of delinquent and non-delinquent females.
As these studies suggests, peer relationships are important for both delinquent and non-delinquent females, which is consistent with developmental literature, demonstrating the major role peers play in the psychological and social development of youth (Hartup, 1999;Vitaro, Boivin, & Bukowski, 2009). The importance of peers during adolescence has been stressed in several developmental theories, including Erikson's (1968) theory of identity development (Bosma & Kunnen, 2001). During adolescence, peer relationships are important sources of support, connection, and intimacy (Rubin, Bukowski, & Parker, 2006), as well as important for developing social skills and competencies such as interpersonal sensitivity, mutual respect, collaboration and perspective taking (Newcomb & Bagwell, 1996). In fact, winning the acceptance of peers could be regarded as one of the major developmental goals during adolescence, which means that peers are important for directing and constraining the adolescent's behavior, and increases the influence of peers (Steinberg, 2010). The elevated susceptibility to peer influence could also include being encouraged by peers to engage in criminal behavior, and due to the strong motivation for peer acceptance among adolescents, criminal behavior is also most likely to occur in the presence of peers (Launay & Dunbar, 2015). Deviant peers have repeatedly been identified as a crucial risk factor for engaging in delinquent acts. In particular, the impact of peers appears to play a key role in delinquent behavior that first appears during adolescence (Piquero, Diamond, Jennings, & Reingle, 2013). Having delinquent peers is a risk factor that both in theory and empirically has been shown to be relevant for juvenile delinquency (McGloin & Dipietro, 2013;Warr, 2002). How delinquent peers influence youth's delinquent behavior has been stressed in socialization theories such as social learning and differential association theory, where these suggest that individuals learn to be delinquent from others such as family members, friends and media figures (Akers, 1998(Akers, , 2009Matsueda, 2001). According to these theories, delinquent behavior is learned in three major ways: through exposure to criminal models whose behavior is imitated, through reinforcement of delinquency, and through teaching of beliefs that promote delinquency. Empirical findings support the connection between peers' delinquency and the youths own delinquent behavior, where deviant peers both expose youths to criminal activities as well as encourage such behavior (Brauer and De Coster, 2015;Farrington, Ttofi & Coid, 2009;Haynie, 2002;Warr, 2002;Wike, Miller, Winn, & Taylor, 2013;Worthen, 2012). Meta analyses of the predictors of juvenile delinquency have found that deviant peers are a major risk factor for adolescents (Wong, Slotboom, & Bijleveld, 2010), and a deviant peer group has been shown to be one of the strongest predictors of female delinquency (Hubbard & Pratt, 2002), even when parents, school and other interpersonal factors are controlled for (Pleydon & Schner, 2001).
Whilst increasingly more researchers have begun to focus on female delinquency (e.g., Chesney-Lind & Pasko, 2013;Miller, 2012;Zahn, 2009), the majority of available literature concerning pathways into juvenile offending, peers as risk factors and interventions is still based on juvenile delinquent males (e.g., Herrenkohl et al., 2000;Thornberry and Krohn 2005). The majority of research concerning delinquency have for the most part used a quantitative approach, even if more scholars are beginning to adapt qualitative approaches to better understand female delinquency (Garcia & Lane, 2013;Herrman & Silverstein, 2012;Miller, 2001;Pugh-Lilly, Neville, & Poulin, 2001). For example, Wike et al., (2013) interviewed eight court-involved adolescent female who were adjudicated but not incarcerated in the juvenile court system in the United States. They found that female offenders describe same-gender friendships as a conduit for their minor delinquent activity, and that these behaviors serve a recreational purpose, such as to relieve boredom or provide excitement. The females in the study also expressed a strong sense of personal agency and efficacy in making their own decisions, mostly denying being influenced by peers and partners to engage in delinquent acts. However, in depth analyses of the material revealed that the girls did in fact describe instances where they were influenced (Wike et al., 2013). Another qualitative study found that the gender of peers, from the young females' perspective, can have different influence on their delinquency (Solomon, 2006). Solomon (2006) reported how a sample of detained females expressed preferring male friends over females, with the explanation that male peers provided safety and protection while female friendships were described in more negative terms. Also, non-romantic male peers, as opposed to romantic partners, were described to play different roles in the lives of delinquent females (Solomon, 2006). As these qualitative studies provide important knowledge on how female offenders themselves experience and view their peer relationships in relation to their delinquency, most have sampled incarcerated adolescents (Solomon, 2006), girls in gangs (Miller, 2001), or "high risk" females in the secure care who display large behavioral problems including violent crimes, in addition to experiencing a range of co-currying risk factors such as abuse and drug addiction (Garcia & Lane, 2013;Herrman & Silverstein, 2012). Little attention has thus been given to convicted adolescent females with limited delinquency who are deemed to be without need of care, where the issue of minor offending is theorized as a part of the growing out of crime thesis (age-crime curve) (Shulman, Steinberg & Piquero, 2013;Wiecko, 2014), or part of Moffitt's (1993) adolescent limited offending approach (Bottrell, Armstrong, & France, 2010;Piquero et al., 2013). For example, studies focusing on community-based measures where female offenders, are not placed in custodial or residential settings, such as those sentenced to youth service in Sweden, are scarce. Youths sentenced to solely youth service (e.g., not in combination with youth care) are deemed to be without special care needs and extensive criminal history (National council, 2011). It can thus be presumed that young females being sentenced to youth service constitute a group of offenders with limited delinquency, which was the target group for our study. However, the knowledge about this specific group of offenders, in contrast to young females placed in residential settings (e.g., Andersson Vogel, 2017;Wångby-Lundh, Klingstedt, Bergman, & Ferrer-Wreder, 2017) is limited. It is important to further investigate this particular group of female offenders as they constitute the majority of females being sentenced for a crime, nevertheless we know very little about them, what they themselves consider important in regards to factors leading to their offending, and in particular how peers have influenced their offending. As this group of convicted females are deemed to be without need of care and thus not incarcerated or placed within residential care, it is the social services and schools that first and mostly come in contact with these young females. More knowledge of the needs of this particular group of convicted young offenders is critical in the development of effective interventions and treatment programs. This knowledge would provide a clearer picture of not only what types of interventions best fits the needs of this group, but also which institutions (social services and/ or schools, etc.) best can provide these interventions. As Resko et al., (2016) argue, it is important from an individual level perspective, to address what female adolescents perceive to be important to them for engaging in delinquency, as this may be a key part of efforts for interventions for female delinquents as well as to prevent delinquency among female adolescents in general.
The aim of the present study was to explore the subjective perspectives of a group of female adolescent offenders sentenced to youth service in order to get their own view on what role their peers play in connection to their delinquent behavior. This was done by analyzing their narratives about the character of their delinquency, what they considered had lead up to their delinquent behavior and their views of the future by focusing on how the females (1) describe pro-social and anti-social peer group interaction, (2) describe the role of peer-interaction in relation to their own delinquent behavior, (3) describe that peer relationships influence their delinquent behavior.

Method
The data used were collected from interviews with female adolescents sentenced to youth service in two major cities in Sweden. The general aim of the interviews was to explore the participants' way of thinking about the crimes they had committed and what they, from their own perspective, might need in order to avoid future criminality. Although the present study focuses its analysis on one aspect of the participants' narratives, peers, the participants' descriptions covered different life areas which were central in connection to their delinquent behavior. However, the role of peers appeared to play a fundamental role and emerged as a significant aspect from the participants' own perspective. The importance of peers in the life of adolescence in general and in particular in connection to adolescent delinquent behavior is theoretically and empirically supported. Also, there is evidence suggesting that peers seem to affect delinquency in a different way than other interpersonal relationships such as family (Brauer & De Coster, 2015), and based on these factors we considered that the domain of peers yielded further analysis on its own.

Participants
The sample included nine adolescent females who had been sentenced to youth service for committing a crime in 2012 and 2013 in two major cities in Sweden. At the time of the interview the participants ranged in age from 15 to 21, however all of them were 15-17 when convicted. Five participants reported living with single mothers or with their mother and her new partner, one lived with both her parents, one reported living with her boyfriend, one reported living with relatives, and one participant reported living alternately with her mother and father. Seven participants were in high school and two were working at the time of the interview.
The young females reported that they had participated in a range of antisocial behaviors such as running away from home, truancy, taking drugs, getting into fights, and bullying schoolmates. They also mentioned having committed crimes such as possession of drugs, battery, retail theft, property damage, vandalism, possession of a weapon and robbery. The felonies of which they had been convicted when recruited included crimes of violence for three of the participants and theft or shoplifting for the remaining five. One of the participants described being convicted of a crime which does not exists, suggesting that she did not fully comprehend or know the offence she was convicted of. 1 Five of the participants had been convicted of crimes prior to the current sentence. For the remaining four, this was their first sentence. All of the participants, except one, reported that they had committed criminal acts prior to the sentence without being convicted and three also described having committed delinquent acts after being sentenced.

Participant Recruitment
Youth service is intended as a penalty option for young people under the age of 21 and consists of unpaid work, and a so-called advocacy program, and intended as a rehabilitative penalty (Swedish Government, 2006). These "specially organized activities" are indented to meet the possible needs these young offenders may encounter, and are thought to be specified in connection with the youth service sanction. Since 2007, youth service has been an independent sanctioning form in Sweden and can be awarded alone or in combination with youth care. Youths sentenced to solely youth service are deemed to have limited "special care needs" and the penalty is the most common referral for delinquent youths (National council, 2011). In 2012 around 300 female adolescents were assigned to youth service in Sweden, of which an estimation is made that 15-30 girls were sentenced in the three major cities (Stockholm, Gothenburg, Malmö) (Brottsförebyggande rådet (Brå), 2012). Of these, all females enrolled in youth service programs during October 2012 and February 2013 in two major cities in Sweden, were given written information about the study through social workers, responsible for the implementation of sentence. The information letter contained a brief background of the study asking if the young female was interested in participating during the performance of their youth service. One person was considered by the social worker unfit to take part due to mental health issues and was therefore not asked to participate. The ones who showed interest in participating were contacted by the researchers to schedule a time and place for an interview. Three young females showed interest in participating by leaving their contact-information to the social workers but declined to participate when contacted by the researcher.
Before the interviews, the participants were given information making it clear that research participation was voluntary and independent of their participation in the youth service or any other contact with the social services, and that data would be treated strictly confidentially and not passed on to the police or the social services. They were also told that they were allowed to discontinue their participation at any time during the interview and that they were free to decline to answer questions they felt uncomfortable answering. Written consent was obtained from each participant prior to the interview. In accordance to Act concerning the Ethical Review of Research Involving Humans (2003:460), youths in Sweden between 15 and 17 who have had adequate information about the research project and an awareness of possible consequences of participating, and are assessed to have an appropriate level of maturity and capacity for insight, can give consent to research without the parent's authorization. On the basis that parental content thus is not required for youth who have attained the age of 15 in Sweden, the participants' guardians were not asked for consent. The present study was approved by the Stockholm Regional Ethics Board.

Interviews
A semi-structured interview protocol was developed based on a literature review covering risk factors and intervention and prevention programs relating to female delinquency. It consisted of open-ended questions, which were generated and discussed by the first and second authors of the present study. The interview guide addressed four central topics: (1) the participants' own views of their delinquency, (2) their views of their sentence, (3) their views of the what lead up to their delinquent behavior and (4) their views of the future. Each interview began by asking the participant to tell some background details about themselves (such as age, where they are from, their living situation and so on). Next, the participants' attitudes and beliefs about their delinquent behavior were covered, as well as their views of what lead them to their delinquent behavior and their views of the future. Asking them if they had any questions or if they wanted to add something concluded the interviews. Although the interviewer had these topics to cover, interviews were conducted in a manner that was sensitive to the interviewee and her narrative, allowing for prompts and exploration of topics that rose spontaneously. No specific questions regarding peers were part of the interview guide, however, since this topic spontaneously rose in all of the interviews, follow up questions regarding peers were asked to all of the participants, although they slightly differed. One pilot interview was conducted with the first participant recruited, which allowed the interviewer to test and become comfortable with the interview protocol. No major revisions were made and the pilot interview was therefore included in the analysis. Interviews took place in private rooms provided by the social workers or private rooms in local libraries. The first author conducted all interviews with the participants, which lasted about 40-70 min each. The interviews were audio recorded and transcribed verbatim by the researcher who conducted the interviews. The transcription included contextual information such as silence and pauses in conversation as well as nonverbal sounds (e.g., laughs, sighs) but did not include nonlinguistic observations (e.g., facial expressions, body language and emotions). All of the participants' names and personal information were omitted and/or changed in the transcriptions as well as in the write-up of the study. For their participation, the participants were offered a gift voucher of 200 Swedish kronor.

Coding and Analysis Process
The data were analyzed using the Consensual Qualitative Research (CQR) method (Hill, Thompson, & Williams, 1997;Hill et al., 2005). Fundamental features of this qualitative methodology include use of several judges throughout the analysis processes to ensure multiple viewpoints as well as using a minimum of one auditor to review the work of the judges to reduce the effects of groupthink. It is also a well-suited and commonly used method involving applied research within psychology and psychotherapy research (Hill, 2012).

The Research Team
The primary team of researchers consisted of five people, three judges and two auditors (one external and one internal). All of the researchers included had experience of working with similar research topics in research and practice. All of the members had read relevant literature on CQR (Hill, 2012;Hill et al., 1997Hill et al., , 2005 prior to taking part in this study, as well as exemplar studies (Diamond et al., 2011;Sander et al., 2010). The interviewer had prior experience and training in conducting general psychological interviews.

The Analysis Process
Preliminary analyses of the material identified distinct domains covering life areas that the participants themselves expressed was of relevance to their delinquent behavior. For the purpose of this study, domains concerning peers were selected for further analysis.
The coding procedure consisted of three steps (Hill, 2012;Hill et al., 1997Hill et al., , 2005. The first two steps of the coding process, coding of domains and abstracting core ideas, included all of the domains, while the last step of the coding process, the cross-analysis, was performed solely on the domain peers that was the main focus of the present study.

Coding of Domains
Each team member began by independently reading through three interviews and categorized all the material that related to the same topic areas-be it a phrase, several sentences, or a paragraph-into an appropriate domain. The team then convened to discuss suggested domains, and reach consensus about the placement of text and the naming of each domain. Once a final list containing 10 different domains had been agreed upon through consensus, the research team went back and individually coded the remaining interviews accordingly (Hill et al., 2005). Afterwards, the research team met yet again and resolved discrepancies about the placement of text into different domains through consensus. No new domains emerged when coding the rest of the interviews. The first auditing took place after three interviews had been coded into domains. The auditors' role during this part was to check whether the raw material was in the correct domain and to make sure that no important domain was overlooked. Auditors' input was considered and discussed by the team members and no major changes were made based on the comments given.

Abstracting Core Ideas
Creating core ideas meant summarizing the content of each domain with the intent to capture the core idea in as few words as possible and with enhanced clarity, while at the same time remaining as close as possible to the text (Hill, 2012). Each judge read one interview independently to construct core ideas for every domain. The research team then discussed and came to consensus about content and the exact wording of the core idea. Once a common understanding of the core idea process had been achieved, one of the team members wrote the core ideas for the remaining interviews and the rest of the team reviewed them and met to reach consensus. The core ideas for all interviews were sent to the auditors for review and the judges met to discuss the auditors' feedback. The feedback from the auditors included comments on parts of statements that were not reflected in the core ideas or regarding material that was unintentionally omitted from the core ideas. Appropriate changes were made before moving to cross-analysis of the data.

Cross-Analysis and Identifying Themes
The final step in the coding process was the cross-analysis, which included identifying common themes across responses within the domain of peers (Hill, 2012). The researchers independently reviewed the core ideas within the domain and suggested potential themes. The group subsequently met to compare themes and reach agreement through consensus regarding which themes best represented the data and how to name them. The third and final time the material was sent to the auditors was after the cross-analysis. For the cross-analysis they were constructed to see if the themes constructed faithfully represents the data. The auditors' suggestions included thoughts about the level of abstraction in the themes created as well as segments that should or should not be included. The comments from the auditors were all considered and discussed by the coders and incorporated if agreed on by consensus.

Results
Five themes within the domain of peers resulted from the cross-analysis (see Fig. 1). Examples of theme contents were included in the results in the form of quotes using the females' own words (for more example of quotes see Table 1).

Considering Peers as a Fundamental Part of Life
The participants described both negative and positive aspects of their peer relationships where their peers and romantic partners were described as fundamental part of their lives. They all described having close, meaningful relationships with peers, where some described their friendships as important to them as their families and others described them as vital for their wellbeing. For example, one participant described the importance of her peers in the following way: If I didn't have friends I wouldn't be able to make it through the day.. if I were to be alone I don't know it just doesn't work that way for me… I don't know how it would be to just sit alone at a table feeling all alone.. no that would never work.
The participants talked about spending all their free time with their peers and some considered them as close as siblings. The young females also described their peers as important people in their lives who they could confide in. They described how much having someone close to talk to in times of need meant to them and they described how they could talk to their friends and boyfriends about difficult things such as fighting with their parents or having a hard time in school.
The importance of the peer group was also shown in how the participants described that they put their peers' needs before their own and how they went above and beyond to protect these friendships. The following statement from one participant illustrates how she put herself in jeopardy in order to protect her friends: If I see any of my girls get hit of course I´m gonna get involved. I can't care about how bad the situation is, I still get involved. I don't care if it ends up with somebody getting stabbed, they're still my girlfriends and I want somebody … I want to be able to protect them.
The participants also described negative aspects of their peer relationships such as having conflicts with peers as well as romantic partners. This could entail fighting, bickering Fig. 1 The figure illustrates all five themes within the domain of Peers. The first theme "Considering peers as fundamental part of life" has two arrows pointing at the direction of delinquent as well as pro-social peers, illustrating that the participants described having peers who both engage in delinquent behavior and who don't, and the importance of peers, whereas delinquent or not. The three different themes, "Committing crime as a way to socialize", "Collectively creating peer pressures and delinquent norms" and "Being both offender and victim in a delinquent context" illustrates that the participants describe being in a delinquent context, where the norm-breaking behavior is described as a way to socialize and that this behavior often occur in the unstructured hanging out with delinquent peers, where they collectively create peer pressures and delinquent norms of how to behave. The arrow which includes the theme "Being aware of the importance of pro-social peers and absence of delinquent peers in order to avoid future delinquency" illustrates the relationship between delinquent peers and pro-social peers and how the absence of hanging out with delinquent peers is considered necessary to avoid delinquent behavior and that pro-social peers can help in that process and having verbal arguments, as well as being mad at each other for keeping secrets, or engaging in physical fights. For some, conflicts with peers went as far as resulting in the ending of the friendship. For example, one participant described how she stopped being friends with her former best friend after she found out that the friend had taken drugs behind her back as well as slept with her boyfriend at the time.
Feelings of distrust and/or loneliness relative to peers were also expressed by the participants. In some cases these two feelings were connected while in other cases the participants felt one or the other. One of the participants expressed her feelings of loneliness by saying: I have a best friend but we live pretty far away from each other and I work and she works so our lives totally crashes so we have never any time to see each other and it's really hard but that's the way it is and I know that if I ever need her I can always call her and even though we don't see each other that often it feels like just yesterday when we see each other.
Feeling of loneliness could also include not talking to friends about private matters, or feeling betrayed by your friends. For example, one participant described how she felt abandoned by all her friends during a period where she did not feel well psychologically and emotionally, and she expressed how her friends withdrew from her and did not make an effort to express concern about her wellbeing.

Committing Crime as a Way to Socialize
Another theme was how committing crimes was described as a way to socialize, like any other type of activities that they did with their peers. This is illustrated in one of the participants' description of how a typical day hanging out with her peers could look like: You meet up and hang out drinking beer and smoking, maybe taking something else [referring to drugs], and then you go home and continue hanging out, and maybe the next day you barbeque and stuff like that.
Another participant painted a similar picture which entailed hanging out with friends whilst committing crimes: Like me and my girls who live close by we live in the same neighborhood we used to go down to the gro- . even if we hung out for several years like every day morning til' night I mean imagine not getting tired of a person it's so weird" "Some people eh some of them meant so much to me I loved them like they were my siblings" 2. Committing crime as a way to socialize "Mm yes ehm we weren't together before but we had decided to do it the day before and it was just me and my friend and then we met up and went to a mall and then we went to another mall together we went to the first store and that's how it started" "We talked to each other like what did you take and tried to make it into a fun moment even if we were feeling bad we just laughed and laughed and laughed" 3. Collectively creating peer pressures and delinquent norms "Yeah for example the can tell me to run away from home I mean it's my choice that I do it but I get these weird ideas like yeah let's do this" "We were about six people and if me and my friend did something like that the others would do it as well thinking we wanna try it as well" 4. Being both offender and victim in a delinquent context "I don't know but as you got older and it became more obvious than before when you were younger .. before it wasn't even on my mind but now it was all around me. And then I thought I might as well be a part of that" "And then one of the guys walked up to her and hit her cap over so I walked up to him and like what are you doing .. and that's when I got a kick in the stomach I think that's where he kicked me first and I asked what the hell are you doing and he kicked me again and I flew back and I was all blue here.. it hurt so bad I thought I'm never going back to town again" 5. Being aware of the importance of pro-social peers and the absence of delinquent peers in order to avoid future delinquency "Mm they'll be in touch now and then but I'll tell them I don't want anything to do with you unfortunately that's the way it is you know some of them are still using and I don't tolerate that I can't hang out with people who take drugs" "Ehm so I've changed my friends and got new hobbies I had no interest in anything when I was just hanging out in town now I keep up in school and think it's fun to have started working and stuff like that" cery store and just take chocolate bars and hide in the bushes and just sit there and eat them.
When the participants described what they did when they spent time together with peers they mentioned different forms of unstructured activities such as being at each other's houses, hanging out in the mall, going to the movies, watching movies and TV, cooking, playing games on the computer, talking, barbecuing, playing video-games, playing music and dancing as well as engaging in norm-breaking behavior such as drinking alcohol, smoking cigarettes, taking drugs and committing crimes.
When describing their delinquency they mentioned behavior such as violent crimes, for example hitting and/or kicking others, as well as shoplifting, driving a stolen car, vandalism and arson. The participants gave different reasons as to why their delinquency took place in the company of others. One of the participants described for example the reinforcing qualities of committing crimes in the company of peers and talked about feeling more secure in the company of others: This theft that happened last week I did that with a friend as well because I wouldn't go by myself and shoplift things alone.. no it feels safer, it's like security to have someone there with you.. if I was alone.. no I don't think I would have the courage to do it.
Others talked about how the criminal activities gave a sense of belonging and comradeship, but were also described as fun, thrill-seeking escapades to do together.

Collectively Creating Peer Pressures and Delinquent Norms
One theme included descriptions of how the participants both influenced and were influenced by peers when it came to their delinquency. Most of the female offenders described the individuals in the peer groups as equals and said that everyone together helped to create peer pressure against everyone in the group, as articulated by one of the participants: Maybe we do stupid things together but it's not like anyone says hey let's go do stupid things. I don't know; it's just like if one of us does something, let's say buys something or not buys but steals for example a candy bar and says to the others I'm not gonna share this with you then automatically, if the others also want candy, they have to steal it too.
Another participant described how it was not always clear who was pressuring whom and that they could influence each other to do things that perhaps no one really wanted to do in the first place: When I was convicted of vandalism then it was like first I don't know if it was my or her idea; it was just I triggered her and she triggered me, so really no one wanted to do it but it still happened.
For others, the delinquent behavior was a direct influence of seeing their friends engaging in a certain behavior. One participant explained it by saying: Without them I would never have taken drugs because I wasn't going around thinking that I was going to try pot one day. It was just that it was right there in front of me and when I saw my friends doing it I thought I might as well try it and take a hit or two.
Some of the participants emphasized that they had let friends or romantic partners decide for them, while others talked about the great influence their friends had over their own behavior. For example, one participant talked about how she had let her boyfriend decide for her, and how she started taking drugs once he started: When we were together I mean I was so in love with him that I sort of let him decide everything. I always said yes because I was so in love. I just wanted to say yes and let him have his way so he could feel good.

Being Both Offender and Victim in a Delinquent Context
One theme included descriptions of how the participants were part of a delinquent context where they witnessed and observed peers and same aged companions engage in norm-breaking behavior, as well as being both victims and offenders of delinquent acts against other peers themselves. This is illustrated by one participant who described it in the following way: I have friends and acquaintances people who went to my school who I heard did things everything from drugs to like shop lifting and robbery all sorts of things but I was like I was never like that.
Being part of a delinquent context entailed having peers, partners, classmates or acquaintances that engaged in different forms of norm-breaking behavior. Some participants had close friends that had committed and/or had been convicted of a crime. Others had friends that took drugs, ran away from home or who displayed violent behavior. Some of the participants lived in areas where youths the same age engaged in delinquent behavior.
Others had current or former boyfriends with more comprehensive and more serious antisocial behavior. These behaviors were usually drug-related, where some of the participants' boyfriends either sold or used drugs, but also included delinquent behavior such as driving a car while under the influence of drugs, and running away from the police.
Descriptions of committing delinquent acts against other peers included stealing from friends, where one participant for example described an incident where she stole another girl's jacket at a house party. It also included acts such as hitting and fighting with peers, where for example one participant described getting into physical fights with several schoolmates.
In addition to being part of a social context in which delinquency occurred and they themselves committed crimes against other friends, the participants also described being victims of some sort of serious abuse, by another person the same age, a friend or an acquaintance. These experiences included physical abuse, being drugged, or raped, as was the case for one of the girls who described the following incident: When I was in high school I was at a friend's party or her brother had a party and everybody got really drunk, or not really drunk but everybody was drinking, and I woke up in the middle of the night and her brother was carrying me away so .. I was raped by her brother.

Being Aware of the Importance of Pro-social Peers and the Absence of Delinquent Peers in Order to Avoid Future Delinquency
In order to avoid future delinquent behavior, the participants talked about the importance of pro-social peers as well as the need to break up with delinquent friends. One participant expressed it by saying "in order to change old habits one needs to quit old acquaintances". The young females said that avoiding delinquent peers, and not letting them influence their own behavior, was one of the main factors that would help them avoid engaging in further delinquent behavior.
Having structured activities to take part in or having a job to go to instead of simply hanging out with delinquent friends was also described as helping them refrain from delinquent behavior. One participant said that the key to staying away from delinquent behavior was to "hang out with the right people who don't do stupid weird stuff you know, and make decisions on my own and not let others decide for me-that's when things go wrong". Others said that they had learned a lesson by hanging out with former delinquent peers, or had the conviction to never socialize with delinquent peers again. For example, one participant said: "I feel that I will never hang out with people like that again. If I could choose I would go back in time and undo it." Some had reevaluated their friendships with deviant peers and concluded that they had a bad influence on them and had ended their friendship, as expressed by one of the participants: I don't know, I just don't want them in my life. I didn't feel they were good for me, they had other thoughts and values and morals than me… to them it seemed like life was all about getting as high as possible and those were not my beliefs anymore since I quit all of that and I didn't wanna sit around watching them throw their life away by taken drugs so I erased them from my life.
All of the participants described at least one, and typically several friends, who did not commit crimes. These peers were usually old friends that had been retained and were available alongside newly acquired delinquent friends. For example, one participant mentioned friends that did not take drugs or engage in delinquency, who had always stood by her side, and described how important these friendships had been to her. Having meaningful and important relationships with peers outside the context of their delinquent behavior were described as important for different reasons, one being that they helped the young offenders to avoid future deviant behavior. For some participants this help meant having non-delinquent friends who rejected or disapproved of their criminal behavior. For example, one girl described how her non-delinquent friends distanced themselves from her when she started doing drugs with her new friends, and how they took her back once the drug use had stopped.

Discussion
The purpose of the present study was to explore female adolescent offenders, sentenced to youth service, own perspective of what role their peers play in connection to their delinquent behavior. The results can be summarized in several main findings. One main finding was that committing crime with friends was portrayed as a way to socialize. Taking part in norm-breaking behavior was described to have a lot in common with pro-social activities typical of adolescents in general. For example, committing crimes in the company of friends was described as contributing to a positive sense of belonging, and led to sharing experiences. Although not focusing on peer relationships per se, Bottrell et al., (2010) reported similar findings, where delinquent escapades were described by young people, like other social recreations typical of adolescents, such as listening to music, cooking, and watching movies, to have the same meaningful qualities of social bonding and connection. Beyond these factors, committing crimes in the company of peers was also described as giving the participants more confidence and courage to do things that they might have never done alone. This is consistent with findings reported by Miller et al., (2012) reported similar findings where delinquent female adolescents spoke about the reinforcing qualities and affiliations of delinquent behavior and peers (e.g., thrills, protection, sense of camaraderie). Presumably, the delinquent group provides opportunities for socialization as well as makes it easier to excuse, justify and avoid responsibility for delinquent behavior (McGloin & DiPietro, 2013;Miller et al., 2012).
Another main finding was how the participants' feelings of being influenced by friends and boyfriends to commit delinquent acts was the result of group dynamics, where the group members collectively created peer pressures and delinquent norms that serve as a frame of reference for how to behave in that particular group. Even if some of the participants afterwards felt that they had let others decide for them, or had done things they did not really want to do, no one explicitly described being forced to behave or act in a certain way. This fits well with Brown's (2004) definitions of peer pressure as typical for most groups of adolescents. It is thus important not to mistake influence with coercion, as the present findings suggest that peer influence is not always a one-directional process but rather a reciprocal group process in which the participants are both recipients and producers of influence in the group at the same time.
Also, the young females showed an awareness of the influence delinquent peers had on their own delinquent behavior. When asked what was required for them to avoid recidivism, most of them pointed to the need to stop hanging out with friends who committed crimes. In order to do so, having access to pro-social peers was described to be crucial. For example, having non-delinquent peers who rejected their delinquent behavior, or making them choose between the two types of friends, was described to have a pro-social influence. Having non-delinquent friends can thus for the participants contribute to avoiding the kind of encapsulation effect Haynie (2002) describes in interaction with solely antisocial peers.
In line with other types of research studying the connection between peers and delinquency (Pleydon & Schner, 2001;Solomon, 2006), the results also show that interviewees seemed to have similar relationships with their friends as female adolescents who do not engage in delinquent behavior. The participants described having friends who engage in delinquent behavior as well as friends who engage in prosocial behavior, whom they were close to. The closeness and importance the girls attributed to their peers, whether delinquent or not, could be interpreted as a way to break free from parents, in order to become more autonomous. This indicates that the females' relationships with peers were consistent with typical development during adolescence. Age-wise, the participants were in a period between childhood and adulthood where they in different ways were trying to find themselves and discover what they wanted to do with their lives, and their friendship groups can be seen as an arena in which to explore identity concerns. The negative aspects of their friendships, with some describing these relationships as superficial, false or complicated, could also be interpreted as typical for adolescent relationships. Having conflicts and experiencing failure in peer relationships is a part of normal development in adolescence as friendship relations during adolescence are generally unstable (Brown, 2004). Relationships with delinquent peers have been shown to be characterized by less communication and more peer pressure (Pleydon & Schner, 2001), and some researchers have suggested that delinquent female adolescents have more trouble than female adolescents in general maintaining friendships with other females due to generalized distrust Solomon, 2006). However, one could interpret the present findings about the participants' experience of loneliness and betrayal, not as indicating a lack of ability to create genuine relationships, but rather as showing that it is peer pressure and the delinquent behavior itself that complicates these relationships.
Finally, the participants described being part of a delinquent context where they were both offenders and victims of different types of delinquent acts by peers, with their victimization mostly occurring in parallel with their own delinquent behavior. However, no one explicitly said that her victimization had led to, or caused, her delinquent behavior. Self-report studies have shown that youths who commit crimes are more likely to be victims of crime committed against them (Wikstrom & Butterworth, 2006), which is in line with our results. Although it is well established that female victimization and peer pressure are two important risk factors for understanding female delinquent behavior (Chesney-Lind, 1989;Giordano, Longmore & Manning, 2006), portraying adolescent female offenders as sole victims of deviant peers and romantic partners, or describing their delinquent behavior as a direct cause of their victimization, is to oversimplify a more complicated process.

Implications and Future Directions
Previous, mostly quantitative research, has repeatedly shown the connection between peers and delinquency. The present findings are important as they indicate that the participants have an awareness of the role of peers for their delinquent behavior themselves and can articulate these thoughts quite insightfully. These results point to the importance for practitioners to ask young delinquent females themselves what they need in order to avoid recidivism. For example, the participants expressed that their friendships with peers, whether delinquent or not, have great meaning and serve a positive function for them. Their friendships with peers who engage in delinquency were not described as solely destructive, and fulfilled many developmental needs. In the eyes of adults, the influence of delinquent peers can be seen as something merely negative. However, it is of great importance, not least from a practitioner's viewpoint, to recognize the positive feelings the participants attribute to their peer groups, delinquent or not, and to provide other social contexts that fulfill the same type of needs. In line with previous findings, the present result suggests that one way of reducing delinquency among female adolescents is to replace those behaviors with other more prosocial behavior that meet adolescents' need for belonging, intimacy and autonomy (Bottrell et al., 2010;Farrell, Thompson, Mehari, 2016). It is also important for practitioners to recognize that female delinquency occurs within a peer context, and is amplified by group dynamics that unfold within the young females' circle of friends, where they can be both recipients and producers of influence in the peer group. To gain a more comprehensive understanding of how these group dynamics manifest, in the context of delinquent behavior, future studies are challenged with finding different ways to more closely study these group dynamics.
As the participants themselves show an awareness of the impact of peers' delinquent behavior on their own delinquent behavior, and also describe their pro-social peers as a factor that had helped them refrain from deviant behavior, it is also important to encourage friendships with non-delinquent friends. Integration of positive peer relationships within intervention efforts has been shown to divert youth from problem behaviors, where for example studies have found that peers' prosocial behavior decreases delinquent attitudes (Carson, 2013) as well as promotes prosocial behavior (Barry & Wentzel, 2006). Since all of the participants described having access to peers who do not engage in delinquency and recon that these relationships were a source of support in avoiding future delinquency, practitioners could encourage them to move closer to these friends, as this is more realistic and palatable, than to develop entirely new network of same aged companions.
It is thus important that prevention as well as intervention efforts target enhance prosocial influences rather than solely focusing on reducing deviant peer influences. This is consistent with broader efforts to promote positive development that attempt to not only reduce risk factors but also increase promotive factors (Catalano, Hawkins, Berglund, Pollard, & Arthur, 2002). Also of importance is that the participants described how they themselves as well as their peers both engaged in delinquent behavior as well as in prosocial behavior. These results suggest, as has been argued by Farrell et al., (2016), that pro-social and delinquent behaviors are not exchangeable terms in the life of delinquent females, but rather two extremes on a continuum. This has implications for future research within this field suggesting that describing adolescents' behavior (e.g., delinquent behavior) may be more appropriate than labeling them themselves as delinquents (Farrell et al., 2016).
The participants described how most of their free time was spent in unstructured activities with peers in the absence of authority figures. Osgood, Wilson, O'Malley, Bachman, & Johnston, (1996) have argued that informal socializing with peers can increase the risk of offending. Accordingly, the presence of peers makes delinquent acts more easy, fun and rewarding, while the absence of authority figures reduces the opportunity for social control responses to delinquency (Hoeben, Meldrum, & Young, 2016). This is consistent with empirical findings that show a negative relationship between structured activities and antisocial behavior (Mahoney & Stattin, 2000), as well as pro-social effects of structured "after school" activities in girls (Hansen, Larson, & Dworkin, 2003). The participants said that having a structured activity to take part in would give them less time to spend with deviant peers, thus making it easier to avoid committing delinquent acts. Some participants also expressed a wish to have a job, so they could earn their own money and not be financially dependent on their parents. Molnar et al., (2005) report similar findings in their study in which delinquent girls themselves viewed extra-curricular activities as beneficial in many ways, especially for avoiding future delinquent behavior. Bottrell et al., (2010) results also show that lack of facilities and organized leisure activities is more likely to get adolescents in trouble. Helping and guiding delinquent females into extracurricular activities, further study, or a work place may in fact fulfill those needs as well as provide the young females with pro-social peers, encouragement, guidance and an entrée into constructive pursuits. One arena where after school activities and organized leisure activities best can be provided are for example through the schools. The school system can help these young offenders find other social contexts to take part in that fulfill the need for belonging, intimacy and autonomy as well as encourage socializing with pro-social peers. Engaging young offenders to structured community resources has been shown to foster positive peer affiliation and prosocial behavior, thus not only reducing delinquency, but also promote positive development for youths in general (Smith, Faulk, & Sizer, 2016). Interventions aimed at helping the young females develop strategies for becoming more academically successful or guidance into further studies are also recommended. Finally, as the participants wish to eventually start working and getting into the labor market, it seems important to adjust the unpaid work part of youth service to not only serve as a way of punishment but also to serve as a way for the young females to get work experience and being an introduction into working life.

Limitations and Conclusions
There are some limitations to this study that need to be addressed. Recruiting participants was difficult, which has lead to the small convenience sample used. In addition, there could be a systematic difference between delinquent females who chose to participate in the study and those who refused. As this is a sample of girls who are presumed to have limited "special care needs" (National council, 2011) it can be presumed that the participants in the present study represent the healthier end of the spectrum. Presumably, the results may not be of relevance for female offenders with more severe problem behavior. Also, it should be noted that one female enrolled in the youth service program was excluded from participating as she was considered, by the social worker in charge, unfit to take part due to mental health issues. The fact that these were girls living in two larger cities could also be reflected in the results. Sampling participants from smaller towns or female offenders living on the countryside could have yielded different results. As the interview questions did not specifically focus on peers but rather asked more generally about the participants' views of what had lead them to their delinquent behavior, questions relating specifically to peers were not asked unless in forms of follow up questions relating to topics raised by the participants themselves. This means that some aspects of this subject could not in depth be explored nor analyzed. However, one can also view the spontaneous focus on peers without any directions from the interviewer as a strength in that it shows that peers are a major part of their delinquent behavior and what they themselves consider important when freely describing their delinquent behavior. Finally, one need to consider if the participants could have had different motives to present themselves in more positive terms by for example undermining their delinquency or their own agency in the decisions leading up to their delinquent behavior, and how this might have affected the results.
The adolescent female offender's own view of their peer relationships and delinquent behavior offered new perspectives on the subject. It also adds to the litterateur by showing the group nature of female delinquency, in that most of the participants' delinquent behavior occurred in the company of friends, and delinquency was described as a way for the young offenders to socialize with peers. Committing crime with peers was described to have much in common with other pro-social activities in the sense that it fulfilled the same developmental needs. In addition, the participants talked about group dynamics, where they were both recipients and contributors of influence, as the main contributing factor leading them to commit delinquent acts. Finally, the female offenders were well aware of what they needed to avoid further recidivism, which was to exclude delinquent peers from their networks as well as having more structured activities to take part in. These results points to the importance of asking young female offenders about their own views of their delinquency, as their perspectives provide insight into how intervention efforts best can be fitted to meet the specific needs expressed by the young offenders.