Patterns of Aggressive Behaviour and Victimization in Adolescents

The objective of the presented research was to find the family determinants for undertaking the aggressor or victim role. The obtained results enabled the description of environmental (family-related) and developmental factors that have a bearing on the formation of perpetrator or victim identity. For that purpose, two groups of variables were identified. The first group included child-independent variables shaping the socioeconomic status of the family (parents' education, material status, number of siblings), while the second group pertained to the patterns of attachment to each parent. The sample consisted of 120 adolescents aged 13 to 20. The research tools were Mini – DIA, the Inventory of Parent and Peer Attachment – IPPA, and Buss-Perry aggression questionnaire. The results revealed a number of determinants for persons involved in perpetration or victimization, such as the type of relationship with parents (secure or insecure pattern), personal experience of being in the victim or aggressor role, and the level of hostility. The resulting " determinant bundles " may inform professionals in their work with adolescents in the field of prevention or therapy.


INTRODUCTION
Adolescence is a stage of life that is particularly important in human development, mainly because of intense physical, emotional and social changes that occur in the functioning of a young person.The adolescent phase is characterized by rapid cognitive and emotional growth and the resulting increase of risk behaviour [1].Researchers pointed to the fact that this is the period of experimenting with psychoactive substances, sexual activity, aggression, exceeding social norms and breaking principles and other behaviour disorders [2].
In modern literature, assuming the role of aggressor and/or victim is treated as a readiness for perpetration or victimization in interpersonal contacts.That readiness consisted of an emotional-moral component and constant temperamental features containing such elements as: irritability, lack of adequate emotion control, strength of emotional arousal, ability to control one's emotions shaped in the course of socialization, habits, scripts and behaviour patterns and tasks resulting from the performed social role [3].
Many teenagers enter the adolescence period with relatively serious problems that affect the way in which they function.Still, many of them begin to experience family and peer-related problems only when they reach the adolescent stage in life.The change of the nature of the relationship with parents and peer group is connected with strengthening of aggressive behaviour or the victimization experience [4,5].That behaviour may assume various forms, from *Address correspondence to this author at the University of Zielona Góra, Poland; E-mail: marzanna@farnicka.plverbal and psychological aggression to physical and indirect violence.Many youngsters engage in overt violence against peers, such as calling names, kicking, bullying or pushing.Others use the so-called relational aggression, characterized by the abuse of contact with another person with the motive to inflict harm.
Research related to the evaluation of aggressive behaviour and victimization experience indicates that victimization coefficients are about twice as high as aggression coefficients [6].At the same time, behaviour patterns tend to change in time.Longitudinal studies show that aggression and victimization tend to increase in early adolescence and gradually decreases towards the end of the life period [7,8].Social psychologists explain that regularity as experimenting with roles connected with belonging to various peer groups and increasing the strength of peer influence [8].From the point of view of developmental psychology, these changes are attributed to the normative identity crisis and specific instability of behaviour [1].Factors connected with the origin and persistence of aggressive behaviour in children, adolescents and young adults are also sought in disturbances of developmental processes in the biological, psychological and social sphere [2,9].Relevant literature points to the importance of risk factors and protection factors from the field of genetics, hormonal processes, intellectual functioning and social environment of children and adolescents that are connected with aggressive behaviour.At the same time, other significant variables are self-esteem, level of self-control, identity style, sex, family situation, and manner in which parental roles are performed [10].Moreover, the role of the family system, its structure, life situation, emotional climate, family conflicts, experience of violence and upbringing style have also been recognized as determinants of aggressive behaviour in children and adolescents [11,12].
Research devoted to determinants of child and adolescent aggression increasingly emphasizes the role of attachment as a factor shaping or inhibiting young people's readiness for aggression.This trend is mainly due to studies from the stream of developmental psychopathology.Attachment patterns developed in a child-parent/carer relationship contribute to individual differences and determine the manner of organization of the child's behaviour towards the attachment figure in the first years of life.The patterns in turn decide the way the child will function outside the relationship with parent/carer at the later stages of life.Thus, one may conclude that they become precursors of later social relations of an individual [13,14].
The results of analyses related to the impact of attachment on the individual's further development and adaptation indicate that attachment type can be connected with the occurrence of psychopathological symptoms in children.And thus, children with anxious or disorganized style of attachment to their mothers are more likely to display disturbances in emotion regulation and are more exposed to problem externalization and development of aggressive behaviour [15][16][17][18][19].Although most analyses focus on the mother-child relationship and child behaviour in the first years of life, attachment does play a significant role also later in life [20,21].As development progresses, attachment patterns are internalized and shape the attribution of self, parents and the external world as unconscious cognitiveaffective constructs.The quality of early care influences the way in which the child later perceives his/her own person and the loved ones, the way in which he or she interprets life experience and the choice of stress-relieving strategies.Research on childhood reveals increased aggressiveness in children with anxious attachment pattern, both in lowincome families [22] and in medium-income ones [23].In later studies, analyses were made of the relations between non-observance of norms and aggression in childhood and antisocial behaviour in adolescence, although only some of the studies explore their connections with attachment.Numerous contemporary studies on the subject of attachment in adolescents focus more on the current quality of their relationship with parents than on the inner attachment models developed in childhood.For instance, Simons, Paternite and Shore [24] have found that the quality of relationship with mother as perceived by teenagers is connected with their evaluation of their own aggressiveness.Other research studies utilize the concept of attachment in relations with parents for the evaluation of their impact on shaping other intimate relations [25].In studies focusing on the general level of aggressiveness, the connection between poor quality of parenting experienced in childhood and problems with behaviour in adolescence has been pointed out [26].The research presented in this article aims at finding the patterns of aggression and victimization determinants in adolescents and their relations with attachment level.

Research Problem
This research examines whether the aggressor or victim behaviours in early and late adolescence are conditioned by specific patterns of attachment.In the testing procedure of the assumed research model two groups of independent variables were employed.The first group includes parents' education level, their financial status and the number of the child's siblings.Those variables, being relatively constant, are shaped prior to dependent variables and may be considered a certain context of a person's development.The second group is attachment patterns in relations with one's mother and father, which are treated as dynamic developmental variables.It is assumed that the perception of attachment and its particular factors is a variable that tells us about the process in the family creating its psychological context which in the child's adolescent years is shaped by both sides of the relationship.On the other hand, the readiness for assuming the aggressor and/or victim role is treated as a dependent variable.
The research procedure was intended to answer the question whether certain determinant patterns can be established in adolescents' readiness for assuming the aggressor or victim role.
The research hypothesis assumed that independent variables, such as the family's socio-economic conditions, do not have a direct impact on the frequency of aggressive or victim behaviour.A more complex relationship was assumed, in which attachment is a process determining the frequency of undertaken aggressor or victim behaviour.Research conducted to date points out to the necessity of treating attachment and its constituents as process variables.Therefore, sharing that view, the existence of a relation between perceived attachment and independent variables, i.e. the life context of a given family and the age and sex of the respondent have been assumed.

Measures
For the purpose of examining a person's experience and frequency of being in the role of aggressor or victim, the Mini-DIA by Österman-Björkvist [27] was employed.That method examines the frequency of being a victim or perpetrator in the spheres of physical aggression, verbal aggression and indirect aggression, also known in Polish literature as psychological or social aggression.The respondent evaluates the frequency of examined behaviour in a 5-step scale (always -5, often -4, sometimes -3, seldom -2 and never -1).The scale yielded a Cronbach's alpha of 0.89.
Attachment was evaluated with the two subscales (mother and father attachment) of the Inventory of Parent and Peer Attachment (IPPA).IPPA was developed by Armsden and Greenberg [28] -experimental version, serving the purpose of evaluation by young people of affectivecognitive patterns of attachment as the source of psychological security.The theoretical framework for IPPA is the attachment theory originally formulated by Bowlby [29] and recently elaborated upon by other researchers.Four broad dimensions of attachment are evaluated: attachment bond, level of mutual trust, quality of communication and extent of anger and alienation.Each scale contains 25 questions coded on the 5-point Likert scale.The obtained scores fitted within the range of 25 to 100 points, and for the particular subscales they were as follows: Attachment Bond and Trust 10-50, Communication 9-45, and Alienation 6-30.Cronbach's alpha reliability coefficients are as follows: Attachment Bond -Mother subscale 0.87, Attachment Bond -Father subscale 0.89.The psychometric validity tests produced satisfactory results.The test is widely used in the USA and only sporadically used for research in Poland [28,30].
The tool used for measuring the level of personal variables (hostility and anger) was the Buss-Perry Aggression Questionnaire (the Aggression Questionnaire) designed by Arnold Buss and Mark Perry in 1992 [31].The scores are normalized on a scale of 0 to 1, with 1 being the highest level of aggression.The questionnaire returns scores for 4 dimensions of aggression: Physical Aggression, Verbal Aggression, Anger, and Hostility.The first two factors represent a motor or instrumental component; anger, which implies psychological activation and preparation for aggression, is the emotional or affective component and hostility represents the cognitive component.Thus, the questionnaire was made up of 29 items, yielding a minimum score of 29 points and a maximum score of 145.The internal consistency coefficients were as follows: Physical Aggression, α= .85;Verbal Aggression, α= .72;Anger, α= .83and Hostility, α= .77,with the internal consistency being α= .89[31].
In order to evaluate the socio-demographic variables, a questionnaire was used that gathered information about the age and sex of respondents, number of siblings, mother's education, father's education and the family's financial situation.

Demographics
The sample consisted of 120 members aged between 13 and 20, mainly secondary school and university students.There were 72 girls (60%) and 48 boys (40 %) in the sample.The average age was 18.68 (cf.Table 1).
In the examined population the average number of siblings was two (91%).Only children formed 6% of the total and persons with 4 and more siblings accounted for 3%.The family's financial situation was evaluated as average by 44% of subjects, as poor by 53% and only 3% considered the financial situation of their families to be good.The predominant type of mother's education was secondary (38%) and vocational (37%).Twenty percent of respondents had mothers with higher education and 5% with primary education.Fathers predominantly had vocational (42%) or secondary education (40%).Higher education of the father was recorded for 14% of respondents, while 4% of them had fathers with primary education.

RESULTS
The main research question pertains to the determinants of aggressor and victim behaviour as a function of attachment pattern and socio-demographic characteristics.The analysis was intended to obtain uniform research subjects separated on the basis of their similarity in readiness for assuming the aggressor or victim role.The cluster analysis with the k-means method (Quick Cluster) was used.In order to make comparisons between clusters, the ANOVA test and the Kruskal-Wallis test were used.
In the entire sample (N= 120) three clusters of persons were identified.The similarities within each cluster were independent of the respondents' age.
The clusters were named as follows: "assertive profile", "victimization/victim profile" and "perpetrator/aggressor profile".The parameters of particular variables in particular clusters are presented in Table 2.
The first cluster -the so-called "assertive profile"consisted of persons with low scores in the frequency of assuming the aggressor and victim role (victim's M = 2.24, SE = 0.189, SD = 1.4,aggressor's M = 1.95,SE = 0.169, SD = 0.1.2).The second cluster consisted of persons who scored high in the frequency of assuming the victim role and had average scores for aggressor role.That profile was called the victimization or victim's profile (victim's M = 6.63,SE = 0.256, SD = 1.52; aggressor's M =3.2, SE = 0.265, SD = 1.57).The third profile, called the "aggressive profile" or "perpetrator's profile", consisted of respondents who scored high for both aggressor and victim roles but act more often as aggressors than victims (victim's M = 4.39, SE= 0.379, SD = 2; aggressor's M = 6.25,SE = 0.302, SD = 1.6).More importantly, the cluster analysis demonstrated that all perpetrators consider himself or herself a victim as well.
The profile most often encountered in the sample was the assertive one, with 55 persons (46%) found to fit into it.The victimization cluster profile was represented by 35 respondents (29%) and the aggressive profile by 30 respondents (25%).
Once the groups had been established, a subsequent question was whether young people from particular clusters differed in the measured variables.The differences found are presented in Tables 3 and 4.
In the identified profiles, the respondents' sex and variables, such as mother's and father's education and the family material status, proved insignificant for the frequency  of undertaken aggressor/perpetrator and victim behaviour (cf.Table 3).
However, the analysis of determinants for undertaking the aggressor/victim role pointed out the significance of trust towards mother and towards both parents.No other significant differences between groups emerged, neither did age, number of siblings or the aggressiveness level (cf.Table 4).
A post-hoc comparison indicates that persons from the assertive profile have considerably higher scores in attachment in the Trust-Mother dimension (3, in  The results allow one to compile the characteristics of selected clusters on the basis of the examined variables.Young people from the assertive profile (i.e.those who rarely act as victims or aggressors) are characterized by a certain specific pattern in their attachment, i.e. higher level of trust towards both parents and the lowest level of hostility.
Adolescents from the victimization and aggression profiles tend to display a lower level of trust towards mother and towards both parents than the assertive group.They also have a higher level of hostility.At the same time, representatives of the victimization profile (more often victims than perpetrators) display lower trust in relations with father and subjectively lower level of wealth, and their mothers have lower education than mothers of respondents from the other two groups.

DISCUSSION
The objective of this study was to better understand the behaviour of adolescents who assume the perpetrator and/or victim role by pursuing two lines of explanation -whether readiness for entering the two roles is characterized by certain behaviour patterns and in what way the sex and age of respondents, socio-demographic features of their families and their attachment styles are connected with the frequency of such behaviour.Although future studies are needed to boost the confidence level of the results, the conclusions of the present study have important implications.
The presented research was able to obtain three distinct behaviour patterns connected with assuming the role of aggressor or victim by adolescents.The first profile, called the assertive one, refers to persons who can function well in their peer group.Such persons can easily cope with difficulties, and are ready to control their own emotional states and manage emotions.Persons from that profile display stable, secure attachment patterns.The remaining two profiles are within the domain of psychopathology and indicate the existence of two behaviour patterns: assuming the victim role (intensification of victimization experience) and assuming the role of aggressor.This is in line with the

Victim profile
Aggressor profile Assertive profile current state of research undertaken from the perspective of developmental psychopathology, which treats maladjustment as a continuum of two forms of difficulties: externalization or a lack of control (e.g.aggression), and internalization or excessive control (e.g.social withdrawal) [32].The victimization profile is largely connected with internalization problems because, as research shows, difficulties with overcontrol beginning already in childhood (such as social withdrawal) are the reasons behind loneliness and depression [32][33][34][35][36]. Also, social withdrawal entails and may be the cause of such problems as peer rejection [37] or victimization [38,39].The results obtained from this research are consistent with other empirical reports related to relationships between the level of victimization and (among other things) the insensitive attitude of parents towards the child and insecure attachment patterns [40].Besides, the longer the child's social withdrawal lasts, the greater the feeling of loneliness, depression, peer rejection and victimization.
The number of people in each of the three profiles indicates that assertive-type persons are most numerous, i.e. those who do not encounter problems with either perpetration or victimization.
Interesting results have been obtained for persons who most often assume the aggressor role.Such "aggressive profile" persons report at the same time a strong sense of being victims.This result may be explained from the sociocognitive point of view [41] describing the development of anger reaction in children and its consequences in the form of aggressive behaviour.The key role here is played by dysfunctions in the ability to process information [42].The results obtained in the presented research indirectly attest to the occurrence of cognitive deficits in aggressors, consisting in faulty interpretation of stimulating events.This confirms earlier reports from research on distortions in socio-cognitive evaluations in young perpetrators.Those distortions manifest themselves, for instance, in the tendency of very aggressive boys to make hostile attributions [43].It has been found that the tendency for hostile attributions is visible both in adolescence and in earlier years, which indicates that the discussed cognitive distortions are formed during first socialization experiences with family and peers.The research has also shown that apart from incorrect evaluation of intentions of other interaction participants, aggressive boys also have the tendency to underestimate their own aggressive behaviour [44].Due to mistakes in the evaluation of one's own and other people's behaviour at early stages of the conflict, aggressive boys more or less overtly ascribe responsibility for the conflict to their peers, placing themselves in the role of victims.As a result, perpetrators may consider their behaviour fully justified, which is confirmed by research results.
According to circular and ecological theories we might assume that the processes and behaviors that are taking place are characterized by a circular nature.It means that the attachment relationship is constantly changing during interactions between parents and child.On the other hand we could say (based on our and previous studies) that insecure patterns of attachment may prognosticate readiness for undertaking aggressive behaviour.They would potentially influence the frequency of aggressive behaviour and/or victimization.A pattern of such behaviour is stored as teenage experience and may be activated in every subsequent situation, becoming in a sense the determinant of readiness for certain types of behaviour [44][45][46][47].Those experiences may facilitate the access to certain behaviour, become a source of self-knowledge, form the basis for creating one's self-image, and influence the emerging personality [1].However, that hypothesis was confirmed only partly.The profiles identified in the cluster analysis indicate that respondents from the victimization and aggression profile are characterized by a weakened level of attachment.Persons who tend to assume a victim role have decreased level of trust in the relationship with their mothers.
For persons from the "aggressive" profile, a specific result was also obtained for the parental dyad (the highest level of trust for mothers, average for fathers and the highest level of alienation and hostility for both mothers and fathers).Those results are consistent with research on parental attitudes in families of aggressive or victimized children.The studies conducted by Dominiak-Kochanek, Frączek and Konopka [48] clearly point out to the existence of problems with aggressive behaviour in adolescents whose parents have contradictory or crossed attitudes (e.g.high level of trust for mother, low level for father, just like for the "victim" group).In a similar vein, the research of Farnicka and Liberska [46] indicates the link between undertaking aggressive behaviour and hostile, rejecting and inconsistent attitudes.Those attitudes may be characterized by parents' alienation and excessive trust which lead to a lack of parental control/ intervention when boundaries are crossed.Such a manner of parental care is considered detrimental to identity development, particularly in adolescencea period when a child needs clear-cut boundaries and their redefinition in contacts with parents [1,12,49].
Persons from the assertive profile displayed the safest attitude patterns as compared with their peers who engaged in aggressive behaviour or acted as victims.Those results are consistent with hitherto conducted research that indicates the existence of a relationship between secure attachment and low level of aggression [50,51].That relationship may be explained by the results from related studies that associated securely attached children with better self-image, higher empathy level, higher level of positive emotions and lower level of negative emotions [52].For older children in particular, connections have been made between secure attachment patterns and social competences.Securely attached children, as compared to their insecure counterparts, had more friends, were more socially attractive, relied more on their parents (who tend to be emotionally available if the situation requires it), and were more able to develop trust towards others in social relations [53].Such a state of affairs facilitates the development of such features (mentioned above) as high self-esteem, empathy and positive affect which makes those children less aggressive and less hostile in relation to others.It is highly probable that those children will be more assertive in various social situations, and rarely assume the role of aggressor and/or victim.Meanwhile, children with insecure attachment patterns cannot develop those positive features (self-esteem, friendliness, etc.).What is particularly important is that they have difficulty developing trust towards others, which in turn leads to aggression and behaviour disorders.Intimidation, physical and verbal aggression, bullyingall these are forms of attention-seeking behaviour and the mounting sense of harm (heightening the frequency of undertaking the victim role or the sense of being a victim) increases aggression.As long as such children are denied adequate attention at home, they will try to elicit it, through aggression, from their peer group.It is highly probable that such behaviour will intensify in early adolescence due to the process of identity development and the necessity to renegotiate relations with parents.Thus, children with secure attachment patterns most probably have more positive relations with their parents, particularly in adolescence, and vice versachildren with insecure attachment patterns have more negative relations at home, and their negative behaviour reflects those relations.It is also highly probable that insecure children are less accepted and more often rejected by parents and more frequently exposed to corporeal punishment and coercion in the upbringing process.Relationships between rejection, corporeal punishment and aggression are quite well documented empirically [54,55].
The analysis of the remaining socio-demographic factors such as mother's education level, the number of siblings and the perception of the family's financial status failed to prove their significance for assuming the aggressor or victim role.That result is consistent with earlier research.Researchers such as Fagot and Kavanagh [56], or Lewis, Feiring, McGuffog and Jaskir [23] already noted the links between aggressiveness and attachment and family material status.Moreover, research results have clearly indicated that the dimension of hostility is important for undertaking the assertive or aggressive behaviour.The sense of hostility indicates the significance of the ability to cope with emotional states and indicates the reason of cognitive disturbances that may occur both in the victim and in the perpetrator at the moment of interpretation of the stimuli from a given situation.The hostility level is also important for the interactions within the family (secondary impact on attachment bond) and it may lower the level of trust in relations between parents and children.That conclusion is in line with reports made available so far and the hypothesis about the lack of direct relationship between variables.The results obtained indicate the significance of psychological processes and states that are influenced by relatively stable attachment patterns and cognitive patterns of reality interpretation (hostility).

CONCLUSIONS
In the present study, the authors attempted to describe the family-related determinants (static and dynamic) for the undertaking of aggressor and/or victim role, with particular focus on attachment patterns.The research was exploratory and has proven that psycho-social factors play an important role in shaping aggressive behaviour in teenagers.The obtained results have allowed the authors to describe the specific environmental (family-related) and developmental factors that influence the forming of a perpetrator or victim identity.Of particular interest is the relationship between entering the aggressor role and the sense of being harmed and being victimized in social relations.This partly explains the increased irritability of persons demonstrating aggressive behaviour.Moreover, one should notice the specific insecure attachment patterns among persons who appear aggressive or prone to take the victim role.The results match the hitherto conducted research in this field [22,23,56].In subsequent studies, the connection between non-observance of norms and children aggressiveness and antisocial behaviour in the adolescence period was analysed, although only some studies explore their relations with attachment.Numerous contemporary studies on the subject of attachment in adolescents focus more on the current quality of their relationship with parents than on the inner attachment models developed in childhood.For instance, as we said above, Simons, Paternite and Shore [24] have found that the quality of relationship with mother as perceived by teenagers is connected with their evaluation of their own aggressiveness.Thus, areas were identified which should definitely be taken into consideration in the process of diagnosing and when designing preventive and therapeutic activities with regard to problem behaviour of young people connected with aggression or victimization.The presented studies have their limitations, as they describe the issue predominantly from the cross-sectional perspective.Longitudinal studies would show the individual dynamics, not only the cohort one.Still, the analysis of the results has proven that numerous (but not all) assumed dependencies have actually been found in the studied groups.Their further exploration in a widened research model may produce interesting results about the determinants of aggressive and victim types of behaviour.