Goal oriented and risk taking behavior: The roles of multiple systems for Caucasian and Arab-American adolescents

Abstract

With Bronfenbrenner’s (1977) ecological theory and other multifactor models (e.g. Pianta, 1999; Prinstein, Boergers, & Spirito, 2001) underlying this study design, the purpose was to examine, simultaneously, key variables in multiple life contexts (microsystem, mesosystem, exosystem levels) for their individual and combined roles in predicting adolescent risk-taking and goal-oriented behaviors. Predictor variables were parenting behaviors (parenting style, monitoring, and involvement), the risk-taking and goal-oriented behavior of peers, and religiosity (attendance, involvement, and importance). General media consumption was also explored. The participants in this study were 272 9th to 12th grade Caucasian and Arab-American high school students (124 males and 148 females) from a suburban public school district in the midwestern United States (Mean age = 15.64). Results revealed several themes, including that peers appeared to have the primary role in explaining variance in risk behaviors, while parents have the primary role in explaining goal-oriented behavior. Religiosity contributed minimally. There were several noteworthy differences between the two cultural groups regarding which factors explained the most variance in criterion variables. Results are explored in more detail and implications for prevention and intervention are discussed.

Keywords

Adolescence risk behavior family goal oriented behavior school

Adolescence is a key time for the development of future plans and goals. Cognitive development allows expansion of the ability to think through what the future holds. However, adolescence has also been well documented as a time in which individuals display the greatest amount of risk-taking behaviors. US national research (2008) has indicated that adolescents partake in astounding amounts of risk-taking behaviors. For example, 47.8% of all high school students had engaged in sex at the time they were surveyed; 26% reported heavy drinking; 75% reported having had at least one alcoholic drink in their lifetime; 50.3% had indicated that they had tried cigarettes, and 35.5% had indicated that they had been in a physical fight in the past 12 months. These behaviors can lead to serious illness, injury, or death, and clearly need continued study. On the other hand, goal-oriented behaviors have been less well studied and have been selected for this study to identify positive outcomes in adolescents, which will be reviewed throughout this article. For the purpose of the current study, goal-oriented behaviors are described by academic achievement (e.g. grades) and involvement in after school activities (e.g. sports, academic enrichment clubs, and community activities such as church, community service, etc.). The overall goal of this study was to better understand which factors and combinations of factors most significantly predict both risk-taking behaviors and goal-oriented behaviors.

Research has been extensive in attempting to identify links between a variety of factors and risk-taking behaviors. These factors have ranged from biological (e.g. onset of puberty and brain development; Price, 2005) to peers and family (Jaccard, Blanton, & Dodge, 2005; Martin & Martin, 2000) to political trauma (Schiff & Fang, 2014). There is less research on predictors of goal-oriented behaviors and even less specifically on Arabic youth samples. Past research on predictors of both adolescent risk-taking behaviors and goal-oriented behaviors has primarily isolated only one or a few factors at a time in an attempt to explain variance in outcomes. However, individuals are continuously interacting with a set of complex social relationships and within numerous life contexts. According to Brofenbrenner (1989), the attributes of the person at a given time in his or her life are actually a combined function of the person’s attributes and the environment over the course of that person’s life up to that time. Bronfenbrenner’s (1977, 1979) Ecological Perspective of Human Development has four ecological systems, constituting interdependent, active structures, ranging from proximal processes, such as a mother–child interaction in the family setting (microsystem), to the more distal influences comprising broader social categories, such as government and culture (macrosystem), with the meso and exosystems in between. The nature of these systems is interactive; their influence operates in a reciprocal pattern. As adolescents age, the exosystem and the macrosystem become increasingly important (Muss, 1996). The microsystem, the mesosystem, and the exosystem are the three structures that will be primarily focused upon in the current study.

Other researchers have also focused on multiple life contexts, for example, Pianta emphasizes the school context (1999), and some have explicitly focused on risk-taking behaviors in developing multi-factor models. For example, Prinstein et al. (2001) discussed a cumulative risk factor model on the effects of aggregated risk factors for adolescents. Their results suggested that the number of adolescents’ risk factors, including peers’ risk behaviors, family dysfunction, social acceptance, and depression, was strongly associated with the likelihood that adolescents engaged in risk-taking behaviors. Also, Ben-Arieh and Attar-Schwartz (2013) used an ecological perspective to examine Middle-Eastern adolescents’ views on their rights and ability to participate in four different ecological contexts (family, school, community, and the larger sociopolitical system). Clearly, a contextual model is an important guide for studying risk-taking and other behaviors in adolescents.goal-oriented behaviors.

The role of family in goal-oriented and health compromising behaviors

The family system is the most proximal influence on adolescent development and thus is an important microsystem level factor to consider. Several factors have been implicated in adolescents’ behavior. The most prominent of these factors include parental monitoring, parental involvement, and parental communication with the adolescent.

Parental monitoring

Research has indicated the need for parents to monitor their adolescents’ interactions with peers. Laird, Criss, Pettit, Dodge, and Bates (2008), reported a correlation between decreased levels of parents’ knowledge of their child’s peers and whereabouts and increased levels of risk behaviors and increased number of friends who are involved in risky behaviors. Henry, Merten, Plunkett, and Sands (2008) reported that Latino adolescents from immigrant families who saw their parents as having greater knowledge of their friends, whereabouts, and activities reported greater academic motivation and, in turn, showed higher grade point averages (GPA). Similarly, Annunziata, Hogue, Faw, and Liddle (2006) reported that when parental monitoring was high, within at-risk African American families, adolescents were more engaged in academics. No clearly parallel studies were found at this time for Middle-Eastern adolescents.

Parental involvement

Parent involvement (commonly defined as helping with homework and involvement in extracurricular activities) is another important characteristic in looking at the parent–adolescent relationship and how that relationship may discourage risk-taking behaviors and support goal-oriented behaviors. Suldo, Mihalas, Powell, and French (2008) reported that increased parental involvement was positively associated with decreased adolescent substance use. Parental involvement in their adolescents’ schooling has not only been associated with lower amounts of risk-taking behaviors but has also been associated with increased amounts of school engagement (Annunziata et al., 2006) and academic competence (Steinberg, Blatt-Eisengart, & Cauffman, 2006). Increased family involvement has been positively linked to school engagement (Taylor & Lopez, 2005) and school engagement has been positively linked to lower levels of risk-taking behaviors (Connell, Halpern-Felsher, Clifford, Crichlow, & Usinger, 1995). No clear parallel studies were found at this time for Middle-Eastern adolescents.

Parent communication

Research has indicated that parent–adolescent communication about life events is an important factor for adolescent development. Somers and Paulson (2000) reported that increased parent communication about sexuality has been indicated as a predictor of decreased risky behaviors such as sexual risk-taking behaviors. Research has also indicated that adolescents engage in less sexually risky behaviors when their parents, who they perceive to have a certain level of expertise and trustworthiness, discuss the social and moral consequences of early sexual activity, as opposed to solely communicating the consequences of pregnancy and STDs (Guilamos-Ramos, Jaccard, Dittus, & Bouris, 2006). It is important to understand the relations between parental communication and adolescents’ decision-making processes. Research on parental communication and adolescent academic achievement is needed. No clearly parallel studies were found at this time for Middle-Eastern or Arab-American adolescents goal-oriented behaviors.

The role of peers in health compromising and goal-oriented behaviors

Peers have frequently been linked to being a powerful predictor of adolescent risk behaviors. One author has gone as far as indicating that 50% of the variance in adolescent personality is genetic in origin and the remaining 50% primarily reflects the influence of peers (Harris, 1998). Research has indicated that adolescents with best friends involved in risk-taking behaviors such as smoking, drinking, drug use (Bahr, Hoffman, & Yang, 2005), and deviant behaviors (Prinstein et al., 2001) are more likely to engage in risk-taking behaviors than individuals who do not have best friends involved in risk-taking behaviors. However, Spoth, Redmond, Hockaday, and Yoo (1996) have indicated that adolescents’ affiliation with goal-oriented peers is predictive of abstinence from alcohol use. Similarly, Prinstein et al. (2001) indicated that adolescents with high proportions of friends who engaged in goal-oriented behaviors (e.g. assisting troubled teens, involvement in school activities, etc.) were less likely to engage in violent and substance use behaviors themselves.

Goal-oriented peers are not only associated with decreased delinquent behaviors but are also associated with increased academic success. LeCroy and Krysik (2008) reported a positive association with pro-academic peers and higher grade point average and greater attachment to school in Hispanic adolescents. Not only do peers with goal-oriented friends display increased attachment to school but also adolescents who engage in goal-oriented behaviors (e.g. academic clubs, sports, etc.) tend to seek out other goal-oriented peers whereas adolescents who do not participate in goal-oriented behaviors do not display the same support seeking behaviors (Fredricks & Eccles, 2005). No studies explicitly on Middle-Eastern or Arab-American students were found at this time of this literature review.

The role of religion in goal-oriented and health compromising behaviors

Religion is another key socialization agent for many adolescents (Wallace & Williams, 1997). Direct involvement in religion at a mesosystem level is of focus in the current proposal. King and Furrow (2004) reported that religious involvement might aid in positive development through the potentially increased social resources (e.g. interested adults) available to adolescents. Rostosky, Regnerus, and Wright (2003) reported that religiosity (e.g. church attendance, attendance of church youth activities, and the overall view of religious importance) indirectly affects sexual initiation through one’s sexual ideology or belief system founded on anticipated negative consequences from engaging in sexual intercourse.

Research on religion has indicated that many adolescents believe in God and are actively involved in some type of church affiliation. Religious beliefs have been affiliated with several pro-social behaviors including increased academic achievement (Regnerus & Elder, 2003). Sinha, Cnaan, and Gelles (2007) reported that adolescents who are more religious are less likely than their peers to engage in risk-taking behaviors (e.g. carrying weapons, getting into fights, drinking and driving, and drug use). Involvement in religion has also been linked to increased academic achievement in a sample of emerging adult black males (Byfield, 2008). Specifically, of those who are academically successful, living both in the UK and the US, a common factor was that most of the men were religious and believed in God. Religiosity has also been positively related with decreased maladjustment in Indonesian adolescents (Sallquist, Eisenberg, & French, 2010).

The role of mass media in goal-oriented and health compromising behaviors

According to Arnett (1995), media has many uses for adolescents such as entertainment, identity formation, coping, and culture identification. Media can also be a means for adolescents to disengage from stress, anxiety, and negative emotions (Larson, 1995). Brown and Witherspoon (2002) reported that 8- to 18-year-olds spend an average 6–7 hours a day with some form of mass media, whether it is television, music, magazines, or the Internet. Adolescents use media for many reasons but with this much time devoted to media consumption it is important to understand how media may be related to adolescents’ goal-oriented and risk-taking behaviors.

Adolescent media consumption has been linked with several risk-taking behaviors such as greater sexual experience ( Somers & Tynan, 2006), obesity (Vandewater, Shim, & Caplovitz, 2004), delinquency (Kremar & Greene, 2000), and adolescent females’ poorer self-perceptions (Botte, 2000). With these findings on sexual activity, obesity, violence, etc., it is important to understand to what extent the mass media is related to adolescents’ goal-oriented and risk-taking behaviors. While the mass media as a whole (music, movies/television, magazines, and Internet) has been associated with risk-taking behaviors there is a need for research on the relations between media exposure and goal-oriented behaviors. These studies tend to be with Caucasian, middle class, American students, and none were found with Middle Eastern or Arab-American students.

Summary and limitations of past research

Many studies that have examined individual relationships between various life contexts (e.g. family, peers, religion, and media) and both risk-taking and goal-oriented behavior have been noted throughout this article. However, it is more likely that a larger combination of factors need simultaneous exploration in order to better understand variance in risk-taking and goal-oriented outcomes. Indeed, as described earlier, there have been some studies that have utilized multi-factor models and hypotheses. However, the context and variables in this study extend on those in past studies. While past research has touched on various individual and some combinations of components that are likely related to adolescent risk-taking and goal-oriented behaviors, none have encompassed the multiple variables that this study will explore. Additionally, risk-taking behaviors tend to be of greater focus and research is needed that also considers goal-oriented outcome behaviors. Importantly, most prior studies have been with Caucasian students and one of the goals of this study was to examine these variables also in an Arab-American subsample in order to build the literature on these associations for Arab-American youth as well.

Purpose of the current study

Based on the aforementioned literature review and limitations of prior research, the purpose of the current study was to explore several key contexts that adolescents are concurrently exposed to, including peer relationships, family relationships, religion, and media. These contexts are part of the various systems of Bronfenbrenner’s bioecological theory (2005). Ultimately, the purpose of this study was to study several key contextual variables reflecting major influences on adolescent behavior, and examine their individual and combined contributions to both risk-taking and goal-oriented adolescent behavior. Additionally, two different adolescent groups were studied to determine similarities and differences. Specifically, no research was identified during the time of this study on patterns for Caucasian and Arab adolescents among these variables of interest. Thus, both groups were included to explore their similarities and differences. This study was expected to answer questions that will significantly add to the current literature base.

Method

Participants

Participants were 272 adolescents from a sampling of required classes in a 9th to 12th grade suburban public high school in the Midwestern United States. They ranged from 14- to 19-years-old (Mean = 15.64, SD = 1.21), with 148 females and 124 males. Participants were either Caucasian (N = 133) or Middle Eastern and/or Arab (N = 139).

Measures

A demographic measure was created for the purposes of this study. It included questions about age, gender, and school grade. All other variables are described in the following sections.

Risk-taking behavior

Two subscales of the Adolescent Risk Questionnaire (ARQ; Gullone & Moore, 2000; Gullone, Moore, Moss, & Boyd, 2000) were used: 1) Rebellious behaviors (five items, i.e. taking drugs); 2) Reckless behaviors (five items, i.e. having unprotected sex). Adolescents were asked to rate the frequency that they engage in the particular behavior. Items were rated on a five-point Likert scale, from never engaging in the risk behavior (1) to engaging in the behavior very often (5). To determine each adolescent’s risk-taking behavior an average composite mean score was calculated. Cronbach’s alphas have exceeded 0.8 and test–retest reliability was 0.78 in the research by Gullone and colleagues. They also established good convergent and divergent validity for their measure. Cronbach’s alphas for the current sample were 0.84 for the Rebellious subscale and 0.71 for the Reckless subscale.

Goal-oriented behavior

This was measured by both academic performance in school and overall involvement in extracurricular activities. School achievement was assessed using self-reported GPA. Adolescents were asked to assess their GPA by answering the question, ‘What grades do you most often receive?’, with the following response options: Mostly As, Mostly As and Bs, Mostly Bs, Mostly Bs and Cs, Mostly Cs, Mostly Cs and Ds, Mostly Ds, Mostly Ds and Es, or Mostly Es. The letter grades were coded as 1 (mostly Es) through 9 (mostly As). In addition, students were asked to provide their most recent grades in each of the core academic areas: English Language Arts, History/Social Studies, Mathematics, and Science. From those, a GPA was calculated. The two measures were compared for consistency. The GPA for overall academic performance was used in analyses. Students also listed all extracurricular activities they were involved in. Overall involvement in extracurricular activities was computed as a tally of number of activities listed.

Parenting variables

Parental communication, parental involvement, and parental monitoring were each measured. The Parental Involvement Scale (Paulson, 1994) consists of a total of 22 items, e.g. ‘My mother/father usually goes to parent-teacher conferences’. Three subscales comprise the total involvement scale: Values towards achievement, interest in schoolwork, and involvement in school functions. Responses across all three scales are summed to create a total involvement score. Adolescents were asked to estimate the frequency with which their parents engage in each particular behavior. Adolescents responded to the instruments twice, once for each parent. Items are rated on a five-point Likert scale, from ‘very unlikely’ that they would engage in the particular behavior (1) to ‘very likely’ that they would engage in the particular behavior (5). Adolescents were instructed to respond about a mother/mother-figure and a father/father-figure. Participants were instructed to speak to the researcher if they had any questions regarding who qualifies as a parent figure (e.g. step-parent, guardian, etc.). Responses were then averaged between mothers and fathers to create a total parental involvement scale. Paulson found Cronbach’s alpha typically exceeded 0.7 on the adolescents’ reports of maternal and paternal parenting with the exception of the School Functions subscale within the Parental Involvement Scale (α = 0.67). The measure has also been shown to have good construct and convergent validity. The parent involvement alpha for the current sample was 0.78 for mothers and 0.85 for fathers.

Parental communication was assessed using the Parent-Child Relationship Survey (PCRS) developed by Fine, Moreland, and Schewebel (1983). The PCRS is a 24-item instrument designed to measure perceptions of their parent–child relationship. The PCRS comes in two forms, one for assessing the individual’s relationship with the mother and the other for assessing the individual’s relationship with the father. For the purpose of the current study the subscales of interest were: 1) Father communication (N = five items, e.g. ‘How comfortable would you be approaching your father about a romantic relationship?’); and 2) Mother communication (N = seven items, e.g. ‘How confident are you that your mother would not ridicule or make fun of you if you were to talk about a problem?’). Five items on the mother and father communication scales overlap. The two additional items on the mother communication scale are ‘How confident are you that your mother would not ridicule or make fun of you if you were to talk about a problem?’ and ‘How confident are you that your mother would help you when you have a problem?’. Items are rated on a seven-point Likert scale ranging from ‘not at all’ (1) to ‘extremely’ (7). A total mean score was calculated to determine the amount of communication between the adolescents and each of their parents. From there, a total score across both parents was computed. The PCRS has strong internal consistency (α = 0.61 and 0.88–0.94) by Fine et al. (1983). Alphas for the current sample were 0.86 for mothers and 0.89 for fathers.

Finally, The Parental Monitoring Instrument (PMI) (Cottrell et al., 2007) was used to assess the amount and type of monitoring adolescents perceive that their parents do. The PMI was developed to assess how frequently parents employ a variety of specific monitoring strategies. The PMI consists of 27 items from seven subscales. Those subscales are: Indirect monitoring, Direct monitoring, School monitoring, Health monitoring, Computer monitoring, Phone monitoring, and Restrictive monitoring. The adolescent is asked to indicate the number of times in the past four months their parents participated in a certain act (e.g. ‘How many times in the last four months has your parent asked to meet your friends?’). Items are rated on a four-point Likert, with 1 (0 times), 2 (1–2 times), 3 (3–4 times), and 4 (5-plus times). Response options are 0 times, 1–2 times, etc.). A total composite was calculated and used for analyses. Students responded across both parents collectively. Cottrell et al. (2007) found Cronbach’s alphas between 0.69 and 0.80 for the various subscales, as well as good construct validity Structural Equations Modeling (SEM). For the current sample, the Cronbach’s α was 0.90.

Peer variables

The Family, Friends, and Self Form (FFS) was used to measure adolescents’ perceptions of their peer’s behaviors (Simpson & McBride, 1992). The complete FFS consists of 60 items with three parts and numerous subscales. For the purpose of the current study the subscales to be included were: 1) Friends ‘conventional involvement’ (N = seven items, e.g. ‘How many of your friends do homework after school or at night?’) (which was labeled ‘peer goal-orientation’ for the purposes of this study); and 2) friends ‘trouble’ (N = seven items, e.g. ‘How many of your friends have been in trouble with the police because of alcohol or drugs?’) (labeled ‘peer risk behavior’ for the purposes of this study). Adolescents were asked to rate and to estimate how many of their peers engage in each particular behavior. Items are rated on a five-point Likert scale, from ‘none engaging in the behavior’ (1) to ‘all engaging in the behavior’ (5). Each subscale was summed for analyses. Simpson and McBride (1992) found acceptable reliability of the FFS, with coefficient alpha reliability between 0.73 and 0.86, 0.82, and solid construct validity via factor analysis. In the current sample, alphas were 0.65 for conventional involvement and 0.88 for trouble.

Religiosity

Religious involvement was assessed using a religiosity scale created by Rostosky, Regnerus, and Wright (2003) based on a National Longitudinal Survey of Adolescent Health sample of 3,691 adolescents. The religiosity scale is a three-item instrument designed to measure the frequency of attendance at religious services, frequency of attendance at religious youth activities, and self-rated importance of religion. The religiosity assessment contains three questions, the first two reflect religious involvement: 1) ‘How often have you attended church/synagogue/mosque/religious services in the past 12 months?’; and 2) ‘Many churches, synagogues, mosques and other places of worship have special activities for young people, such as Bible classes, retreats, youth groups, or choir. In the past 12 months, how often have you taken part in such activities?’. Items are rated on a seven-point Likert scale ranging from ‘never’ (0) to ‘more than one a week’ (6). A total mean score is calculated to determine the amount of religious involvement. The third question is: 3) How important is your religious faith to you? This item is rated on a four-point Likert scale ranging from ‘not important’ (0) to ‘more important than anything else’ (3). The first two items are later collapsed into a four-point Likert scale (0 = 0; 1or 2 = 1; 3 or 4 = 2; 5 or 6 = 3). These items were shown to have adequate internal consistency by Rostoskey et al. (2003) (α = 0.69–0.70) and Rostosky, Danner, and Riggle (2008) (α = 0.71–0.80). These three questions were summed to form a composite ‘religiosity’ score. Internal consistency for the three items in the current sample was alpha = 0.69.

Media consumption

This measure was exploratory and surveyedamounts of time spent with each of four mediums (TV, internet, music, and magazines), which were then totaled and labeled ‘all media’. There was also an item that asked how much time adolescents spend using any type of media that they would not want their parents to know that they were using. This was labeled, for convenience, ‘negative media’.

Procedure

Instruments were completed under researcher supervision at school. Full HIC approval was granted. Parental consent and adolescent assent were both obtained after an explanation that participation was strictly voluntary, and that their choice to participate or not had no influence on their grade in the class, and that their teachers would not be privy to individual participation. It was explained that all information was anonymous and they were informed not to write their name anywhere on the questionnaires. A total of 46 students or their parents chose not to participate.

Results

The outcome variables of interest in this study were risk-taking behaviors (e.g. unprotected sexual activity, smoking, drinking, truancy, speeding, etc.) and goal-oriented behaviors (e.g. academic achievement, attendance, involvement in teams and clubs, etc.). Environmental variables were family relationships and peer relationships (conceptualized here at the microsystem and mesosystem levels) and religion and media exposure (conceptualized here at the exosystem level).

Preliminary analyses

Analyses of Variance (ANOVA) revealed significantdifferences between the two primary cultural groups included (Caucasian and Arab American) on many of the independent and dependent variables. We were interested in these patterns of differences. Thus, rather than statistically controlling for ethnic group, the primary analyses were run separately for each group. See Tables 1 and 2 for descriptive statistics.

Table 1.

Caucasian descriptive statistics.

Independent variables	Mean	SD	Range
Independent variables	Mean	SD	Minimum	Maximum
Independent variables
Parental Monitoring	2.12	0.53	1.00	3.30
Mother Demandingness	3.27	0.66	1.27	4.87
Father Demandingness	3.07	0.76	1.00	4.93
Mother Responsiveness	3.31	0.70	1.27	4.80
Father Responsiveness	3.11	0.69	1.33	4.80
Mother Involvement	3.90	0.62	1.77	4.82
Father Involvement	3.68	0.65	2.09	4.68
Mother Communication	4.92	1.43	1.43	7.00
Father Communication	4.04	1.63	1.00	7.00
Goal-Oriented Peers	2.29	0.62	.57	4.00
Risk-Taking Peers	1.00	0.80	.00	4.00
Religious Attendance	1.88	1.63	0.00	5.00
Religious Involvement	1.24	1.60	0.00	5.00
Religious Importance	1.37	0.97	0.00	3.00
Hours of Television Viewed Per Week	13.09	17.69	0.00	102.00
Hours Spent Listening to Music Per Week	30.11	37.72	0.00	200.00
Hours Spent Reading Magazines Per Week	2.03	5.58	0.00	50.00
Hours Spent on the Internet Per Week	23.34	29.80	0.00	150.00
Total Hours Per Week With Media	54.01	52.21	0.00	211.00
Approx. Hours Per Week With Negative Media	12.36	29.68	0.00	168.00
Dependent variables
Thrillseeking Behavior	1.58	0.56	1.00	4.86
Rebellious Behavior	1.97	1.01	1.00	5.00
Reckless Behavior	1.35	0.51	1.00	3.40
Antisocial Behavior	1.92	0.67	1.00	4.20
Average Risk-Taking Behavior	1.69	0.52	1.00	3.55
Extracurricular Activities	1.92	1.58	0.00	6.00
Overall Grades	2.83	1.60	1.00	7.00

Note: Possible ranges: Parental Monitoring – 1 to 4; Mother and Father Demandingness – 1 to 5; Mother and Father Responsiveness – 1 to 4; Mother and Father Responsiveness – 1 to 5; Mother and Father Involvement – 1 to 5; Mother and Father Communication – 1 to 7; Goal-Oriented and Risk-Taking Peers – 1 to 4; Television, Music, Magazines, Internet, Media Total, and Negative Media were self-reported hours; Extracurricular Activities were self-reported totals; Overall Grades – 1 to 9.

Table 2.

Arab American descriptive statistics.

Independent variables	Mean	SD	Range
Independent variables	Mean	SD	Minimum	Maximum
Independent variables
Parental Monitoring	2.24	0.67	1.00	4.00
Mother Demandingness	3.3	0.52	1.93	4.8
Father Demandingness	3.13	0.73	1.53	4.8
Mother Responsiveness	3.21	0.6	1.87	4.53
Father Responsiveness	3.12	0.59	1.67	4.47
Mother Involvement	3.76	0.62	2.23	4.86
Father Involvement	3.5	0.66	1.77	4.59
Mother Communication	5.01	1.41	1.43	7.0
Father Communication	4.09	1.82	1.00	7.0
Goal-Oriented Peers	2.32	0.61	0.57	3.71
Risk-Taking Peers	1.2	0.95	0.00	3.57
Religious Attendance	2.39	1.68	0.00	5.0
Religious Involvement	1.49	1.61	0.00	5.0
Religious Importance	2.29	0.81	0.00	3.0
Hours of Television Viewed Per Week	12.95	19.6	0.00	130.00
Hours Spent Listening to Music Per Week	15.6	26.39	0.00	135.00
Hours Spent Reading Magazines Per Week	1.72	4.4	0.00	40.00
Hours Spent on the Internet Per Week	14.14	23.65	0.00	120.00
Total Hours Per Week With Media	35.27	49.59	0.00	300.00
Approx. Hours Per Week With Negative Media	7.36	24.28	0.00	205.00
Dependent variables
Thrillseeking Behavior	1.54	0.66	1.00	5.0
Rebellious Behavior	1.56	0.82	1.00	5.0
Reckless Behavior	1.41	0.76	1.00	5.0
Antisocial Behavior	2.06	0.97	1.00	5.0
Average Risk-Taking Behavior	1.63	0.68	1.00	5.0
Extracurricular Activities	1.30	1.40	0.00	6.0
Overall Grades	3.22	1.80	1.00	9.0

Note. Possible ranges: Parental Monitoring – 1 to 4; Mother and Father Demandingness – 1 to 5; Mother and Father Responsiveness – 1 to 4; Mother and Father Responsiveness – 1 to 5; Mother and Father Involvement – 1 to 5; Mother and Father Communication – 1 to 7; Goal-Oriented and Risk-Taking Peers – 1 to 4; Television, Music, Magazines, Internet, Media Total, and Negative Media were self-reported hours; Extracurricular Activities were self-reported totals; Overall Grades – 1 to 9.

Given the nature of our research questions and sample size, multiple linear regression analysis was selected as the most appropriate tool to answer our questions. To assess the potential for multicollinearity among predictor variables, Pearson product moment correlations were run. A correlation of 0.8 or greater was used as the cutoff for identifying multicollinearity. All decisions on the statistical significance of the findings were made using a criterion alpha level of 0.05. No issues of multicollinearity were identified.

Model development

In an attempt to identify the most significantly contributing factors to risk-taking and goal-oriented behaviors, we used a multi-tiered multiple regression approach using the stepwise method of analysis. Given the sample size and number of predictor variables across the contexts being measured in this study, this was selected as an appropriate model reduction approach. For each criterion variable (i.e. Total Risk-Taking, Rebellious Behavior, Reckless Behavior, Total GPA, and Extra-Curricular Activities), an initial model was explored including the following predictor groups: Mother variables (Mother Demandingness, Mother Responsiveness, Mother Involvement, and Mother Communication), father variables (Father Demandingness, Father Responsiveness, Father Involvement and Father Communication), peer variables (Peer Goal-Oriented and Peer Risk-Taking), religious variables (Religious Attendance, Religious Involvement and Religious Importance), and media variables (Hours of Television, Hours of Music, Hours of Magazines, Hours of Internet, Total Media Hours, and Bad Media Hours) in succession. This was repeated for each criterion variable – Total Risk-Taking, Rebellious Behavior, Reckless Behavior, GPA, and Extra-Curricular Activities – and for each racial/ethnic subgroup separately. This process is referred to as Stage I in each of the following subsections.

Caucasian subsample

Stage I

The criterion variable Total Risk-Taking was regressed separately on each of the five predictor groups (i.e. mother variables, father variables, peer variables, religious variables, and media variables), with a separate regression analysis for each predictor group.After completing a regression with each predictor group and examining results across all five analyses, Mother Demandingness (β = −0.259; p < 0.01), Peer Risk-Taking (β = 0.637; p < 0.01), and Religious Importance (β = −0.277; p < 0.01) each entered as statistically significant. The father variables and media variables predictor groups each did not provide any significant predictors in relation to Total Risk-Taking. The criterion variable Rebellious Behavior yielded similar results. Mother Demandingness (β = −0.320; p < 0.01) contributed as a significant predictor, as did Peer Risk-Taking (β = 0.708; p < 0.01) and Religious Attendance (β = −0.326; p < 0.01) in each separate model. The dependent variable Reckless Behavior, too, yielded similar results with Mother Demandingness (β = −0.319; p < 0.01) entering as a statistically significant predictor along with Father Involvement (β = −0.208; p < 0.05), Peer Risk-Taking (β = 0.612; p < 0.01), and Religious Importance (β = −0.263; p < 0.01).

This process was repeated for the two remaining criterion variables, GPA and Extra-Curricular Activities. When GPA was regressed on the five predictor groups separately, Mother Involvement (β = −0.377; p < 0.01), Father Involvement (β = −0.342; p < 0.01), Peer Risk-Taking (β = 0.235; p < 0.01), Peer Goal-Oriented (β = −0.258; p < 0.01), Religious Attendance (β = −0.226; p < 0.01), and Internet Hours (β = 0.178; p < 0.05) entered to be statistically significant predictors of the variable. The Extra-Curricular Activities criterion yielded Mother Involvement (β = 0.384; p < 0.01), Father Involvement (β = 0.388; p < 0.01), Peer Goal-Oriented (β = 0.436; p < 0.01), and Religious Activities (β = 0.299; p < 0.01) as statistically significant predictors.

After completing these initial regression analyses, the researchers then used those results to develop a refined model. Using the significant predictors in Stage I of the analysis, new multiple regression models were formulated for further exploration. The models for Total Risk-Taking, Rebellious Behavior, and Reckless Behavior appeared to be very similar, and thus, we used Total Risk-Taking to represent all risk behaviors and no longer included Rebellious Behavior and Reckless Behavior as separate analyses.

Stage II

In the following models, beta weights and significance levels are included in parentheses only for the variables that reached statistical significance. See Table 3. In the final Total Risk-Taking model, we included the predictors Mother Demandingness, Father Involvement, Peer Risk-Taking (β = 0.650; p < 0.01), Religious Importance, and Religious Attendance, based on the Stage I results. The one significant predictor(i.e. Peer Risk-Taking) accounted for approximately 42% of the explained variance for the Total Risk-Taking criterion variable.

Table 3.

Multiple regression models – Stage II – Caucasian.

	Variables entered	Model	B	SEB	β	Model Adj R²
Risk-Taking Behavior Model^a	Mother Demandingness	Peer Risk-Taking	0.42	0.047	0.650	0.417**
	Father Involvement
	Peer Risk-Taking
	Religious Importance
	Religious Attendance
GPA Model^b	Mother Involvement	Mother Involvement	−0.662	0.243	−0.257
	Father Involvement	Peer Risk-Taking	0.515	0.181	0.267	0.167**
	Peer Risk-Taking
	Peer Goal-Oriented
	Religious Attendance
	Internet media
	Extra-Curricular Model^c	Mother Involvement	Father Involvement	0.673	0.217	0.272
	Father Involvement	Peer Goal-Oriented	0.832	0.218	0.335	0.223**
	Peer Goal-Oriented
	Religious Activities

Note: ^**p < 0.01; ^aF = 77.67, p = 0.000; ^bF = 11.61, p = 0.000; ^cF = 16.80, p = 0.000.

The final GPA model included the predictors Mother Involvement (β = −0.257; p < 0.01), Father Involvement, Peer Risk-Taking (β = 0.267; p < 0.01), Peer Goal-Oriented Behavior, Religious Attendance, and Internet Media Hours. The significant variables accounted for approximately 17% of the overall explained variance in the model. The final Extra-Curricular model included the predictors Mother Involvement, Father Involvement (β = 0.272; p < 0.01), Peer Goal-Oriented Behavior (β = 0.335; p < 0.01), and Religious Activities. The two statistically significant variables accounted for approximately 22% of the variance in the overall model.

Arab American subsample

Stage I

The criterion variable Total Risk-Taking was regressed separately on each of the five predictor groups (i.e. mother variables, father variables, peer variables, religious variables, and media variables,with a separate regression analysis for each predictor group. After completing a regression analysis with each predictor group, and examining results across all five analyses, Mother Demandingness (β = −0.222; p < 0.05) and Peer Risk-Taking (β = 0.446; p < 0.01) each entered as statistically significant. The criterion variable Rebellious Behavior yielded similar results. Mother Demandingness (β = −0.263; p < 0.01) and Father Demandingness (β = −0.281; p < 0.01) contributed as significant predictors, as did Peer Risk-Taking (β = 0.387; p < 0.01). The criterion variable Reckless Behavior yielded only two significant predictors,Mother Demandingness (β = −0.311; p < 0.01) and Peer Risk-Taking (β = 0.372; p < 0.01).

This process was repeated for the two remaining criterion variables, GPA and Extra-Curricular Activities. When GPA was regressed on the five predictor groups separately, Mother Involvement (β = −0.282; p < 0.01), Father Involvement (β = −0.368; p < 0.01), Peer Risk-Taking (β = 0.318; p < 0.01), Peer Goal-Oriented (β = −0.269; p < 0.01), and Bad Media Hours (β = 0.310; p < 0.01) entered as statistically significant predictors. The Extra-Curricular Activities criterion yielded only Religious Attendance (β = −0.231; p < 0.05) and Religious Activities (β = 0.301; p < 0.01) as statistically significant predictors.

Stage II

Next, the same method of model refining was used as in the Caucasian sample. In reporting of the following models, beta weights and significance level is included only for the variables that reached statistical significance. See Table 4. The final Total Risk-Taking model included the predictors Mother Demandingness, Father Demandingness, and Peer Risk-Taking (β = 0.429; p < 0.01) due to their previous significance in Stage I analysis. The significant predictor Peer Risk-Taking accounted for approximately 18% of the explained variance for the Total Risk-Taking criterion variable.

Table 4.

Multiple regression models – Stage II – Arab American.

	Variables Entered	Model	B	SEB	β	Model Adj R²

Risk-Taking Behavior Model^a	Mother Demandingness	Peer Risk-Taking	0.248	0.053	0.429	0.175**
	Father Demandingness
	Peer Risk-Taking
GPA Model^b	Mother Involvement	Father Involvement	−0.568	0.221	−0.237
	Father Involvement	Peer Risk-Taking	0.494	0.157	0.289
	Peer Risk-Taking	Bad media	0.024	0.006	0.340	0.242**
	Peer Goal-Oriented
	Bad Media
Extra-Curricular Model^c		Religious Attendance	−0.181	0.081	−0.216
		Religious Activities	0.256	0.085	0.292	0.056**

Note: ^**p < 0.01; ^aF = 77.67, p = 0.000; ^bF = 10.89, p = 0.000; ^cF = 4.90, p = 0. 009.

The final GPA Model included the predictors Mother Involvement, Father Involvement (β = −0.237; p < 0.05), Peer Risk-Taking (β = 0.289; p < 0.01), Peer Goal-Oriented, and Bad Media Hours (β = 0.340; p < 0.01). The significant variables accounted for approximately 24% of the overall explained variance in the model. The final Extra-Curricular model included the predictors Religious Attendance (β = -0.216; p < 0.05), and Religious Activities (β = 0.292; p < 0.01). The two statistically significant variables accounted for approximately 6% of the overall model.

Discussion

The purpose of this study wasto measure variables from several key environmental contexts in adolescents’ lives, and examine their individual and combined contributions to both risk-taking behavior and goal-oriented behaviors. Overall, mother demandingness had the strongest association with less risk-taking behaviors among both racial/ethnic groups, followed by associations for both groups between risk-taking peers and risk-taking. Other significant results of interest included a strong relation between both goal-oriented peers and parental involvement and participation in extracurricular activities for Caucasians; however, there were no significant results among those variables for the ArabAmerican subgroup. Interestingly, there was a positive relation between risk-taking peers and GPA for both subgroups. These results are discussed in more detail below.

We hypothesized that the type of variable with each context would correlate in unique ways with the various criterion variables. Indeed, there was at least one variable from each context that did contribute to variance across the criterion variables, but it was the pattern of what contributed in each model that is noteworthy. As an example, for risk-taking behavior, peers’ behaviors were clearly associated but for GPA, parent variables demonstrated a negative relation with GPA while risk-taking peers and negative media had positive associations. Also, for extracurricular activities, parent involvement, peer goal-oriented behaviors, and religion contributed most significantly to the models. Interestingly, overall, there was primarily much overlap between the Caucasian and Arab-American adolescents in terms of which predictors significantly predicted outcomes, with some unique differences.

Peer risk-taking behavior was a significant predictor of total risk-taking behavior in both groups with a slightly higher proportion of variance explained for Caucasians.While the results did support previous research on the association amongst risk-taking peers and risk-taking behaviors, it was surprising that there was such a strong association, while there was minimal to no negative association among other perceived positive variables such as goal-oriented peers, religious involvement, and parental relationships. Clearly, peers play an important role in developing risk-taking behavior. However, other contexts also matter, e.g. for both subgroups, mother demandingness was negatively associated with risk-taking behaviors while religiosity was negatively associated with risk-taking behaviors in Caucasians.

For GPA, what is most striking is that there were only small associations of variables that might be considered ‘positive’ or desirable with grade point average, while several ‘negative’ or undesirable variables had significant positive relations with grade point average. It is these negative associations that were counterintuitive. Specifically, more parental involvement and more goal-oriented peers had the strongest negative association with GPA, while risk-taking peers had the strongest positive relationship with GPA.We did not distinguish between intrusive parental involvement that can be very behaviorally controlling and general supportive parental involvement, which could account for some of these unexpected directions of relations. Future research should address these possibilities.

Regarding extracurricular activities as a goal-oriented behavior outcome variable, only religious activity overlapped between the two groups. For Caucasians, parent variables and goal-oriented peers were positively associated with extracurricular activities. Interestingly, for the Arab-American subgroup there were no significant associations beyond religiosity. These varying patterns indicate a need to consider that the nature of these two groups may be unique in several key ways.

Limitations and directions for future research

We included a large number of variables in an effort to capture the breadth of contexts in which adolescents develop. However, in the model reduction process, some variables that may otherwise have played a small role in these relations may not have emerged in the analyses. Thus, this study emphasizes the most significant variables that explain variance in the risk-taking and goal-oriented outcomes of interest in this sample. Additionally, some of our measures were exploratory and future research will be helpful in expanding their use. In addition to other weaknesses noted above, a weakness of this study was that the sample was from a single public, suburban school in the Midwest, which may affect generalizability. It would be beneficial to expand the current results using a larger number of participants, particularly with the Arab-American subgroup as there is minimal research in this area. Also, the adolescents’ range of responses on each media variable indicated that they did not appear to be intentional responses but rather they poorly estimated the time available in a week. This may at least partially explain why those measures were uncorrelated with anything in the study. These are some of the weaknesses that may affect interpretation of results.

Conclusions

Despite limitations, the results of the current study demonstrate that multiple factors in these two groups of adolescents’ lives, ranging from the microsystem level to the exosystem level, contribute to their risk-taking behaviors and goal-oriented behaviors in overlapping and some unique ways. While results support and add to the breadth of available literature pertaining to adolescence, overall the results indicate that adolescents, like all humans, are complex individuals who receive input from multiple influences at any given time such as parents, peers, religion, and media. With all the different variables impacting the adolescent it is important to look at significant variables and build upon those. For example, while it is evident in the current study and previous research that peers are the strongest predictors of adolescent risk-taking behavior, positive parental involvement and demandingness as well as religious involvement may counteract some of the influence that peers may have on risk-taking decisions. Mediational analyses could be considered in future research.

Implications and applications

There are implications for school and clinical practice, primarily in terms of where to focus intervention efforts. These two major demographic/cultural groups studied here (Caucasian and Arab-American) evidenced similar trends for risk-taking variables but slightly different trends for goal-oriented behaviors leading to some generalizability in potential intervention design. The current study indicates that while adolescents from different cultures may display cultural differences at the heart of adolescent growth they still are faced with similar challenges and choices. Programming of interventions does not need to be distinctly different but rather a general understanding of cultural differences may be needed (e.g. greetings, terminology, proximity), as well as possibly the way that different influences are addressed in any approach.

As previously stated, one variable does not solely impact other variables but rather there are several environmental variables interacting at any given time. It is important for school psychologists and other school-based practitioners to understand the interaction among the adolescents and their environment. However, with respect to the specific differences discovered in this study, there are unique implications for approaching intervention, whether through formal programming or either formal or informal intervention with individuals and small groups. For Caucasian adolescents, these results suggest that when addressing goal-oriented behaviors (specifically GPA and involvement in extracurricular activities), topics addressed should include both parent and peer factors that may influence adolescents. For risk-taking behavior, focusing directly in on peer risk-taking behavior as a critical correlateis, based on these results, likely a more efficient use of intervention time. Similarly, for risk-taking behavior among Arab-American adolescents, it was also peer risk-taking behavior that solely predicted their own risk-taking behavior, and thus, that should similarly be the focus of intervention efforts. On the other hand, interventions for Arab-American adolescents’ goal-oriented behavior will, based on these results, be most effective when interventions are conceptualized through a broader-based ecological lens and include aspects of parenting (involvement), peers, media, and religion alike. These results provide important information about these two unique subgroups of adolecents that can be used to frame prevention and intervention approaches.

Footnotes

Author biographies

Joshua J. Tynan holds a PhD in school psychology and is interested in research on how to best predict risk behavior in adolescents that ultimately can interfere with life outcomes including academics and personal growth. He plans to build on his research with Dr Somers and incorporate other existing research results to develop school-based prevention and intervention efforts.

Cheryl L. Somers holds a PhD in school psychology. Her research emphasizes risk and reliance among adolescents from a variety of backgrounds, and focuses, similar to that of Dr Tynan, on a variety of contextual predictors of adolescent risk behavior and academic success. She and Dr Tynan are working together to craft school-based models that will maximize adolescents’ academic and personal success.

Jamie H. Gleason holds a PhD in educational evaluation and research and is interested in higher level statistical modeling of educational and developmental psychology research questions. He has a special interest in adolescent risk behavior and the role of schools in potentially mitigating that process. He and Dr Somers have published several studies together on adolescent risk behavior.

Barry S. Markman holds a PhD in experimental psychology and is interested in research design and statistical applications, and research on ADHD in adolescents and adults.

Jina Yoon holds a PhD in school psychology. Her primary research emphasis is on teacher–student relationships, school climate, bullying, and enhancing successful social, emotional, and behavioral outcomes in children and adolescents.

References

Annunziata

D. A.

Hogue

Faw

Liddle. (2006) Family functioning and school success in at-risk, inner-city adolescents. Journal of Youth and Adolescence 35: 105–113. doi: 10.1007/s10964-005-9016-3.

Arnett

J. J.

(1995) Adolescents’ uses of media for self-socialization. Journal of Youth and Adolescence 24: 519–533. doi: 10.1007/BF01537054.

Bahr

S. J.

Hoffmann

J. P.

Yang

(2005) Parental and peer influences on the risk of adolescent drug use. Journal of Primary Prevention 26: 529–551. doi: 10.1007/s10935-005-0014-8.

Ben-Arieh

Attar-Schwartz

(2013) An ecological approach to children’s rights and participation: Interrelationships and correlates of rights in different ecologicalsystems. American Journal of Orthopsychiatry 83(1): 94–106. doi: 10.1111/ajop.12003.

Botte

R. A.

(2000) The mirror of television: A comparision of black and white adolescents’ body image. Journal of Communication 50(3): 157–167. doi: 10.1111/j.1460-2466.2000.tb02857.x.

Bronfenbrenner

(2005) Making human being human: Bioecological perspectives on human development, Thousand Oaks, CA: Sage.

Brofenbrenner

(1977) Toward an experimental ecology of human development. American Psychologist.513–531. doi: 10.1037/0003-066x.32.7.513.

Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press.

Brofenbrenner

(1989) Ecological systems theory. Annals of Child Develoment 6: 187–249.

10.

Brown

J. D.

Witherspoon

E. M.

(2002) The mass media and American adolescents’ health. Journal of Adolescent Health 31: 153–170. doi: https://dx-doi-org.web.bisu.edu.cn/10.1016/S1054-139X((02)00507-4.

11.

Byfield

(2008) The impact of religion on the educational achievement of black boys: A UK and USA study. British Journal of Sociology of Education 29: 189–199. doi: 10.1080/01425690701837547.

12.

Connell

J. P.

Halpern-Felsher

B. L.

Clifford

Crichlow

Usinger

(1995) Hanging in there: Behavioral, psychological and contextual factors affecting whether African American adolescents stay in high school. Journal of Adolescent Research 10: 41–63. doi: 10.1177/0743554895101004.

13.

Cottrell

S. A.

Branstetter

Cottrell

Harris

C. V.

Rishel

Stanton

B. F.

(2007) Development and validation of a parental monitoring instrument: Measuring how parents monitor adolescents’ activities and risk behaviors. Family Journal 15(4): 328–335. doi: 10.1177/1066480707303748.

14.

Fine

M. A.

Moreland

J. R.

Schwebel

A. I.

(1983) Long-term effects of divorce on parent–child relationships. Developmental Psychology 19: 703–713. doi: 10.1177/1066480701093005.

15.

Fredricks

J. A.

Eccles

J. S.

(2005) Developmental benefits of extracurricular involvement: Do peer characteristics mediate the link between activities and youth outcomes? Journal of Youth and Adolescence 34: 507–520. doi: 10.1007/s10964-005-8933-5.

16.

Guilamo-Ramos

Jaccard

Dittus

Bouris

A. M.

(2006) Parental expertise, trustworthiness, and accessibility: Parent adolescent communication and adolescent risk behavior. Journal of Marriage and Family 68: 1229–1246. doi: 10.1111/j.1741-3737.2006.00325.x.

17.

Gullone

Moore

S. M.

(2000) Adolescent risk-taking and the five-factor model of personality. Journal of Adolescence 23: 393–407. doi: 10.1006/jado.2000.0327.

18.

Gullone

Moore

Moss

Boyd

(2000) The adolescent risk-taking hispanic questionnaire: Development and psychometric evaluation. Journal of Adolescent Research 15: 231–250. doi: 10.1177/0743558400152003.

19.

Harris

(1998) The nurture assumption: Why children turn out the way they do, New York, NY: Free Press.

20.

Henry

C. S.

Merten

M. J.

Plunkett

S. W.

Sands

(2008) Neighborhood, parenting, and adolescent factors and academic achievement in Latino adolescents from immigrant families. Family Relations 57: 579–590. doi: 10.1111/j.1741-3729.2008.00524.x.

21.

Jaccard

Blanton

Dodge

(2005) Peer Influences on risk behavior an analysis of the effects of a close friend. Developmental Psychology 41: 135–147. doi: 10.1037/0012-1649.41.1.135.

22.

King

P. E.

Furrow

J. L.

(2004) Religion as a resource for positive youth development: Religion, social capital, and moral outcomes. Developmental Psychology 40: 703–713. doi: 10.1037/1941-1022.

23.

Kremar

Greene

(2000) Connections between violent television exposure and adolescent risk taking. Journal of Media Psychology 2(3): 195–217. doi: 10.1207/S1532785XMEP0203_1.

24.

Laird

R. D.

Criss

M. M.

Pettit

G. S.

Dodge

K. A.

Bates

J. E.

(2008) Parents’ monitoring knowledge attenuates the link between antisocial friends and adolescent delinquent behavior. Journal of Abnormal Child Psychology 36: 299–310. doi: 10.1007/s10802-007-9178-4.

25.

Larson

(1995) Secrets in the bedroom: Adolescents’ private use of media. Journal of Youth and Adolescence. Special Issue: Adolescents’ Uses of the Media 24: 535–550. doi: 10.1007/BF01537055.

26.

LeCroy

C. W.

Krysik

(2008) Predictors of academic achievement and school attachment among Hispanic adolescents. Children and Schools 30: 197–209. doi: 10.1093/cs/30.4.197.

27.

Martin

(2000) Understanding dysfunctional and functional family behaviors for the at-risk adolescent. Adolescence 35: 785–792. doi: 10.1037/a0036239.

28.

Muss

R. E.

(1996) Theories of adolescence, New York, NY: McGraw-Hill.

29.

Paulson

S. E.

(1994) Relations of parenting style and paternal involvement with ninth grade students’ achievement. Journal of Early Adolescence 14: 250–267. doi: 10.1177/027243169401400208.

30.

Pianta

R. C.

(1999) Enhancing relationships between children and teachers, Washington, DC: American Psychological Association, pp. 45–64.

31.

Price

L. F.

(2005) The biology of risk taking. Educational Leadership 62: 22–26.

32.

Prinstein

M. J.

Boergers

Spirito

(2001) Adolescents’ and their friends’ health- risk behavior: Factors that alter or add to peer influence. Journal of Pediatric Psychology 26(5): 287–298. doi: 10.1093/jpepsy/26.5.287.

33.

Regnerus

Elder

G. H.

(2003) Staying on track in school: Religious influences in high- and low-risk settings. Journal for the Scientific Study of Religion 42: 633–649. doi: 10.1046/j.1468-5906.2003.00208.x.

34.

Rostosky

S. S.

Danner

Riggle

E. D. B.

(2008) Religiosity and alchohol use in sexual minority and heterosexual youth and young adults. Journal of Youth Adolescence 37: 552–563. doi: 10.1007/s10964-007-9251-x.

35.

Rostosky

S. S.

Regnerus

M. D.

Wright

M. L. C.

(2003) Coital debut: The role of religiosity and sex attitudes in the add health survey. Journal of Sex Research 40: 358–367. doi: 10.1177/0739986308322912.

36.

Sallquist

Eisenberg

French

D. C.

(2010) Indonesian adolescents’ spiritual and religious experiences and their longitudinal relations with socioemotional functioning. Developmental Psychology 46(3): 699–716. doi: 10.1037/a0018879.

37.

Schiff

Fang

(2014) Adolescent substance use in Israel: The roles of exposure topolitical traumas and posttraumatic stress symptoms. Psychology of Addictive Behaviors 28(2): 453–463. doi: 10.1037/a0035603.

38.

Simpson

McBride

(1992) Family, friends, and self (FFS) assessment scales for Mexican American youth. Hispanic Journal of Behavioral Sciences 14: 327–340. doi: 10.1177/07399863920143003.

39.

Sinha

J. W.

Cnaan

R. A.

Gelles

R. J.

(2007) Adolescent risk behaviors and religion: Findings from a national study. Journal of Adolescence 30: 231–249. doi: 10.1016/j.adolescence.2006.02.005.

40.

Somers

C. L.

Paulson

S. E.

(2000) Students’ perceptions of parent–adolescent closeness and communication about sexuality: Relations with sexual knowledge, attitudes, and behaviors. Journal of Adolescence 23: 629–644. doi: 10.1006/jado.2000.0340.

41.

Somers

C. L.

Tynan

J. J.

(2006) Sexual dialogue and content on television and adolescent sexual outcomes. Adolescence 41(161): 15–38.

42.

Spoth

Redmond

Hockaday

Yoo

(1996) Protective factors and young adolescent tendency to abstain from alcohol use: A model using two waves of intervention study data. American Journal of Community Psychology 24: 749–770. doi: 10.1007/BF02511033.

43.

Steinberg

Blatt-Eisengart

Cauffman

(2006) Patterns of competence and adjustment among adolescents from authoritative, authoritarian, indulgent, and neglectful homes: A replication in a sample of serious juvenile offenders. Journal of Research on Adolescence 16: 47–58. doi: 10.1111/j.1532-7795.2006.00119.x.

44.

Suldo

S. M.

Mihalas

Powell

French

(2008) Ecological predictors of substance use in middle school students. School Psychology Quarterly 23: 373–388. doi: 10.1037/1045-3830.23.3.373.

45.

Taylor

R. D.

Lopez

E. I.

(2005) Family management practice, school achievement, and problem behavior in African-American adolescents: Mediating processes. Applied Developmental Psychology 26: 39–49. doi: 10.1037/a0021369.

46.

Vandewater

E. A.

Shim

Caplovitz

A. G.

(2004) Linking obesity and activity level with children’s television and video game use. Journal of Adolescence 27: 71–85. doi: 10.1016/j.adolescence.2003.10.003.

47.

Wallace

J. M.

Williams

D. R.

(1997) Religion and adolescent health-compromising behavior. In: Schulenberg

Matts

J. L.

Hurrelmann

(eds) Health risks and developmental transitions during adolescents, Cambridge: Cambridge University Press, pp. 444–468.