The Relationship between Risk-Taking with a Sibling and Risk-taking in General in Adolescence: The Role of Gender Composition
Thesis Forensic Child and Youth Care Sciences Graduate School of Child Development and Education University of Amsterdam Student: Simone Lochs Student number: 12926930 Thesis supervisor: dr. I. N. Defoe Second supervisor: dr. M. J. Noom
Abstract
This correlational research studies the relationship between adolescents’ risk-taking with a sibling and risk-taking in general. It is expected that higher levels on risk-taking with a sibling are related to higher level on risk-taking in general, which is consistent with the Cognitive Social Learning Theory. Furthermore, the role of gender composition (e.g. both male, both female, and mixed dyads) as a moderator is studied. Based on the Evolutionary Personality Theory, environmental theories, and the Cognitive Social Learning Theory, it is expected same sex siblings influence each other more, and particularly male sibling dyads have a stronger moderation effect than the other dyads. The data originated from the ‘Adolescent Risk-Taking Project’ (ART), a prospective longitudinal study in the Netherlands. From wave 2 of this data, 405 participants (46.2 % female; Mage=14.64) were selected, whom had a sibling between 11 and 21 years old. Results show higher levels of risk-taking with a sibling are related to higher levels of risk-taking in general. Gender composition was not a significant moderator for this relationship. These findings suggest implementing siblings into
interventions regarding risk-taking should be considered, since such interventions are rare. Also, limitations and strengths of this study are discussed.
The Relationship between Risk-Taking with a Sibling and Risk-taking in General in Adolescence: The Role of Gender Composition
Risk-taking can have negative consequences for an individual but can also harm others. Everyone is facing decisions about risk-taking every single day. This can be about simple things such as speeding in traffic, jumping of a helicopter, or not using a condom with somebody you just met (Figner et al., 2009). The action can seem simple, but the
consequences of these actions could be about life and death. Negative consequences of risk-taking behavior are, for example, getting a traffic fine or an STD infection. But jumping of a helicopter could also result in dying, if something goes wrong with the parachute. Other examples of risk-taking are delinquency and substance use, because participating in this kind of behavior brings certain risks along (Figner et al., 2009). Delinquency could result in going to jail and substance use could result in reckless behavior. Since risk-taking is integrated in many dimensions of life (e.g., social, financial, recreational), it is an important subject to study (Weber et al, 2002). It is important to take the social context into consideration when researching risk-taking (Willoughby et al., 2014). Therefore, in this study the relationship between risk-taking with a sibling and risk-taking in general will be studied. Furthermore, the moderating role of gender composition within the sibling dyad in adolescent risk-taking will be studied.
Adolescent Risk-Taking
Risk-taking can be defined as choosing the riskiest option out of the possible choices (Defoe et al., 2015). One of these choices is the riskiest, but that does not necessarily mean the others are not. It is often the case that the riskier option comes with a more positive outcome and the safer option with a more negative one at first. However, the riskier option could result in a way more negative outcome than choosing the safest option right away. This is called the increasing-risk dynamic (Figner &Weber, 2011). In short, this increasing-risk
dynamic suggests that riskier options result in a positive outcome in short terms, and the safer options result in negative outcomes in short term, whereas in the long run the safer option is often more positive (Figner & Weber, 2011).
Risk-taking is not an individual trait, but can be influenced by the decision maker, the characteristics of the situation, and most importantly a combination of the two (Figner & Weber, 2011). A factor shaping the decision maker is, for example, age. Generally, people often think adolescents are more likely to take risks than other age groups. They are searching for their identity, and in order to do so, show experimental behavior, including risk taking behavior (Holt et al., 2012). However, research, particularly experimental studies, often shows otherwise (Figner et al., 2009). Another example of a factor shaping the decision maker is an individual’s risk perception or in other words, weighting the costs up to the benefits. Besides individual traits, the characteristics of the situation are of importance, such as what the
decision is about or whether it is a cold, deliberative or a hot, affective process. Choices made in a social context are often considered affective (Figner et al., 2009). For example, an
affective choice could be to use a condom or not or whether to gamble, whereas a deliberative choice could be investing money (Figner & Weber, 2011). Adolescents appear to only show increased risk-taking in affective choices, in deliberative choices they show the same level of risk-taking as adults and children (Figner & Weber, 2011). Thus, risk-taking is influenced by the decision maker (e.g. age, risk perception) and the characteristics of the (social) situation, but also by an interaction between the two.
The Sibling Context and Adolescent Risk-Taking
Siblings form an important component of the social context for adolescents (Dunn, 2002). Most children interact more with their siblings than with their peers or parents (Dunn, 2002; Larson & Verma, 1999). The interaction between siblings is emotionally intense, since it teaches children positive and negative ways of interacting with others (Katz et al.,1992).
This intense sibling relationship can be seen in the perspective of the Cognitive Social Learning Theory (Bandura, 1977). This theory is about observational learning, where one imitates people in their social environment. Besides, the theory also describes reciprocal determinism and self-efficacy as important factors. Reciprocal determinism is the interaction between a child and their social environment, constantly influencing one another. Self-efficacy is about how competent children view themselves to be. If a child perceives his/her-self as highly competent, it is more likely to solve social problems, in comparison to other children, who don’t perceive their selves as competent (Bandura, 2006). In short,
observational learning, reciprocal determinism and self-efficacy have an impact on how adolescents are influenced by their peers, particularly their siblings.
Although (experimental) research on sibling influence is limited, the few amount of such studies do show a significant sibling effect. Nevertheless, as far as known, no
(correlational) research into risk-taking with a sibling and risk-taking in general has been done, that would be comparable to the current study. There is one experimental study about the influence of sibling presence in risk-taking, which shows less risk-taking when a sibling is present. This could be because, especially if the sibling is younger, adolescents want to give a positive example (Defoe, 2016). Another experimental study, which was done using a
computerized driving task during a functional MRI scan, showed that higher sibling closeness was associated with lower externalizing behavior, such as recklessness, anger, lack of
concentration and lack of inhibition (Rogers et al., 2018). This is in line with a meta-analysis by Buist et al. (2013), that showed a better relationship between siblings is related to less externalizing and internalizing problems. Correlational studies that would be comparable to this study have not been found, but there are some studies that show sibling risk-taking is related to adolescents’ risk-taking. For example, a longitudinal study using interviews by Trim et al. (2006) showed that older sibling alcohol use predicts the alcohol use of the
younger sibling. This is in line with a cross-sectional study by Samek et al. (2018), which used computerized assessments, self-reported questionnaires and diagnostic interviews. Other correlational studies found sibling delinquency is a risk factor for adolescents’ delinquency (Walter, 2018; Walters, 2020; Abderhalden & Evans, 2018; Buist, 2010). For example, more sibling arrests are related to more arrests for an adolescent offender (Abderhalden & Evans, 2018). Buist (2010) found the older siblings’ delinquency is related to the younger siblings’ delinquency two years later. Another study shows smoking habits of a sibling are related to adolescents’ tobacco use (Von Bothmer, 2002). Experimental studies on sibling influences in adolescent risk-taking are scarce. However, there is a small but growing amount of
experimental research on peer influence in adolescent risk-taking, showing that the presence of a peer increases risk-taking (Cavalca et al., 2013; Gardner & Steinberg, 2005; Chein et al., 2011; Smith et al., 2014).). Peers are individuals who you know that are of the same age as yourself (Holt et al., 2012). They could for example be in the same class or live in the same neighborhood (Holt et al., 2012). Considering this definition, siblings can be seen as a special kind of peer, since you know them and they mostly are of similar age. Taking previous research into consideration, this study will focus on the relationship between risk-taking with a sibling and risk-taking in general, since siblings could play an important role and yet research is scarce. The expectation is that adolescents who take more risks with their siblings also take more risks in general. As far as known, correlational research into risk-taking in presence of a sibling and risk-taking in general is non-existent. Therefore, this hypothesis is based on sibling studies that showed a relationship between the siblings’ risk-taking and adolescent risk-taking. Besides, there seems to be a relation between the presence of a peer and increased risk-taking. Sibling presence will be assessed by asking the participants whether they have ever performed in risk-taking in presence of their sibling.
There are gender differences in risk-taking, which can be explained by both evolutionary factors, environmental factors and a combination of the two, since evolution always happens and is influenced by a certain environment (Byrnes et al., 1999; Foley & Gamble, 2009). For example, gender differences in adolescent risk-taking can be explained by the Evolutionary Personality Theory. This theory considers traits in humans that are caused by adaptive demands of evolutionary history (Holt et al., 2012). According to this theory, males are more extraverted than females and therefore show more risk-taking (Nettle, 2006). An explanation for this is that risk-taking is associated with an increased chance to breed successfully (Pawlowski et al., 2008). In short, according to the evolutionary perspective, men are more extraverted and therefore more risk-taking, in order to breed successfully.
Moreover, environmental theories also explain gender differences. The environment of adolescents influences their development (Bronfenbrenner, 1994). Through parenting a
gender identity is formed by children, which is a sense of femaleness or maleness. During life, adolescents experience gender typing, where different genders are treated differently and expected to behave in different ways (Ellemers, 2018). This is also the case when it comes to risk-taking (Fisk, 2016). Like any other kind of behavior, patterns are often expected of males and females, which could have negative consequences. For example, labelling girls and women as more risk-averse could limit them in their opportunities and benefits of risk-taking (Maxfield et al., 2010). A literature review by Fisk (2016) supports this, claiming risk-taking is a gendered action. Women take fewer risks, because it is less rewarding for them. Also, there are institutional gatekeepers that give women fewer opportunities to take risks, even though women often have a better educational status than men (Fisk, 2016). Therefore, there is an underrepresentation of women at the top of hierarchies. Men are for example more likely to be a leader, top earner or hold a key political function (Fisk, 2016).
Of note, a meta-analysis of 150 studies showed that risk-taking behavior is higher among males than females in the majority of topics asked about in questionnaires. Men are more likely to weight the benefits of the risk behavior over the costs than women (Byrnes et al., 1999). Other studies also support these findings (Gardner & Steinberg, 2005; Parsons et al., 2000). For example, Parsons and colleagues (2000) did a study on 704 college students between 17 and 25 years old, using a survey. They found women see more benefits in condom use and more costs in unprotected sex than men do. Men also take more risks when
performing a gambling task (Weller et al., 2010) or making financial decisions (Jianakoplos & Bernasek, 1998).
Gender Composition Effects in Social Influence
Studies about gender composition in social influence are scarce, especially studies about mixed gender groups. Gender composition is about the type of genders in the dyad, in other words whether they are of the same gender or not. This is of importance when
considering sibling influence, since research shows that men and women differ in taking risks and in the ways they influence each other (Trim et al., 2006; Boyle et al., 2001; Byrnes et al., 1999). One relevant study, which was a longitudinal study using interviews, showed that sibling influence is larger when siblings are of the same sex (Trim et al., 2006). This is in line with a study by Boyle et al. (2001) that showed similarity in substance use between siblings through surveys. This effect is stronger when the siblings are of the same sex, with higher similarities when the siblings are both male, meaning more substance use in the older sibling was related to more substance use in the younger sibling. Thus, same sex siblings seem to influence each other more than mixed sex siblings (Trim et al., 2006; Boyle et al., 2001). Also, the aforementioned study of Gardner and Steinberg (2005) reported that peer effects in risk-taking are larger for men in comparison to females, meaning males are more influenced than females. However, this study only compared same sex groups and studies the effects of
peer presence. Similarly, another study found that adolescent males in antisocial or risky situations are more likely to be influenced by peers in comparison to females (Brown et al., 1986). In conclusion, the current study contributes to the scientific knowledge about gender composition, because it not only considers same sex siblings, but mixed sex siblings as well. Namely, there are some studies that indicate there is a difference between same and mixed sex siblings (Boyle et al., 2001; Trim et al., 2006), but also between female and male siblings (Brown et al., 1986) or peers (Gardner & Steinberg, 2005).
In addition to the available research, the Cognitive Social Learning Theory could also be used to explain the gender composition effect in risk-taking with a sibling (Bandura, 1977). People imitate each other’s behavior, and therefore search for a model to copy. According to this theory, people are more likely to copy behavior if their model is more like them, so if they are the same sex (Bandura, 1977). In other words, if a same sex sibling would show risk-taking, an adolescent is more likely to copy this behavior.
Zooming in on mixed gender studies, the limited amount of such studies show conflicting results. The few studies that are available are correlational. For example, a correlational study by Whiteman et al. (2014) found no moderation of gender composition in the relation between older sibling deviant and sexual behavior and the behavior of the
younger sibling in those domains. In other words, there was no difference in same sex and mixed sex siblings in the relationship between the risk behavior of the older and the younger sibling. Another longitudinal study using interviews found that there is a bidirectional effect in siblings’ risk behavior, but also found no gender composition effect (Whiteman et al., 2017). Another study by Defoe et al. (2013) also found that there was similarity in
adolescents’ externalizing behavior, for both same and mixed sex siblings. However, a study which used questionnaires showed that there is a gender composition effect (Huijsmans et al., 2019). This effect was found for early adolescent delinquency influenced by sibling
delinquency, but only for sisters and mixed sibling pairs. For men, sibling delinquency seemed to influence late adolescent delinquency instead of early adolescent delinquency (Huijsmans et al., 2019). Another study by Buist (2010) into older and younger sibling delinquency found a moderation of gender composition. This study found significant results for same sex pairs (e.g. brother and sister pairs), but not for mixed sex pairs (e.g. older
brother/younger sister and older sister/younger brother sibling pairs).Thus, although results on gender composition are conflicted, there could possibly be an effect, as can be explained by the Cognitive Social Learning Theory (Bandura, 1977). In conclusion, the influence is expected to be larger when the siblings would be of the same sex and men are expected to be more susceptible to influence than women, as it is the case with peers. Therefore, in this study, same sex siblings are expected to influence each other more than mixed sex siblings. Also, the male sibling dyads are expected to have the strongest moderation effect.
Present Study
This study focuses on whether sibling presence is related to risk-taking in adolescents between 12 and 15 years old, using a self-reported questionnaire. Also, adolescents’ own motives for engaging in risk-taking will be analyzed and provided as descriptive information about the sample, adding a qualitative touch to this study. Most of the participants in the current study are in the early (12 and 13 years old) to mid-stage (14 and 15 years old) of adolescence, but will be referred to as adolescents. Furthermore, the moderation of gender composition will be studied. In this study, same and mixed gender dyads are compared to each other, resulting in 3 groups: both female, both male, and mixed dyad (i.e., containing a male and female). A self-reported questionnaire will be used to assess adolescents risk-taking. Risk-taking will be assessed by asking questions about soft drug use, alcohol use, smoking and delinquency, of which a composite score will be computed. Sibling presence will be measured by asking whether the adolescence have ever performed in risk-taking with their
sibling, for which a composite score will also be computed for the different risk behavior (i.e., soft drug use, alcohol use, smoking and delinquency).
The hypotheses are as follows. For hypothesis 1, we expect that sibling presence when engaging in risk-taking is related to higher levels of overall risk-taking in adolescents. This is based on the available studies on siblings (e.g. Rogers et al., 2018; Trim et al., 2006), on studies about risk-taking and the influence of peers (e.g. Gardner & Steinberg, 2005; Smith et al., 2014; Cavalca et al., 2013), and on the Cognitive Social Learning Theory by Bandura (1977). For hypothesis 2, we expect that sibling presence when engaging in risk-taking is related to higher levels of adolescent risk-taking, especially when they are of the same gender (versus mix gender). Furthermore, we predict that sibling presence is related to highest levels of risk-taking in male sibling dyads compared to female-female and male-female dyads. This hypothesis is based on the Evolutionary Personality Theory (Holt et al., 2012), environmental theories (Bronfenbrenner, 1994), and the Cognitive Social Learning Theory (Bandura, 1977). There are also studies which support this hypothesis (Boyle et al., 2001; Trim et al., 2006; Huijsmans et al., 2019; Buist, 2010), even though other studies found different results (Whiteman et al., 2014, Whiteman et al., 2017, Defoe et al., 2013).
Method Sample
The data originated from an existing prospective longitudinal study among
adolescents in the Netherlands, called ‘The Adolescent Risk-Taking (ART) Project (Defoe et al., 2016). This study contained three waves, which were collected in 2012, 2013 and 2014. The ART project examined adolescent risk-taking in multiple domains. Participants were recruited via seven schools, located in different cities in the Netherlands. From wave 2, 405 participants (46.2% female, 53.8% male) were selected, which had a sibling between 11 and 21 years old. The average of the participants is 14.64 years (SD=1.29) old. They were in their
second (N = 171, 42.2% female) or fourth year (N = 223, 55.1% female) of either
“preparatory middle-level applied education” (VMBO in Dutch) or “higher general continued education” (HAVO in Dutch). At the start of the study, most adolescents (93.2%) reported they were born in the Netherlands, whereas 61.6% identified as Dutch (Defoe et al., 2016). The others identified as Turkish or Turkish-Dutch (9.3%), Surinamese or Surinamese-Dutch (7.4%), Moroccan or Moroccan-Dutch (5.5%) or other ethnicities (16.2%) (Defoe et al., 2016). Most of the adolescents’ parents lived together or were married (68.4%), whereas 24.8% were divorced or separated (Defoe et al., 2016). For 11.0% of the fathers and 11.8% of the mothers no highest education level was reported. This is partly because parents (44.9% fathers; 46.5% mothers) were born in foreign countries with other education systems than the Netherlands. From the data that was reported on parents’ education level, 3.8% of mothers and 10.5% of fathers had an university degree, 35.8% of mothers and 28% of fathers had a lower or middle level of vocational training, and 6.7% of mothers and 6.4% of fathers had not completed high school education (Defoe et al., 2020).
In addition to the other questions, adolescents filled in a couple of open questions about why they would engage in risk-taking. In order to orientate on this topic, open question about soft drug use, alcohol use and smoking have been coded, using axial coding (Kendall, 1999). Inspired by prior literature (Bandura, 1977, Cooper, 2015 Furby & Beyth-Marom, 1992, Zang et al., 2019) different categories have been established, based on the answers participants gave.
Being cool/impressing other was mentioned most as a reason to engage in risk-taking. Besides this, enjoying it (‘lekker’ in Dutch) was the second most mentioned. For drinking alcohol specifically, having a good time was mentioned a lot. For using soft drugs this was the feeling it gives you and for smoking it was addiction. Other motivations that were frequently said were fitting in, experimenting, stress, and having fun.
Procedure
The schools that participated in this research are from six different regions in the Netherlands. Eight schools participated in the first wave after contact through telephone calls and emails, resulting in 810 potential participants for wave 1. A letter was sent to the parents to inform them about the research, with an option to refuse by sending a dissent letter back. If parents did not send a dissent letter back, passive permission was assumed (Defoe et al., 2016). Of the 810 potential participants, 9.75% did not participate because they did not receive parental permission, they did not want to participate their selves or due to other reasons, such as illness or absence on the day of data-collection (Defoe et al., 2016). In the second wave, one of the schools stopped participating due to organisational reasons. Before the participants filled in the questionnaire during school, written and verbal instructions were provided by trained research assistants. Besides the digital questionnaire, a cognitive task and a videogame were completed by participants, which will not be used in this study (Defoe et al., 2016).
Measures
General adolescents risk-taking. This construct contains four subcategories (alcohol use, soft drug use, smoking and stealing) that are combined as one average score for
adolescents risk-taking. A higher score means a higher level of participating in risk-taking. The individual scores on the subcategories are described below. The average score was calculated by coding answer 0 as 0 and answers 1 to 4 as 1 (Field, 2013). A Cronbach’s alpha of .670 was calculated for this construct, which is questionable (Santos, 1999).
Alcohol use was measured with one item, namely ‘Do you drink alcohol?’. Participants could answer on a scale from 0 = No to 5 = Yes, every day.
Soft drug use was measured with one item, namely ‘Do you use soft drugs
(cannabis/weed/hash/marihuana)?’. Participants could answer on a scale from 0 = No, I have never used soft drugs to 5 = Yes, every day.
Smoking was measured with one item, namely ‘Do you smoke sometimes (cigarette, cigar, shag, or pipe tobacco)?’. Participants could answer on a scale from 0 = No, I have never smokes to 5 = Yes, every day.
Stealing was measured with three items, namely ‘Have you ever stolen something from a shop or a department store?’, ‘Have you ever stolen a bike, scooter or motor bike?’, and ‘Have you ever stolen a wallet, purse, or something else from someone?’. Participants could answer on a scale from 0 = Never to 4 = Yes, in the past 12 months 3 times or more.
Risk-taking in the presence of siblings. Only participants that had a sibling between 11 and 21 years old where selected for this study. For smoking, drinking and soft drug use, participants were asked the same items that were used to measure adolescents risk-taking, with the addition of ‘in the presence of your sibling’. For example, alcohol use in the presence of a sibling was measured with the question ‘How many times did you drink alcohol in the presence of your sibling?’. For stealing, the question ‘How many times have you stolen something in the presence of your sibling?’ was used. The answer options were on a scale from 0 = No to 4 = 6 times or more, with an extra option 9 = Not applicable. An average score of adolescents risk-taking with the presence of a sibling was retrieved from the answer on the four subcategories (alcohol use, soft drugs use, smoking and stealing). A higher score means higher levels of participating in risk-taking with a sibling. The scores were calculated by coding answer 0 as 0 and answers 1 to 4 as 1. A Cronbach’s alpha of .518 was calculated, which is questionable (Santos, 1999).
Gender composition. This construct was examined with two items to divide the participants into three categories, namely sisters, brothers and mixed dyads. The first item is
‘Are you a boy or a girl?’, with the answer options 0 = boy and 1 = girl. The second item is ‘When asking questions about your sibling, we mean the sibling you mentioned as the one you spend most time with. What is the gender of the sibling you spend most time with?’, with the answer options 0 = male and 1 = female. The combination of these items determines in which of the three groups the participant will be placed.
Strategy of Analyses
The first hypothesis about the direct relationship between sibling presence and risk-taking is tested with a multiple regression analysis using the PROCESS macro version 3.4 (Hayes, 2019). The second hypothesis about the moderation effect of gender composition on the relationship between sibling presence while engaging in risk-taking and risk-taking general is also tested with a multiple regression analysis, using the PROCESS macro version 3.4 (Hayes, 2019). Other studies about risk-taking have also used a type of regression analysis (e.g. Fernie et al., 2010; Steinberg, 2010).The PROCESS macro is more advanced than a normal regression analysis, since it centres predictors, computes the interaction term automatically and does simple slope analyses (Field, 2013) .
Before running the analyses, all subjects with a sibling between 11 and 21 years old were selected in the data file and assumptions were tested. First, stem-and-leaf plots and boxplots showed whether the variables were normally distributed. This wasn’t the case, but since the PROCESS macro accounts for that, this is not of concern (Field, 2013). Second, the assumption of normality, linearity and homoscedasticity of residuals was tested, which were all met. Then, Mahalanobis distance was tested, but no outliers were found that needed to be deleted. The critical value χ2 for df = 5 (20.52) was not violated (Allen et al., 2014). Last, VIF and tolerance values showed there was no multicollinearity.
Since all the assumptions were either met or taken care of, a multiple regression could be performed using the PROCESS macro. Since the moderator is categorical, dummy coding
was needed, which has been done automatically by the PROCESS macro. The moderator is selected to be ‘multicategorical’ and the option ‘indicator’ is chosen as a coding system. With dummy coding, one group functions as a reference group, to which the other two groups are compared. Because in this research, the differences between all three groups (e.g. both female, both male, and mixed dyad) are needed, the analysis has been done twice (Alkharusi, 2012). The first time, the male dyad functioned as a reference group for gender composition and the second time this was the mixed dyad. Both main effects and interaction effects were simultaneously tested in the multiple regression analysis with the PROCESS macro (Field, 2013). Other studies into risk-taking have also used dummy coding to compare groups (e.g. Raffaelli & Crockett, 2003). In addition, a regular multiple regression analysis has been performed in SPSS, in order to compute Beta values and check if the results were comparable to the results of the PROCESS macro.
Results Descriptive Statistics
Means, standard deviations and Pearson correlations for the sample are displayed in Table 1. In Table 2.1, 2.2, and 2.3 the means, standard deviations and Pearson correlations of each separate group are displayed (e.g. both male, both female and mixed dyad). Higher scores on risk-taking with a sibling were correlated to higher scores on risk-taking, for both the whole sample and the different groups. Small differences can be seen for the different groups; the correlation is the highest for the male group and the lowest for the female group. Nevertheless, all groups show a medium effect size individually according to Cohen’s criteria, as does the whole sample (Cohen, 1988).
Table 1
Means, Standard Deviations Pearson Correlation Coefficients (N = 405)
Note. * p < .05 ; ** p < .01 Table 2.1
Means, Standard Deviations Pearson Correlation Coefficients of the Both Male Group (N = 129)
Note. * p < .05 ; ** p < .01 Table 2.2
Means, Standard Deviations Pearson Correlation Coefficients of the Both Female Group (N = 98) Note. * p < .05 ; ** p < .01 M SD 1 2 1. Risk-taking in presence of siblings 0.92 0.29 - .572** 2. Risk-taking 0.27 0.76 .572** - M SD 1 2 1. Risk-taking in presence of siblings 0.12 0.22 - .603** 2. Risk-taking 0.29 0.32 .603** - M SD 1 2 1. Risk-taking in presence of siblings 0.08 0.15 - .609** 2. Risk-taking 0.23 0.26 .609** -
Table 2.3
Means, Standard Deviations Pearson Correlation Coefficients of the Mixed Group (N = 178)
Note. * p < .05 ; ** p < .01
Relationship between Risk-Taking with a Sibling and Risk-Taking in General
A significant relation was found between risk-taking with a sibling and risk-taking in general, using a multiple regression analysis (F (1, 404) = 195.87, p < .001). Therefore, this model is useful in predicting levels of risk-taking. Risk-taking with a sibling can account for 32.7 % of the total variance of risk-taking (R2 = .327). The results of this analysis show a significant positive relationship between risk-taking with a sibling and risk-taking in general (B =.41, SE = .04, t = 10.21, p < .001, 95% CI [.333, .492]). This result denotes that higher levels of risk-taking with a sibling is related to higher levels of risk-taking in general. In conclusion, the first hypothesis can be accepted based on this multiple regression analysis, performed with the PROCESS macro. There is a medium effect according to Cohen’s criteria (Cohen, 1988).
Moderation of Gender Composition
Regarding the moderation effect, the multiple regression analysis shows the total model is significant (R2 = .331, F (5, 399) = 39.47, p < .001). However, the interaction effect of gender composition as a moderator is not significant (R2 = .001, F (2, 399) = .35, p =.703). The analysis was done twice to compare all three groups with each other. In Table 3 the results the first multiple regression analysis are displayed, with the male dyad as a reference group. Beta values have been retrieved from a regular multiple regression analysis in SPSS,
M SD 1 2
1. Risk-taking in presence of siblings
0.08 0.16 - .521**
since the PROCESS macro does not display these values. No significant differences between the male dyad in comparison to the mixed and female dyad was found. No interaction effect was found either. The results from a regular multiple regression analysis performed in SPSS were comparable to the results from the PROCESS macro.
Table 3
Results of the multiple regression analysis with risk-taking as an outcome variable and the both male dyad as a reference group for the variable gender composition (N=405)
Note. LLCI = lower level confidence interval; ULCI = upper level confidence interval
The second time the multiple regression was performed, the mixed dyad was used as a reference group. In this analysis the mixed and female group were compared, in addition to the first time the multiple regression was performed. No significant differences between groups were found either, as is displayed in Table 4. Beta values have been retrieved from a regular multiple regression analysis in SPSS, since the PROCESS macro does not display these values. Also, no interaction effect was found, which is in line with the results in Table 3. The results from the regular multiple regression analysis performed in SPSS were comparable
B β SD t p LLCI ULCI
Constant .188 - .024 7.397 .000 .142 .234
1. Risk-taking with a sibling .880 .548 .096 9.220 .000 .693 1.068
2. Male dyad – mixed dyad .010 .018 .031 .334 .739 -.050 .070
3. Male dyad – female dyad -.037 -.055 .036 -1.029 .304 -.107 .034
Interaction 1 x 2 .033 -.013 .147 .226 .821 -.256 .323
to the results from the PROCESS macro.Therefore, the relationship between risk-taking with a sibling and risk-taking in general is not significantly affected by gender composition. In conclusion, the second hypothesis was not accepted.
Table 4
Results of the multiple regression analysis with risk-taking as an outcome variable and the mixed dyad as a reference group for the variable gender composition (N=405)
Note. LLCI = lower level confidence interval; ULCI = upper level confidence interval Discussion
In this study the relationship between risk-taking with a sibling and risk-taking in general in adolescence was studied, as well as to what extent this relationship is moderated by gender composition. Risk-taking was measured with questions about soft drug use, alcohol use, smoking and delinquency, which were all combined in a composite score for risk-taking. First, we hypothesized positive relationship between risk-taking with a sibling and adolescent risk-taking in general. Results showed that higher levels of risk-taking with a sibling are indeed related to higher levels of risk-taking in general. Therefore, the first hypothesis was
B β SD t p LLCI ULCI
Constant .198 - .020 10.108 .000 .160 .237
1. Risk-taking with a sibling .914 .569 .112 8.139 .000 .693 1.134
2. Mixed dyad – male dyad -.010 -.017 .031 -.334 .739 -.070 .050
3. Mixed dyad – female dyad -.047 -.070 .033 -1.413 .159 -.112 .018
Interaction 1 x 2 -.033 -.016 .147 -.226 .821 -.323 .256
accepted. The second hypothesis was that the positive relationship between risk-taking in general and risk-taking with a sibling would be stronger for same sex siblings, and that it would be the strongest for male sibling dyad. No significant differences were found between female and male sibling dyads, female and mixed sibling dyads, and male and mixed sibling dyads. This moderation hypothesis of gender composition was rejected.
The Relationship Between Risk-Taking with a Sibling and Risk-Taking in General The finding that engaging in higher levels of risk-taking with a sibling predict higher levels of risk-taking in general, is in line with previous research. Nevertheless, no research could be found that considered adolescents’ risk-taking with a sibling and risk-taking in general, and therefore, other siblings studies are used to compare. For example, a longitudinal study among adolescents by Trim et al. (2006) showed the alcohol use of the younger sibling could be predicted by the alcohol use of the older sibling. This is also the case for smoking and delinquency (Abderhalden & Evans; Buist, 2010; Von Bothmer, 2002). Nevertheless, these studies only investigated one type of risk-taking behavior, whereas the current study takes more types of risk-taking into consideration. Both have different advantages, since the current study investigates a broader concept, and therefore gives more information, and the other studies focussed on one type of behavior, which makes it more specific. A meta-analysis by Buist et al. (2013) found a better relationship between siblings was related to less
internalizing and externalizing problems. The current study does not look at the closeness between siblings specifically, but at the risk-taking behavior in presence of a sibling, which is also of importance to know. In any case, the results of the current study are also hypothesized by the Cognitive Social Learning Theory (Bandura, 1977). This theory states that people imitate other people in their social environment through observational learning. Besides, the interaction between one and their social environment also shapes behavior. This way
this case risk-taking behavior. These results also substantiate that risk-taking can be influenced the characteristics of the situation (Figner & Weber, 2011). It can namely be concluded that risk-taking in general can partly be explained by risk-taking with a sibling. Therefore, this study supports the fact sibling presence is an important factor to take into consideration when investigating risk-taking in adolescents.
The Role of Gender Composition
Subsequently, a moderation effect of gender composition was expected. The positive relationship between risk-taking with a sibling and risk-taking was expected to be stronger for same gender siblings. Also, the moderation effect would be higher for male sibling dyads, in comparison to female and mixed dyads. However, the results showed no moderation effect of gender composition. No significant differences between the three groups (e.g. both male, both female, and mixed dyads) were found either. Hence, the second hypothesis was rejected.
This result is partly in line with other studies, since these are divided whether there is a moderation effect of gender composition. That is, on the one hand, these results are in line with a number of other studies, which also found no moderation by gender composition (Defoe et al., 2013; Whiteman et al., 2014; Whiteman et al., 2017). However, a study into older and younger sibling delinquency found a moderation of gender composition, namely they found a significant results for same sex pairs (e.g. brother and sister pairs), but not for mixed sex pairs (e.g. older brother/younger sister and older sister/younger brother sibling pairs) (Buist, 2010). For same sex pairs, over-time changes in the older sibling’s delinquency was related to the younger sibling’s delinquency two years later, but also to the change pattern in the younger sibling’s delinquency over time (Buist, 2010). Other studies also showed risk-taking increased when same sex siblings were drinking, whereas they did not find this for mixed sex siblings (Trim et al., 2006, Boyle et al., 2001). For example, longitudinal study by Trim et al. (2006) found alcohol use is influenced by siblings, especially if they are of the
same sex. Even though the current study is about siblings and not peers, sibling studies are scarce, and therefore peer studies are used to compare. Of note, in Trim et al. (2006), only one type of risk-taking behavior was studied, namely alcohol. It could be that for alcohol use gender composition does serve as a moderator, whereas for the other types of risk-taking that we studied (i.e., soft drug use, smoking and delinquency) that is not the case. In other words, it could be that in the current study no significant moderation effect of gender composition was found, because the concept of risk-taking was too broad. However, when researching more specific types of risk-taking (e.g. alcohol use), there could be an effect.
These results of the current study are not in line with the previous mentioned Evolutionary Personality Theory, which considers traits in humans that are caused by adaptive demands of evolutionary history (Holt et al., 2012). According to this theory males are more extravert and therefore show more risk-taking (Nettle, 2006). Based on the Cognitive Social Learning Theory by Bandura (1977), people of the same sex would be more likely to copy and reinforce each other’s behavior, which also conflicts with the results of this study. Nevertheless, the current study does not show any differences between the three groups, which is contradictive with both the Evolutionary Personality Theory and the Cognitive Social Learning Theory. This could be because even though there could be a gender
composition moderation effect, it was not found in the current study, as a result of limitations that will be further discussed in the next part of the discussion.
Strengths, Limitations and Future Directions
This study has strengths, but also limitations, which should be held in mind when considering the results. These limitations could be have been why no moderation effect of gender composition was found in the current study. First, it would be recommended to use more questions to test the levels of risk-taking. In this study, only 4 questions were used to measure risk-taking with a sibling and 7 questions were used to measure risk-taking in
general. For delinquency, only questions about stealing were used. The Cronbach’s alpha’s were both questionable and might have been higher if more questions were added (Santos, 1999; Field, 2013). Therefore, in future research it is recommended to use more
comprehensive questions to measure the different constructs. Second, a disadvantage of running a multiple regression with dummy coding when comparing three groups, is that the analysis had to be done twice. The chance of making mistakes is higher this way.
Furthermore, a questionnaire might not be the best way to get information on this topic. Participants could give social desirable answers, which also leads to an distorted image of the scores (Van de Mortel, 2008). Nevertheless, this was an convenient way to get a lot of
information in a short period of time, since questionnaires can give a lot of information in a short period of time. This made the sample large, which is also of value to the research. Furthermore, there is an overlap in the questions about risk-taking in general and risk-taking with a sibling, since the risk-taking in general is partly with a sibling as well. This can be accounted for by doing an experimental study, where a participant has to do a task with and without a sibling. An experimental study could also consider a causal relation, instead of a correlational one. In a correlational study like the current one, the overlap can be accounted for by asking questions about the taking when not in presence of anyone and the risk-taking in presence of a sibling. However, this could underestimate the level of risk-risk-taking, since risk-taking in adolescence often happens in presence of peer and siblings (Holt et al., 2012). Furthermore, the moderation effect of gender composition could be overshadowed by other factors, such as the closeness between the siblings or the age gap. For example, a meta-analysis by Buist et al. (2013) found less internalizing and externalizing problems in
adolescents if the relationship between siblings was better. Also, they found that a smaller age gap made the relation between sibling conflict and internalizing problems stronger. Last, the data could be influenced by a new law that was introduced in 2014 in the Netherlands, raising
the minimum age for smoking and drinking from 16 to 18 years. This could be of influence to risk-taking behavior, because raising this age could cause adolescents to start smoking and drinking at a later age (Defoe et al., 2016). Since the law was introduced in the middle of the data collection, risk-taking nowadays could be different from the risk-taking levels in this study.
This study is of societal relevance, because as discussed earlier risk-taking is present in many domains, every day and for everyone (Weber et al., 2002). Studying the impact of siblings in risk-taking could provide guidelines for future interventions. Studies show most children spend more time in interaction with their sibling than with anyone else, even more than other peers or their parents (Dunn, 2002; Larson & Verma, 1999). Since this study has shown an impact of siblings on risk-taking, siblings could be included in these interventions. These types of interventions do not exist yet for risk-taking behavior, even though they could be effective. Though, this has proved to be effective for other interventions. For example, a study into sibling inventions with children with autism has shown to be a success in
improving social and communication skills (Banda, 2015).
Additionally, the current research is also of scientific relevance, since there is a research gap for both sibling presence in risk-taking and the influence of sibling gender composition. Most studies focus on (non-biologically related) peers (e.g. Cavalca et al., 2013; Gardner & Steinberg, 2005), instead of siblings, even though siblings could be of importance too. Studies regarding gender composition have been done, but mainly focused on comparison of same sex groups (girl-girl versus boy-boy) instead of mixed groups (e.g., boy-girl versus boy-boy) (e.g. Trim et al., 2006; Rogers et al., 2018). Hence the current study has contributed to filling in that existing gap in research, by considering the differences between not only same sex siblings, both also comparing both same sex dyads to mixed sex siblings.
Besides the quantitative part of this research, motivations for participating in risk-taking behavior were included and described in the participants section of the methods. Even though in this study there is no hypothesis about the motivations of adolescents to participate in risk-taking, future studies could consider to do this. Specifically, for the current study, the participants were asked through open questions why they participated in different types of risk-taking behavior: smoking, soft drug use, and alcohol use. Answers were analysed with coding schemes. In conclusion, being cool and impressing friends was mentioned the most for all categories. Enjoying it (‘lekker’ in Dutch) was mentioned a lot for all types of behavior. Also, having fun was mentioned a lot for drinking alcohol, addiction for smoking, and the feeling for soft drugs. With all of the four types of risk-taking behavior being cool and
impressing others was mentioned, which contributes to the claim that risk-taking happens in a social context and is influenced by the characteristics of the situation. This is in line with the first hypothesis as well. Future research could build on this by focussing more on the social context in which the risk-taking is happening. All of the social factors (e.g. parents, siblings, peers) that could be of influence should be identified, so they can be taken into consideration when looking at risk-taking (Boyer, 2006).
Conclusion
This study contributes to the existing research gap concerning risk-taking for both sibling presence and gender composition. In conclusion, for adolescents risk-taking with a sibling present predicts higher levels of adolescent risk-taking in general. These results
substantiate the Cognitive Social Learning Theory by Bandura (1977) and are also in line with previous research. According to this study, gender composition does not seem to moderate this relationship. This is in line with the conflicting research on gender moderation, where some do find a moderation effect and others do not. No differences between the three groups (e.g. both female, both male and mixed gender) have been found. This is not in line with the
Evolutionary Personality Theory, which states men are more extravert and therefore show more risk-taking. Also, it contradicts with the Cognitive Social Learning Theory, which states that adolescents are more likely to copy and reinforce each other’s behavior if they are of the same sex. Future research should further investigate this gender composition moderation to minimize the conflicting results found across the existing studies. Risk-taking is part of our everyday lives in many domains. It is of importance to know which factors have an effect on risk-taking. Considering the results of this study, it is advised that more interventions
References
Abderhalden, F. P., & Evans, S. Z. (2018). An Examination of Sibling Impact of Frequency and Type of Arrest among Chronic Offenders. Criminology, Criminal Justice, Law & Society, 19, 27-45.
Alkharusi, H. (2012). Categorical variables in regression analysis: A comparison of dummy and effect coding. International Journal of Education, 4(2), 202.
https://doi.org/10. 5296/ije.v4i2.1962
Allen, P., Bennett, K., & Heritage, B. (2014). SPSS statistics version 22: A practical guide. Cengage Learning Australia.
Banda, D. R. (2015). Review of sibling interventions with children with autism. Education and Training in Autism and Developmental Disabilities, 50(3), 303-315.
https://www.jstor.org/stable/24827512?seq=1
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84(2), 191-215. https://doi.org/10.1037/0033-295X.84.2.191 Bandura, A. (2006). Toward a psychology of human agency. Perspectives on psychological
science, 1(2), 164-180. https://doi.org/10.1111/j.1745-6916.2006.00011.x Boyer, T. W. (2006). The development of risk-taking: A multi-perspective review.
Developmental review, 26(3), 291-345. https://doi.org/10.1016/j.dr.2006.05.002 Boyle, M. H., Sanford, M., Szatmari, P., Merikangas, K., & Offord, D. R. (2001). Familial
influences on substance use by adolescents and young adults. Canadian Journal of Public Health, 92(3), 206-209. https://doi.org/10.1007/BF03404307
Bronfenbrenner, U. (1994). Ecological models of human development. Readings on the development of children, 2(1), 37-43.
meta-analysis. Psychological bulletin, 125(3), 367-383. https://doi.org/10.1037/00332909.125.3.367
Brown, B. B., Clasen, D. R., & Eicher, S. A. (1986). Perceptions of peer pressure, peer conformity dispositions, and self-reported behavior among adolescents.
Developmental psychology, 22(4), 521-530. https://psycnet.apa.org/buy/1986-29673001
Buist, K. L., Deković, M., & Prinzie, P. (2013). Sibling relationship quality and
psychopathology of children and adolescents: A meta-analysis. Clinical Psychology Review, 33(1), 97-106. https://doi.org/10.1016/j.cpr.2012.10.007
Buist, K. L. (2010). Sibling relationship quality and adolescent delinquency: A latent growth curve approach. Journal of Family Psychology, 24(4), 400-410.
https://doi.org/10.1037/a0020351
Cavalca, E., Kong, G., Liss, T., Reynolds, E. K., Schepis, T. S., Lejuez, C. W., & Krishnan Sarin, S. (2013). A preliminary experimental investigation of peer influence on risk-taking among adolescent smokers and non-smokers. Drug and alcohol dependence, 129(1-2), 163-166. https://doi.org/10.1016/j.drugalcdep.2012.09.020
Chein, J., Albert, D., O’Brien, L., Uckert, K., & Steinberg, L. (2011). Peers increase adolescent risk taking by enhancing activity in the brain’s reward circuitry. Developmental science, 14(2), F1-F10. https://doi.org/10.1111/j.1467-7687.2010.01035.x
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2.nd ed.). Academic Press.
models of substance use: A review of theory and research on motives for using alcohol, marijuana, and tobacco. Oxford Handbooks Online, 1, 1-53.
https://doi.org/10.1093/oxfordhb/9780199381678.013.017
Defoe, I. N. (2016). The puzzle of adolescent risk taking: An experimental-longitudinal investigation of individual, social and cultural influences (unpublished doctoral dissertation). Utrecht University.
Defoe, I. N., Dubas, J. S., Dalmaijer, E. S., & van Aken, M. A. (2020). Is the peer presence effect on heightened adolescent risky decision-making only present in males?. Journal of youth and adolescence, 49(3), 693-705. https://doi.org/10.1007/s10964-019-01179-9
Defoe, I. N., Dubas, J. S., Figner, B., & van Aken, M. A. (2015). A meta-analysis on age differences in risky decision making: Adolescents versus children and adults. Psychological bulletin, 141(1), 48-84. https://doi.org/10.1037/a0038088
Defoe, I. N., Dubas, J.S., Somerville, L. H., Lugtig, P., & van Aken, M. A. (2016). The unique roles of intrapersonal and social factors in adolescent smoking development. Developmental psychology, 52(12), 2044-2056. https://doi.org/10.1037/dev0000198 Dunn, J. (2002). Sibling relationships. Blackwell handbook of childhood social development.
Blackwell publishers.
Ellemers, N. (2018). Gender stereotypes. Annual review of psychology, 69, 275-298. https://doi.org/10.1146/annurev-psych-122216-011719
Fernie, G., Cole, J. C., Goudie, A. J., & Field, M. (2010). Risk-taking but not response inhibition or delay discounting predict alcohol consumption in social drinkers. Drug and alcohol dependence, 112(1-2), 54-61.
https://doi.org/10.1016/j.drugalcdep.2010.05.011
Figner, B. & Weber, E.U. (2011). Who Takes Risks When and Why? Determinants of Risk Taking. Current Directions in Psychological Science, 20(4), 211-216.
https://doi.org/10.1177/0963721411415790
Figner, B., Mackinlay, R. J., Wilkening, F., & Weber, E. U. (2009). Affective and deliberative processes in risky choice: age differences in risk taking in the Columbia Card Task. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(3), 1-75. https://doi.org/10.1037/a0014983
Fisk, S. (2016). Gender stereotypes, risk-taking, and gendered mobility. Advances in group processes, 33, 179-210. https://doi.org/10.1108/S0882-614520160000033007 Foley, R., & Gamble, C. (2009). The ecology of social transitions in human evolution.
Philosophical Transactions of the Royal Society B: Biological Sciences, 364(1533), 3267-3279. https://doi.org/10.1098/rstb.2009.0136
Furby, L., & Beyth-Marom, R. (1992). Risk taking in adolescence: A decision-making perspective. Developmental review, 12(1), 1-44.
https://doi.org/10.1016/0273-2297(92)90002-J
Gardner, M., & Steinberg, L. (2005). Peer influence on risk taking, risk preference, and risky decision making in adolescence and adulthood: an experimental study. Developmental psychology, 41(4), 625-635. https://doi.org/10.1037/0012-1649.41.4.625
Hayes, A. F. (2019). The PROCESS macro for SPSS and SAS (Version 3.4). http://processmacro.org/download.html
Holt, N. J., Bremner, A., Sutherland, E., Vliek, M., Passer, M., & Smith, R. (2012). Psychology: The science of mind and behavior. McGraw-Hill Education.
of siblings in adolescent delinquency next to parents, school, and peers: do gender and age matter?. Journal of developmental and life-course criminology, 5(2), 220-242. https://doi.org/10.1007/s40865-018-0094-9
Jianakoplos, N. A., & Bernasek, A. (1998). Are women more risk averse?. Economic inquiry, 36(4), 620-630. https://doi.org/10.1111/j.1465-7295.1998.tb01740.x
Katz, L. F., Kramer, L., & Gottman, J. M. (1992). Conflict and emotions in marital, sibling, and peer relationships. In C. U. Shantz & W. W. Hartup (Eds.), Cambridge studies in
social and emotional development. Conflict in child and adolescent development
(p. 122–149). Cambridge University Press.
Kendall, J. (1999). Axial coding and the grounded theory controversy. Western journal of nursing research, 21(6), 743-757. https://doi.org/10.1177/019394599902100603 Koo, T., & Li, M. (2016). A Guideline of Selecting and Reporting Intraclass Correlation
Coefficients for Reliability Research. Journal of Chiropractic Medicine, 15(2), 155– 163. https://doi.org/10.1016/j.jcm.2016.02.012
Larson, R. W., & Verma, S. (1999). How children and adolescents spend time across the world: work, play, and developmental opportunities. Psychological bulletin, 125(6), 701-736. https://doi.org/10.1037/0033-2909.125.6.701
Maxfield, S., Shapiro, M., Gupta, V., & Hass, S. (2010). Gender and risk: women, risk taking and risk aversion. Gender in Management: An International Journal, 25(7), 586-604. https://doi.org/10.1108/17542411011081383
Nettle, D. (2006). The evolution of personality variation in humans and other animals. American Psychologist, 61(6), 622-631. https://doi.org/10.1037/0003-066X.61.6.622 Raffaelli, M., & Crockett, L. J. (2003). Sexual risk taking in adolescence: the role of self
regulation and attraction to risk. Developmental psychology, 39(6), 1036-1046. https://doi.org/10.1037/0012-1649.39.6.1036
Rogers, C. R., McCormick, E. M., van Hoorn, J., Ivory, S. L., & Telzer, E. H. (2018). Neural correlates of sibling closeness and association with externalizing behavior in
adolescence. Social cognitive and affective neuroscience, 13(9), 977-988. https://doi.org/10.1093/scan/nsy063
Santos, J. R. A. (1999). Cronbach’s alpha: A tool for assessing the reliability of scales. Journal of extension, 37(2), 1-5. https://www.joe.org/joe/1999april/tt3.php/journal-current-issue.php
Samek, D. R., Goodman, R. J., Riley, L., McGue, M., & Iacono, W. G. (2018). The developmental unfolding of sibling influences on alcohol use over time. Journal of youth and adolescence, 47(2), 349-368. https://doi.org/10.1007/s10964-017-0703-7 Smith, A. R., Chein, J., & Steinberg, L. (2014). Peers increase adolescent risk taking even
when the probabilities of negative outcomes are known. Developmental psychology, 50(5), 1564-1568. https://doi.org/10.1037/a0035696
Steinberg, L. (2010). A dual systems model of adolescent risk‐taking. Developmental Psychobiology: The Journal of the International Society for Developmental Psychobiology, 52(3), 216-224. https://doi.org/10.1002/dev.20445
Trim, R. S., Leuthe, E., & Chassin, L. (2006). Sibling influence on alcohol use in a young adult, high-risk sample. Journal of studies on alcohol, 67(3), 391-398.
https://doi.org/10.15288/jsa.2006.67.391
Parsons, J. T., Halkitis, P. N., Bimbi, D., & Borkowski, T. (2000). Perceptions of the benefits and costs associated with condom use and unprotected sex among late adolescent college students. Journal of adolescence, 23(4), 377-391.
https://doi.org/10.1006/jado.2000.0326
behavior in humans. Evolutionary Psychology, 6(1), 29-42. https://doi.org/10.1177/147470490800600104
Van de Mortel, T. F. (2008). Faking it: social desirability response bias in self-report research. Australian Journal of Advanced Nursing, 25(4), 40-48.
https://search.informit.com.au/documentSummary;dn=210155003844269;res=IELHE A
Von Bothmer, M. I., Mattsson, B., & Fridlund, B. (2002). Influences on adolescent smoking behaviour: siblings’ smoking and norms in the social environment do matter. Health & Social Care in the Community, 10(4), 213-220.
https://doi.org/10.1046/j.13652524.2002.00363.x
Walters, G. D. (2018). Sibling delinquency as a risk factor for future offending: An exploratory analysis. Youth violence and juvenile justice, 16(4), 343-357. https://doi.org/10.1177/1541204017713255
Walters, G. D. (2020). The sibling effect for delinquency: mediation by proactive criminal thinking and moderation by age. International journal of offender therapy and comparative criminology, 64(2-3), 265-288.
https://doi.org/10.1177/0306624X19872963
Weber, E. U., Blais, A. R., & Betz, N. E. (2002). A domain‐specific risk‐attitude scale:
Measuring risk perceptions and risk behaviors. Journal of behavioral decision making, 15(4), 263-290. https://doi.org/10.1002/bdm.414
Weller, J. A., Levin, I. P., & Bechara, A. (2010). Do individual differences in Iowa Gambling Task performance predict adaptive decision making for risky gains and losses?. Journal of Clinical and Experimental Neuropsychology, 32(2), 141-150.
https://doi.org/10.1080/13803390902881926
behaviors from adolescence to young adulthood: Vertical socialization or bidirectional effects?. New directions for child and adolescent development, 2017(156), 67-85. https://doi.org/10.1002/cad.2019
Whiteman, S. D., Zeiders, K. H., Killoren, S. E., Rodriguez, S. A., & Updegraff, K. A. (2014). Sibling influence on Mexican-origin adolescents' deviant and sexual risk behaviors: The role of sibling modeling. Journal of Adolescent Health, 54(5), 587-592. https://doi.org/10.1016/j.jadohealth.2013.10.004
Willoughby, T., Tavernier, R., Hamza, C., Adachi, P. J., & Good, M. (2014). The triadic systems model perspective and adolescent risk taking. Brain and cognition, 89, 114-115. https://doi.org/10.1016/j.bandc.2013.11.001
Zhang, X., Qu, X., Tao, D., & Xue, H. (2019). The association between sensation seeking and driving outcomes: a systematic review and meta-analysis. Accident Analysis &
