PARENTAL ATTACHMENT AS PREDICTOR OF DELINQUENCY

Purpose – The purpose of this study is to fi nd out the relationship between adolescents’ parental attachment and delinquent behaviour in the aspects of trust, communication and alienation. In this study, delinquent behaviour refers to the set of behaviours which is against any mainstream established rules or norms. It is believed that delinquent behaviour has the potential to cause school failure, impairments in socio-emotional development, peer rejection, and adult crime. Methodology – Data were gathered from a sample of 300 secondary school students aged 15-18 years old in Selangor. These students were selected from the “hardcore schools” which was obtained from the so called "hardcore schools" listed by the School Division for serious disciplinary problems. Two instruments, namely surveys on the “Behaviour of Students” and “Parental Attachment”(PA) were utilized to collect the research data. The researchers employed correlational research design in this study. Findings – Data showed that the highest delinquency among the adolescents was misbehaviour in school, followed by crime, vandalism, pornography, dishonesty and drugs. Results also revealed a negative linear relationship between parental attachment (r=-0.282,p=0.0001) and delinquent behaviour, implying that adolescents with better parental attachment had lower levels of delinquency. In addition, fi ndings showed that adolescents trusted in their mothers (M=3.70, SD=0.63) more than they trusted in their fathers (M=3.51, SD=0.72) in relation to communication and alienation. It is suggested that close supportive parents and frequent 100 Malaysian Journal of Learning and Instruction: Vol. 10 (2013): 99-117 communication between parents and their children serve to prevent adolescent delinquency. Signifi cance – This research is signifi cant to parents, school, teachers, counselors and the community.


INTRODUCTION
Delinquent behaviour refers to the set of behaviours that violates the established rules or norms of mainstream society. The delinquency happens within school settings. It is believed that the early stage of delinquent behaviour has the potential to cause school failure, impairments in socio-emotional development, peer rejection, delinquency, and adult crime (Moffi tt, 1993). Furthermore, early onset of delinquency signifi cantly increases the risk of serious, violent, and chronic offending in later years (Loeber & Farrington, 2000).
Past studies have revealed that there are many factors that are said to contribute to delinquent behavior among adolescents (Ezinga, 2011). These factors originate either from personal characteristics or other external factors including the individual's past experiences, interests, attitudes, and expectations that could infl uence their interpersonal relationships. This is consistent with previous research by Patterson et.al (1991), who described the main environmental causes of antisocial and eventually delinquent behaviour. They have studied many different environmental factors such as parents, peers, schools and individuals that can infl uence the wholesome development of the child, physically, emotionally, socially, and spiritually.
According to Sprinthall (1995), early adolescence is a time of both intensive and extensive transitions that affect individuals physically, psychologically and socially. During this period, adolescents are more vulnerable to emotional diffi culties and seek comfort and emotional support either from parents or peers. Thus, parental attachment plays an important role in helping adolescents overcome their emotional diffi culties during childhood and beyond adolescence, especially in early adolescence.

ATTACHMENT THEORY
According to Bowlby (1982), attachment is an emotional bond. This bond comprises comfort, safety, and support to an individual. In the past few decades, John Bowlby's attachment theory has greatly enriched our understanding of social development during infancy and early childhood. More recently, attachment theory has provided a theoretical basis for understanding the presence of emotional and behavioural problems during adolescence (Marsh, McFarland, Allen, McElhaney, & Land, 2003). Therefore, attachment can be seen as an important construct in understanding future development and behavioural patterns of individuals.
Attachment has been defi ned as an intimate and affectionate relationship between two people (Ainsworth, Blehar, Waters, & Wall, 1978). Bowlby (1982) further defi nes attachment as a child being "strongly disposed to seek proximity to and contact with a specifi c fi gure and to do so in certain situations, notably when he is frightened, tired or ill". Thus, during the course of a healthy development, attachment behaviour leads to the development of a bond of affection between child and parent (Goodwin, 2003). Bowlby (1982), in his attachment theory mentioned that people develop the mental models that are related to their relationships with other individuals during infancy, childhood, and adulthood. These mental models are the ones that unify and organize the individual's experiences and beliefs. In his theory, the quality of early attachment relationships is rooted in the degree to which the infant has come to rely on the attachment fi gure as a source of security (Ainsworth et al., 1978). Armsden and Greenberg (1987) suggested that the following three factors shape the quality of parental-child attachment: trust, communication, and alienation. All three factors contribute to the quality of parental attachment, either positively or negatively (Bogels, & Brechman-Toussaint, 2006;Ainsworth et al, 1978;Armsden & Greenberg, 1987;Bowlby, 1988).
Indeed, evidence suggests that securely attached children are more acceptable and less rejected by their peer group compared to those with insecure parental attachments. In addition, attachment security also appears to be related to dyadic peer relationships whereby securely attached children have more friends and have more harmonious, more satisfying and less confl icted friendships than insecurely attached children (Kerns, Klepac, & Cole, 1996). Meanwhile children who are classifi ed as ambivalent, hold negative views of themselves. (Main, Kaplan, & Cassidy, 1985).

SOCIAL CONTROL THEORY
Control theories contend that people will commit crimes if they do not develop normative social bonds. The researchers include this theory because it explains why individuals do not engage in delinquent behaviours rather than why they do. The control perspective states that most people do not commit criminal acts if they are strongly bonded to social institutions such as family and "controlled" by the social bond that could prevent them from being involved in delinquent behaviours. However, if the individuals fail to form a strong bond to their families or any signifi cant model, they are likely to engage in criminal activities.
Among the control theories, Travis Hirschi's social bond theory (1987) is one of the most frequently investigated frameworks for the parent-child relationship and delinquency (Vold, Snipes & Bernard, 2002). Hirschi's (1969) theory mentions that human beings are not innately programmed to conform to social norms or rules, but are rather guided by primitive instincts and hence "naturally capable for committing criminal acts." One of Hirschi's fundamental propositions is that there is an inverse relationship between social bonds and delinquency (Hirschi, 1969). It assumes that delinquent behaviours occur when the individual's bond to society is weak or broken.
According to Hirschi (1969), the concept of social bond consists of four elements: attachment, involvement, commitment, and belief. Hirschi (1969) argues that these four elements are independently signifi cant barriers to the adolescent's delinquent behaviour. According to Hirschi (1969), attachments are most useful in predicting delinquency for those children who have their parents, schools and peers as the primary agents of socialization. The children's attachment to parents is the most crucial among the bonds, and it serves as a foundation upon which a strong positive attachment can be built to serve as a barrier to delinquency (Hirschi, 1969). If the bond of affection to the family is strong, the attachment formed may be able to deter criminal behaviours. The theory emphasizes that the stronger the bond is, the greater the likelihood that the child will internalize the social norms and make rational considerations when faced with moral dilemmas.
Another study that examined the social bond theory is the longitudinal study conducted by Wiatrowski et al. (1981), using a sample of 2,212 school boys to test how the four elements of social bond operating with respect to delinquency. The study shows that parental attachment has a strong negative relationship with delinquency, as predicted by the social control theory. On top of that, the study provides empirical support for the claim that social bonds are not only built in the context of a family but also in schools. These empirical tests of Hirschi's social control theory strengthen the explanation of certain elements of social bonding, particularly attachment, in predicting delinquency among younger adolescents.

PARENTAL ATTACHMENT AND DELINQUENT BEHAVIOUR
Warm and consistent parenting is related to "secure" attachment, which is characterized by the child's use of the attachment fi gure as a secure base. Adolescents with secure parental attachments may contribute to positive views of self and lower levels of depressive symptoms (Kenny et.al, 1993). Securely attached adolescents are less likely to engage in delinquent behaviour. They also manage their school work more successfully, and enjoy more positive relationships with family and peers. In fact, they demonstrate less concern about loneliness and social rejection and display more adaptive coping strategies (Armsden & Greenberg, 1987).
On the contrary, insecure attachment such as alienation, and dismissal attachment are associated with parental negativity and rejection. Adolescents who do not have a trusting or secure attachment to their parent will be anxious, lack confi dence, be depressed and have social withdrawal inclinations (Ryan & Lynch, 1989). This is further elaborated by the fi ndings of Elgar, Knight, Worrall and Sherman (2003), whereby internalizing and externalizing problem behaviours were associated with insecure attachment. Further research by Arbona and Power (2003), found that adolescent with secure parental attachment indicated a higher level of selfesteem and lower level of delinquent behaviour than their insecurely attached counterparts.
Meanwhile, the research by Cooper, Shaver and Collins (1998) on the impact of attachment among adolescents with delinquent behaviour, such as excessive alcohol use and a variety of antisocial behaviours, found that adolescents with insecure-anxious attachment reported higher tendency for truancy and property-related crimes than the more secure adolescents. Insecure groups of adolescents reported higher levels of drug use than the more securely attached group.
The fi ndings also indicate those insecurely attached adolescents were signifi cantly more likely to be involved in casual sex and often pregnancy.
Since most of the studies on parental attachment have been conducted in Western societies, the aim of the present study is to examine the relationship between parental attachment and delinquent behaviour among students in Malaysia. It is hypothesized that higher parental attachment will result in lower engagement in delinquency in school. Finally, the current study investigates the extent to which parental attachment contributes to the development of delinquent behaviour among secondary school students in Malaysia.

METHODOLOGY
Participants in the study were 300 secondary school students (200 boys, 100 girls; 178 Malays, 84 Chinese, 40 Indians) aged 15-18 from 10 schools in Selangor. These schools were selected because they were listed as "hardcore schools" in relation to discipline problems obtained from the Schools Division in the State of Selangor. The respondents were identifi ed with the help of teachers and counselors based on their delinquency behaviour, represented in their school's demerit systems. They were then self-assessed through the "Behaviour of Students" with six subscales of delinquent behaviour to identify the delinquent acts.

MEASURES
Two instruments, namely "Behaviour of Students" and "Parental Attachment"(PA) were used to collect the research data. "Behaviour of Students" was established by Rozumah, Rumaya, Asnarulkhadi, Amna, Mansur and Tan (2003). It was used to assess students' involvement in delinquent behaviour over the last eight months. The instrument comprised 35 items with six subscales to identify the six different types of delinquent acts exhibited bythe respondents. The six subscales of this instrument are categorised as crime, vandalism, drugs, pornography and sexual behaviours, dishonesty and other misbehaviours. The response to the items were scored on a 5 point Likert scale, 5 for "very frequent", 4 for "frequent", 3 for "moderately frequent", 2 for "not frequent", and 1 for "never". Basically, the delinquent behaviour included in the items cover the following offences: stealing something worth RM50.00 or less, stealing something worth RM50.00 or more, shoplifting, buying stolen goods, selling drugs and discipline problems in school such as fi ghting, vandalism, throwing items, verbal bullying, graffi ti in school, threatening or using violence in the classroom. The lowest score is 35 and the highest is 175 points. High scores in each category indicate a high incidence of delinquent act among the students in that particular category.
The parental attachment instrument was developed by Noriah Ishak (1999) based on the work of Armsden and Greenberg (1987) to assess adolescents' perceptions of the positive and negative affective or cognitive dimension of the relationships with their parents. The theoretical framework used in designing the instrument was the attachment theory, originally formulated by John Bowlby (1982). The instrument consists of three broad dimensions. They are degree of mutual trust, quality of communication, and extent of anger and alienation. It also generates a composite score of the overall quality of their parental attachment. The instrument is a self-report questionnaire with a fi ve-point Likert-scale response format. The respondents rated themselves from 1 (never true) to 5 (always true).
The Parental Attachment instrument has two parts: 30 items for mother's attachment and 30 items for father's attachment. Meanwhile, the Cronbach values for mother's attachment, father's attachment are 0.7957 and 0.8438, respectively. These values confi rmed that this instrument is reliable and suitable for secondary school students in Malaysia. Examples of the items are, Trust:"I feel that my mother is a responsible person"; Communication: "I always get my mother's opinion on important issue"; Alienation: "I feel embarrassed when discussing important issues with my mother".

DATA ANALYSIS
The data collected were analyzed using the Statistical Package for the Social Sciences (SPSS Version 19.0). Procedures for descriptive statistics were carried out to provide answers to the research questions for this study. Descriptive statistics of frequency, percentages, mean and standard deviation were used to sort the scores attained from the students based on gender, race, age, crime, verbal aggression, truancy and vandalism in the instruments.
Pearson Correlation Matrix was used to examine the intercorrelations among predictor variables and the criterion measures. Variables were examined from various assumptions underlying multivariate analysis such as normality, linearity, independence of errors and multi-co-llinearity. In this case, parental attachment fi ts the assumptions of the multivariate analysis. In order to examine the amount of variance contributed by each of the predictor variables in infl uencing delinquent behaviours, multiple regression analysis was used.

Levels of delinquent behaviours in the aspects of crime, drugs, vandalism, pornography and sexual behaviours, dishonesty and other misbehaviours
For the fi rst research question, descriptive statistics were used to determine the level of the adolescents' delinquency. The analysis shows the distribution and means scores with standard deviation for all the delinquency categories (Table 1).
Based on the interpretation of mean scales, it was found that generally, the level of other misbehaviours (

Levels of attachment by fathers and mothers in the aspects of trust, communication and alienation among the students with delinquent behaviours
The second research question concerned the level of father's and mother's attachment in the aspects of trust, communication and alienation (Table 2).

a) Level of mother's attachment in trust, communication and alienation
Based on the fi ve-point scale used,

b) Level of father's attachment in trust, communication and alienation
Based on the fi ve-point scale used,

Relationship Between Parental Attachment and Delinquent Behaviour
The third research question is whether there is any signifi cant relationship between parental attachment and delinquent behaviour amongst students in secondary school. The hypothesis for this research question states that the researchers expected a negative linear relationship between parental attachment and delinquent behaviour among students in secondary school. Table 3 shows the Pearson's correlation coeffi cients of delinquent behaviour and parental attachment. As depicted in Table 3, there is a negative signifi cant relationship between delinquent behaviour and parental attachment (r= 0.282, p<0.01). The negative correlation coeffi cients indicate that as the parental attachment increases, the score on delinquent behaviour decreases considerably. In general these results supported the hypothesis.
A Pearson Correlation was conducted as a further test to see the relationship between the aspect of parental attachment (trust, communication, and alienation) and delinquent behaviour. Findings showed a negative signifi cant relationship on communication between parents (r = -0.126, p<0.05) and adolescents while trust and alienation were found to have no signifi cant relationship with delinquency. The negative correlation coefi cients indicates that as communication between parents increases, the score on delinquent behaviour decreases considerably.

To what extent does parental attachment explain the variation in delinquent behaviour among students?
The fourth research question is to determine how well parental attachment predicts delinquency among students. The result of multiple regressions for the parental attachment measures are presented in Table 4. As shown in Table 4, parental attachment was found to be signifi cant in explaining deliquent behaviour (t= -3.567, p=.0001). This means that parental attachment is related to delinquent behaviour. Meanwhile, the beta coeffi cient of parental attachment is -0.202 and this suggests that a one standard deviation increase in parental attachment will be followed by 0.202 standard deviation decrease in delinquent behaviour.
The ANOVA revealed that the F-statistics (F = 12.721) and the corresponding p-value reach statistical signifi cance (p=0.0001) or lower than the alpha value of 0.05. This indicates that the slope of the estimated linear regression model line is not equal to zero confi rming that there is linear relationship between delinquent behaviour and the parental attachment.
However, since none of the factors had tolerance value smaller than 0.10 and VIF statistics were less than 10.0 (Table 4), we conclude that there are no serious multi-collinearity problems among the predictor variables. As a result, it is safe to say that the predictor variables entered in the regression model showed stable and unique characteristics of their own to explain their contribution in predicting delinquent behaviour. It can be concluded that the regression model confi rmed that parental attachment is a signifi cant predictor for delinquent behavior in adolescents.

DISCUSSION
The current study examined the role of parental attachment in predicting adolescents' delinquent behaviour. From the fi ndings, it was found that parental attachment (t=-3.567, p=.0001) was related of delinquent behaviour. It was expected that higher levels of parental attachment would be associated with less engagement in delinquent behaviour. This fi nding is supported by Cooper, Shaver and Collins (1998) who found secure parent-child attachment to be associated with lower rates of adolescent sexual activity and lower teen pregnancy rates. Consistent with the fi ndings, Arbona and Power (2003) believed that adolescents with secure parental attachment tend to have lower levels of delinquent behaviour. Adolescents who have good parental attachment appear to be very well-adjusted. They also possess higher self-esteem, and enjoy frequent and satisfactory communications with their families.
Both previous studies and the current study have been proven to be consistent with attachment theory, which holds that internal working models (formed from child-parent earlier relationship), served as a foundation of beliefs on individuals as well as a guide for all future close relationships (Bowlby, 1982). The fi ndings also support Hirschi's Social Control Theory(1969), which emphasized that family has a profound impact on subsequent behaviours that a child would exhibit later in life. If the attachment between parents and adolescents is strong, adolescents will tend to conform to the norms of society as their own. They would be less involved in any risk-taking behaviour.
Furthermore, this research also shows a negative signifi cant relationship between parental attachment and delinquent behaviour (r = -0.282, p = 0.0001). The negative correlation coeffi cient of 0.282 was small; it indicates that as the parental attachment increases, the score of delinquent behaviour decreases considerably. This result is in line with the fi ndings by Wiatrowski et al. (1981), which reported a strong negative relationship between attachment and delinquency. As such, family environment serves as foundation, in which a strong positive attachment could be a barrier to delinquency.
In this study, the research also found adolescents have higher trust in their mothers (M=3.70, SD=0.63) than fathers (M=3.51, SD=0.72).
The trust between children and parents is partly developed when children are able to reach out to their parents and know that help is available when needed. These fi ndings supported Bowlby's (1988) notion regarding accessibility of parents and healthy development among children. It is manifested by the show of understanding, respect and responsiveness from caregivers (parents) to the adolescents' needs, desires, and perceptions. It is important for relationship-building between children and parents.
The adolescents in this study showed less communication with their fathers (M=3.16 SD=0.73) compared to their mothers (M=3.40,SD=0.69). This fi nding is consistent with Ainsworth (1989) who reported that adolescents are more attached to mothers than to fathers. Communication leaned towards mothers who provide more empathic understanding, listened attentively and are more passionate about their children. These fi ndings support Bowlby's (1988) argument that mothers are better providers than fathers when it comes to psychological support. Mothers are the safety net because they have the natural instinct to ensure children's safety exists over time. This could be also due to the fact that in Asian families, the father is regarded as the head of the family. Besides, the father's role is seen as more serious and more disciplinary.
In addition, Pearson correlation analysis shows a negative signifi cant relationship on communication between parents (r = -0.126, p<0.05) and adolescents while trust and alienation were found to have no signifi cant relationship with delinquency. According to Armsden and Greenberg (1987), effective communication between individuals and their caregivers is essential in developing healthy attachment from young. It helps in developing trust and the feeling of trustworthiness between the two individuals during their period of interaction. Thus, parents not only inculcate the values and norms, but the nature of their relationship with their child reduces opportunities for delinquent engagement.

IMPLICATIONS
This study indicates that Malaysian parents should be more aware and concerned of their children's feelings especially if those children are displaying delinquent behaviour. Moreover, parents should know how to accept their children unconditionally. Parents should help their child to overcome negative emotions and build a strong relationship between parents and child. Besides making more effort to spend more time with their children at home, parents should also communicate with them frequently. Through communication, parents would be able to listen to their children's emotional states and give them an opportunity to say what is on their mind. With communication and trust that are built in the relationship, parents should take the opportunity to know about the people, places and events in the child's life. In addition, parents need to be more aware of their children's potential, and of changes in their children. More attention should be focused on the changes of their children's attitude.
Cultivating problem solving in their children is also important to understand the children's attitudes and emotions. Lastly, parenting qualities such as fairness, understanding, pride, and trust, facilitate positive development and help adolescents avoid social illness.
Meanwhile, the fi ndings of quality attachment to parents are useful to school teachers, and principals for a better understanding of the student delinquency. Teachers can create a learning environment that nurtures not only academic success, but also affective development by improving adolescents' communication skills. Open discussion is an excellent way to expose adolescents to a variety of communicative styles which will give them the opportunity to vent their feelings within a secure and teacher-controlled environment.
Schools should build partnerships with parents to strengthen the moral values of students outside the classroom through community activities. These community activities can include a variety of neighborhoods ranging from the well-kept to the neglected, the affl uent and the poor. The students have the opportunity to show their strengths and potential while doing the activities and at the same time cultivate moral values among them. Parents are encouraged to talk and show their interest in what their children learn and also the happenings in schools.
Besides that, the fi ndings also contribute towards helping school counsellors in providing proper counselling, activities and programmes to manage student behaviour. These fi ndings would give some insights to the counsellors to organize community-based after school programmes such as crime prevention. Parents are encouraged to participate in the activities in order to understand the issue of delinquent behaviour.

LIMITATIONS OF THE STUDY
The population of this study is limited to the students from 'hardcore' secondary schools in the state of Selangor. The population does not fully represent the entire Malaysian student body with delinquent behaviour.
A self-rating instrument was used to measure all the variables in this study because it is a more direct measure of personal variables than any other techniques and it is appropriate for adolescents, depending on their reading abilities (Bracken, 1996). Besides, nearly all instruments that examine attitudes, interest, emotive feelings and perception are self-rating measures (Gay & Airasian, 2005). Nevertheless, this study also acknowledges that self-rating instruments suffer some problems. The researcher could never be sure that the individual is expressing his or her true responses, as opposed to 'socially acceptable' responses.
Finally, it is also important to consider that delinquent behaviour is the result of a wide range of factors, not only parental attachment. Other factors such as school culture, school environment, teachers, and adolescent autonomy may contribute to delinquent behaviour (Ezinga, 2011). These factors, however were not analyzed in this study. This was to avoid confounding results as the study involved many independent variables, a complex research design, and various statistical analyses. Despite these limitations, this study was able to contribute to the literature by expanding our understanding of the etiology of delinquency.

CONCLUSION
Overall, the current study provided strong support for parental attachment in predicting adolescents' delinquent behaviour. The results highlight the importance of parental attachment towards their children. If the bonding of affection is strong, the attachment formed may deter criminal behaviours among secondary students.
Although this study provides increased insight to the processes infl uencing delinquency, still there are many questions unanswered. In fact, delinquency does not exist in a vaccum, and it is necessary to look at a variety of risk factors that may predict delinquency. As such, further research is needed to investigate the specifi c aspects of parental attachment which are most infl uential in the well-being of the adolescent and of additional factors that might interact with adolescents' relationships in predicting delinquency. With this information, it might become possible to enable adolescents to focus on the factors that might steer them away from engaging in the most serious forms of delinquency.