Why Some Adolescents Engage in Risk-Taking Behavior

This study explored adolescents’ reasons for involving themselves, or not, in risk-taking behavior, in two vulnerable areas of North Jakarta. The sample was purposively selected among households with adolescents ranging from 12 to 18 years old living in the two areas. The study involved 401 parents (8% female; 92% male; mean age 45.3 years) and 414 adolescents (57.49% female; 42.51% male; mean age 14.9 years). Parents’ demographic data included educational level, employment status, family income, and expenditure. Adolescents were asked about their perceptions of their relationships with their parents, whether they had been involved in eight risky behaviors (smoking, consuming alcohol, substance use, brawling, crime, physical fighting, heavy petting, and premarital sex), and reasons for engaging in risky behaviors or not. The comparison of proportions of eight risky behavior was tested by different test procedures, namely Z test, Chi-Square and Marascuillo multiple comparison. Results revealed that older adolescents were more likely to be involved in risk-taking, and boys were more likely to engage in risky behaviors than girls. The study also indicated that curiosity and peer pressure were the main reasons adolescents engaged in risky behaviors. Advice from family members, fear of God, and fear of being sinful were reasons adolescents did not engage in risky behaviors.

criminal gangs were influenced by parents' membership in criminal gang. The study of Augustyn et.al. showed that parents are role models of behavior for adolescents. The above explanation indicates the important of good relationship between parents and adolescents.
Previous research on adolescents' risk-taking behavior reported that peer groups have both positive and negative impacts. Maxwell (2002) found that peers might influence adolescents to begin smoking cigarettes and marijuana. However, peers can also influence adolescents to stop consuming alcohol and chewing tobacco. Mason et al. (2014) showed that close friends' attitudes impact adolescents' substance use. Having unresponsive close friends who pay little attention to or do not care what adolescents do might lead to increased substance use. Conversely, close friends who show concern about and disapproval of substance use can affect its decrease.
A study on the relationship between religion, religiosity, and risky behavior reported that religion and religiosity play an important role in protecting adolescents from engaging in risky behavior, such as smoking and alcohol consumptions (Sinha, Cnaan, & Gelles., 2007;Marsiglia, Ayers, & Hoffman, 2012;Brown, et.al., 2014;Charro Baena, Meneses, Caperos, Preitos, & Uroz, 2018); premarital sex and HIV risks behavior (Gyimah, Kodzi, Emina, & Cofie, 2013;Cerqueira-Santos & Koller, 2016;Hasnain, Sinacore, Mensah, & Levy, 2005), fighting and violence (Salas-Wright, Vaughn, & Maynard, 2014). Marsiglia, et.al. (2012) conducted a study with adolescents aged 14-17 in central Mexico and reported that adolescents who have high scores in both intrinsic religiosity and extrinsic religiosity were less likely to take risks by cigarette and alcohol use. Intrinsic religiosity referred to the importance of religion in personal behavior, while extrinsic religiosity referred to individual's involvement in religious activities such as church attendance. The Arli et al.'s study (2016) with Indonesian youth aged 18-24 showed that intrinsic religiosity influenced youth's perception of risky behavior, such as gambling, health and safety, and ethics. Youth with high scores in intrinsic religiosity were less likely to engage in risky behavior. Additionally, Ameri, Mirzakhani, Nabipour, Khanjani, & Sullman (2017) conducted a study with Iranian university students and found that students who engaged more often in religious activities and had stronger intrinsic religiosity were less likely to engage in risky behavior, such as sexual risk-taking, careless driving, violence, smoking, alcohol, and substance use. According to Sinha, et al. (2007) and Mojahed (2014) religion has a very positive role in decreasing risktaking behavior both directly and indirectly. Religious rules and prohibitions are a direct way in which risky behavior is inhibited, while religious activities and communities are indirect ways in which religion inhibits risky behavior. Religious teachings thus serve as a basis for preventing adolescents from engaging in risk-taking behaviors, for instance, premarital sex, alcohol and substance use, and violence.
Poverty is also regarded as predictor of criminal behaviors among adolescents (Sariaslan, Larsson, D'Onofrio, Långström, & Lichtenstein, 2014;Shah, Soomro & Mirjat, 2019). According to Shah et.al (2019) lack of resources to fulfill the needs lead adolescents to engage in criminal acts. Shong, Bakar and Islam (2018) stated that poor and unhappy family conditions as well as school failure are two factors associated with children's criminal behavior. The study of Madise, Zulu and Ciera (2007) reported that poverty is the driving factor for female adolescents to involve in early sexual activity and having sexual activity with multiple partners. These sexual activities are intended to get gifts or money. The explanation above shows that family life is essentials for adolescents to engage in risky behavior. Therefore, this study focuses on the risk-taking behavior of adolescents from vulnerable neighborhood. This study is also interested in identifying adolescents' perception on their connection with their parents. Another study on adolescents' risk-taking behavior also reported that family financial pressure, disadvantaged neighborhoods, unstable life, and lack of hope and certainty are likely to affect adolescents' risky behaviors (Caldwell, Wiebe, & Cleveland, 2006). Unfavorable and uncertain conditions leading to hopelessness indirectly impact adolescents' maladaptive school and coping behaviors (Bolland et al., 2007). McKelvey et al. (2011) reported that community violence negatively impacts children's psychosocial development, including risk-taking behavior. However, families with low conflict can protect children from community violence's negative effect. Thus, McKelvey et al.'s study (2011) showed that good family life and good parenting play important roles in decreasing adolescents' risk-taking behavior.
Referring to the above explanation on the risk-taking behavior among adolescents, this study is intended to provide an overview of eight risk-taking behaviors namely smoking, consuming alcohol, substance use, brawling, crime, physical fighting, heavy petting, and premarital sex of adolescents in two vulnerable areas in North Jakarta Indonesia. The study also provides information on parent-adolescents relationship as perceived by adolescents which regarded as important factor that influence adolescents' engagement in risky behaviors. In this study adolescents' reasons for engaging in or not in risky behavior is also presented. Knowledge of the reasons for adolescents' involvement in risky behavior will provide information for designing future youth behavior change programs (Buckley & Sheehan, 2016) health promotion programs (Dey, Gmel, Studer, & Mohler-Kuo, 2014), and other prevention programs (Oman, Vesely, Kegler, McLeroy, & Aspy, 2003;Morales-Alema, 2011;Hale & Viner, 2016).
This study provides an overview of eight risk-taking behaviors (smoking, consuming alcohol, substance use, brawling, crime, physical fighting, heavy petting, and premarital sex) of adolescents in two vulnerable areas in North Jakarta, Indonesia. The study provides information on parent-adolescent relationships as perceived by adolescents. In addition, the study provides information on adolescents' reasons for engaging in or not in risky behaviors. Knowledge of the reasons for adolescents' involvement in risky behaviors will provide important information for designing future youth behavior change programs (Buckley & Sheehan, 2010), health promotion programs (Dey, Gmel, Studer, & Mohler-Kuo, 2014), and other prevention programs (Oman, Vesely, Kegler, McLeroy, & Aspy, 2003;Morales-Alema, 2011;Hale & Viner, 2016).

Method
Using a nonrandom sampling method, this study was initially conducted as a household survey in two vulnerable neighborhoods in North Jakarta. The neighborhoods are located in two slum areas with poor quality and highdensity housing, poor sanitation, inadequate access to clean water and other infrastructure. Although some household heads in these two neighborhoods have low-level permanent jobs, most work as temporary informal workers, fishermen, motorbike taxi drivers, truck drivers, port workers, or are unemployed. Therefore, these two neighborhoods are considered unsafe and vulnerable. Households with adolescents of 12 to 18 years old were purposively selected. The research defines "parents" as fathers, mothers, or any adults living in the same house with adolescents and available to participate in interviews. Adolescents are identified as boys and girls of 12 to 18 years of age who are still in school, have dropped out of school, have already worked, or are unemployed. Informed consents were obtained from all participants involved.
Parents and adolescents were interviewed in separate places, using a questionnaire developed by the researcher by referring to the previous report conducted by Kusumawardani and Suhardi (2011) on behavioral health risk among adolescents in West Java, Indonesia including smoking, unhealthy behavior, and physical activities. The parental questionnaire obtained information related to educational background, employment, income, and household expenditure. The questionnaire for adolescents related to eight potentially risky behaviors (smoking, consuming alcohol, substance use, brawling, criminal activity, physical fighting, heavy petting, and premarital sex). Expected responses were "yes" or "no." In the adolescent questionnaire, 20 items addressed adolescents' perception of their relationship with their parents. Respondents chose from three alternative answers, that is, "often," "sometimes," and "never." The questionnaire also included items on reasons for taking or not taking risky behaviors.
Two hypotheses were tested in this study. First, whether there is different proportion between boys and girls in related to eight potential risky behaviors (smoking, consuming alcohol, substance use, brawling, criminal activity, physical fighting, heavy petting and premarital sex) (H0: pgirls=pboys against the alternative hypotheses Ha: pgirls≠pboys.). Were boys more likely to be involved in eight potential risky behaviors than girls? The second hypotheses was whether there is different proportion between adolescents age 13-15.9 years and adolescents age 16.18 years in their involvement to eight potential risky Ajisuksmo -Adolescents' risk-taking behavior behaviors (Ho:p13-15.9y=p16-18y against the alternative hypotheses Ha:p13-15y≠p16-18y). Were older adolescents i.e. age 16-18 years more likely to be engaged in eight potential risky behaviors than younger adolescents i.e. age 13-15 years old?

Parents' Characteristics
Parents' educational backgrounds were as follows: More than half (63.4%) had a low level of education (35.7% elementary; 27.7% junior high), from elementary school, grade 1 to junior high school, grade 9. One-third of parents (33.4%) had graduated from senior high school and 2.7% from institutions of higher learning (0.5% did not answer). More than half of parents had been able to fulfill the basic requirement of 9 y compulsory education.
Work and family life were closely related to educational level. The higher the education, the better the jobs and quality of life or well-being. With regard to parents' employment, Table 1 indicates that most parents were informal workers in various sectors (40.6%; n = 163) (e.g., motorbike taxi driver, truck driver, fishing laborers, building construction workers), and some were unemployed (6%; n = 24). Data on parents' employment relates to data on their educational background: Lack of knowledge and skills needed for high levels of employment might have been the reason most parents worked in informal sectors with low wages.
Data indicated that parents were categorized at low socioeconomic levels (Table 1). In previous studies, Ponnett (2014) and Crandall, Magnusson, Novilla, Novilla, and Dyer (2017) reported that family financial problems influence adolescents' behavior and often lead to parental conflict because parents cannot meet the family's needs. Previous studies also reported that serious criminal cases often occur in low-income and low-resource communities, and these circumstances influence adolescents' antisocial behavior (Cook et al., 2009;Tjora, Hetland, Aarø, & Øverland, 2011;Elliott, Avery, Fishman, & Hoshiko, 2002;Djerboua, Chen, & Davison, 2016).  Table 2 characterizes relationships of adolescents and their parents as perceived by the adolescents. Even though numbers are small, some adolescents perceived their relationship with their parents as not strong. Some felt that their parents never supported or motivated them, never paid attention or listened to them, and never praised them. Some mentioned that their parents never fulfilled their needs, never spent time with them, and never provided advice they needed. Table 2 also shows that some adolescents reported that they were scolded with abusive words, often hit or beaten with or without tools, or abused by their parents. Contreras and Cano (2014) stated that communication among family members is the essential dynamic of family relations, and lack of communication is associated with antisocial behaviors. Open communication relates to a democratic parenting style that shows warmth and relates positively to affection. In contrast, problematic communications are related to the authoritarian parenting style and positively related to criticism and rigid manners of setting rules.  Table 3 indicates that age correlates with adolescents' risky behaviors. Risky behaviors were performed mostly by older adolescents from 13 to 18 years. Alarmingly, some children aged 13 to 15 already smoked (6%), drank alcohol (1.4%), and engaged in brawls (2.8%), in physical fights (5.5%), and in heavy petting (0.5%). Table 3 also shows that smoking (5.6%), physical fights (4.6%), and brawling (2.2%) were the riskiest behaviors in which adolescents were involved. Although in relatively small numbers, some adolescents also engaged in consuming alcohol (1%), substance use (0.2%), heavy petting (0.5%), and premarital sex (0.2%).  Table 4 shows adolescents' risk-taking behaviors across genders. Boys were more likely than girls to be involved in risky behaviors (e.g., smoking, drinking alcohol, brawling, and physical fighting). However, even though the number is small, some female participants were also involved in drinking alcohol (0.4%), substance use (0.4%), physical fighting (0.8%), and heavy petting (0.4%). According to Idemudia and Sekano (2015), gender and age are important factors that determine probability of risk-taking behaviors. Erol and Orth (in Idemudia & Sekano, 2015) reported that adolescent boys showed higher levels of risk-taking behaviors than girls. Djerboua et al. (2016) also stated that physical fighting related to injury was more frequent in males than in females. However, Idemudia and Sekano (2015) found no significant gender differences in risk-taking behavior. Meanwhile, Schulte, Ramo, and Brown (2009) identified factors that influenced alcohol-drinking behavior in adolescence and continued into adulthood. Their study's result indicated that certain biological and psychosocial factors appear to impact boys and girls similarly. However, as adolescents shifted into adulthood, biological and psychosocial impact appeared to differ between males and females. Sabri et al. (2017) assessed gender differences regarding two factors of HIV transmission, multiple sexual partners and sharing needles. Results indicated that more men than women had recent multiple sex partners and shared needles. For physical fighting, Djerboua et al. (2016) reported that males were more often involved than females.

Adolescents' Engagement in Risky Behaviors
In this study the comparison of proportions of eight risky behavior, namely smoking, drinking alcohol, drug use, fighting, criminal acts, physical fights, heavy petting, and premarital sex engaged by adolescents was tested by different test procedures. A Z test was used to measure the equality of proportions among the population, a Chi-Square test was performed for testing the independence of risky behavior, and Marascuillo procedure was applied to provide the magnitude of variation in the pairs of proportions.
The two hypotheses, comparison of proportions of eight risky behaviors between girls and boys population as well as between adolescents age 13-15.9 and 16-18 years population were tested by Z test. For the gender population the hypotheses tested was H0: pgirls=pboys against the alternative hypotheses Ha: pgirls≠pboys. The Z test was applied to test the hypotheses of each risky behavior between girls and boys. For the age population the hypotheses tested was Ho:p13-15.9y=p16-18y against the alternative hypotheses Ha: p13-15.9y≠p16-18y. The Z test was applied to test the hypotheses of each risky behavior between adolescents' age 13-15.9 years and 16-18 years. This study did not include the population of age 12-12.9 y since they did not engage in risky behavior. The hypotheses tested in the Z test was performed with the following formula.
Based on gender population, the result of Z test showed for smoking Z= 5.294 with p value = 0.00, brawling Z = 3.497 with p value = 0.00, and physical fight Z= 4.273 with p value = 0.00. Small Z value showed for other five risky behaviors, namely drinking alcohol, substance use, criminal activity, heavy petting and premarital sex. The result of this Z test indicated that the proportions of three risky behaviors, namely smoking, brawling and physical fight are not equal between girls and boys population. Meanwhile, based on age population, the result of Z test showed that all Z-values are small, less than 1.96 and all p values are more than 0.05. This study used α = 0.05 level of significant which means that the proportions of eight risky behaviors between age 13-15.9 and age 16-18 years are equal. There was no difference between adolescents age 13-15.9 and 16-18 years in engaging with eight risky behaviors (see table 3) The Chi-Square test was performed to test the hypotheses Ho:p1=p2=….p7=p8 against the alternative hypotheses that not all eight risky behaviors proportions are equal, Ha=not all pi are equal (i=1, 2,3….8). The result of Chi-Square test was 2 = 77.016 with the p value = 0.00. This study choose 0.05 level of significant ( 2 = 77.016; p<0.05), which can be concluded that not all proportions of eight risky behaviors engaged by adolescents are equal.
In order to identify the equality of the proportion of risky behaviors, a multiple comparisons procedures known as Marascuillo procedure was performed. The result showed that there is difference on the proportion of risky behaviors that adolescents engaged in, namely between smoking vs substance use, smoking vs criminal activity, smoking vs heavy petting, smoking vs premarital sex, substance use vs physical fights, criminal activity vs physical fights, physical fights vs heavy petting, and physical fights vs premarital sex (see table 5)

Reasons to Engage in Risky Behaviors
Knowledge about reasons for risky behaviors is very important. In this study, seven factors influenced adolescents to engage in risky behaviors: (1) pressure from friends and (2) family, (3) desire after watching a film, (4) after viewing the Internet, (5) after reading a book or magazine, and (6) after seeing something directly, and (7) wanting to try something or wanting to know more about something. Table 6 indicates that wanting to try or wanting to know (n = 21; 50%) and pressure from friends (n = 17; 40.5%) were the most frequent reasons adolescents engage in risky behaviors.

Reasons Not to Engage in Risky Behaviors
Adolescents reported eight reasons not to engage in risky behaviors: (1) being advised by family; (2) by religious leaders; (3) by other adults, such as teachers, community leaders, etc.; or 4) by friends or peer groups; 5) fear of God for committing sins; 6) of being scolded by parents; 7) of destroying the future; and 8) of harm; and (9) others. Table 7 indicates that being advised by family (68%) was higher than being advised by teachers and community leaders (23.9%), religious leaders (20.7%), and friends or peers (14.5%). Data suggests that families play an important role in influencing adolescents not to engage in risky behaviors. This follows Morales-Alema (2011) who indicated that parental involvement is a significant buffer for high-risk sexual behavior.

Discussion and Conclusion
This study was conducted in vulnerable areas of North Jakarta, where more parents have low educational and employment backgrounds as well as lowincome and expenditure levels.
Previous studies reported that family income influences development of adolescents' behavior, that is, financial problems lead to adolescents' maladaptive coping due to the family's inability to meet their basic needs (Ponett, 2014;Crandall et al., 2017). Adolescents from low-income families take risky behaviors as a form of maladaptive coping with family financial stress. Besides that, most low-income families live in communities with high crime rates, high poverty, and few resources; antisocial behaviors are perceived as normative (Cook et al., 2009). Djerboua et al. (2016) reported that family prosperity influences adolescents' risk-taking behavior. Adolescents from families with low socioeconomic levels have higher risk of being involved in physical fights and having related injuries. Tjora et al. (2011) reported that parents' socioeconomic status was significantly associated, directly and indirectly, with adolescents' initiation and development of smoking behavior. Conversely, high socioeconomic status had direct negative association with adolescents' smoking behavior.
A study conducted by Jang & Johnson (2010) showed that adolescents imitate the behavior of those around them, especially their parents when they see their parents smoke and use drugs. Parents directly show their adolescents that smoking and using drugs is part of an acceptable lifestyle. Additionally, Elliott et al. (2002) reported that experiencing and witnessing family violence contributed to risky sexual behavior among young female adolescents. Meanwhile, the study of Herrera & McCloskey (2003) indicated that childhood sexual abuse appeared to be the powerful predictor of girls' criminal behavior. Study by Iverson, Jimenez, Harrington, & Resick (2011) reported that exposure to family violence during childhood, including childhood physical abuse, childhood sexual abuse, and witnessing parental violence contribute to the risk for victimization of intimate partner violence for both male and female. Ccommunities at lower socioeconomic levels have low level of education and do not have enough knowledge about parenting and childcare. Therefore, children in poor communities are often exposed to family violence.
The present study also showed that fear of God for committing sins, fear of parents, and fear of damaging the future were reasons adolescents did not engage in risky behaviors. This study supports Landor et al. (2011) in that a negative relationship exists between religious beliefs and risky sexual behavior; parental religiosity affects authoritative parenting and youth religiosity. The present study also supports the findings of previous studies (Marsiglia, et.al., 2012;Arli et.al., 2016;Ameri, et.al., 2017) that religion and religiosity can directly and indirectly influence adolescents' perceptions of risky behavior and the eagerness to engage in risky behavior. Thus, it is very important to integrate religious aspects while developing intervention programs for adolescents to prevent them from engaging in risky behavior. Despite peers' influence on sexual behavior, the parental role was found strong enough to facilitate adolescents in deciding not to engage in risky sexual behavior. Parents play an important role in protecting their adolescents from risky activities. They play a role in preparing their adolescents to be responsible for the various decisions that they make, including in their sexual behavior. Communication with and supervision of adolescents by their parents greatly helps adolescents to not feel alone in facing challenges in their development.
The present study indicated that even though the percentage is small, less than 10% of adolescents living in two vulnerable areas of North Jakarta engage in risky behaviors, namely, smoking, consuming alcohol, substance use, brawling, crime, physical fighting, heavy petting, and premarital sex. However, this study showed that there is difference on the proportion of risky behaviors that adolescents engaged in.
The study also revealed that some parent-adolescent relationships are not strong, thus likely influencing adolescents' engagement in risky behaviors. Fear of God because of sins, fear of parents, and fear of damaging the future were the most common reasons for adolescents not to involve themselves in risky behaviors. Intervention programs intended for young people, therefore, will also be relevant for communities at a lower socioeconomic level.