Age Differences in Risk Taking: The Effect of Physical Risk Exposure on the Link Between Moral Development and Delinquency in Mid- and Late Adolescents
Master Thesis Forensic Child and Youth Care Sciences Graduate School of Child Development and Education University of Amsterdam L.Y. Rezelman 11037717 First supervisor: dr. I.N. Defoe Second supervisor: dr. M.J. Noom Amsterdam (July 2020)
Abstract
Adolescence is often described in the literature as a distinct stage of life in which individuals have an increased chance of risk-taking behavior such as delinquency. However, adolescents do not always take more risks than children or adults in experimental tasks. An explanation for this discrepancy could be that in real life age differences in physical access to risk conducive situations (i.e., physical risk exposure) are present. According to Situational Action Theory, risk exposure is an important predictor for delinquency along with moral development. This study therefore examined the moderation effect of physical risk exposure on the link between moral development and delinquency among mid- and late adolescents. Physical risk exposure was hereby assessed through neighborhood quality (with categories advantaged, weak and insufficient-largely insufficient) and availability of a police station. The sample consisted of N = 582 Dutch adolescents ranging from 12 to 18 years old (M = 14.60;
SD = 1.30) who filled in a questionnaire. Results from a hierarchical linear regression analysis
showed that after controlling for age, moral development was a significant predictor for delinquency. Moreover, adolescents with lower levels of moral development engaged in higher levels of delinquency, especially when they were exposed to more risk conducive situations (i.e., weak disadvantaged neighborhood or absence of a police station). The identification of which proxies for physical risk exposure moderate the link between moral development and delinquency, can be a new starting point in effective prevention and intervention programs for juvenile delinquents.
Keywords: moral development, delinquency, physical risk exposure, availability of a police
Age Differences in Risk Taking: The Effect of Physical Risk Exposure on the Link Between Moral Development and Delinquency in Mid- and Late Adolescents
Delinquency among youth is an extensive problem, as it is a burden for the entire society. Committing a crime in adolescence is namely a risk factor to come into contact with criminality again as an adult (Mowen, Brent, & Bares, 2017). Delinquency can be described as a legal term that refers to individuals who commit acts that violate the law (Smith, 2008). It is a regularly occurring form of risk-taking behavior (Junger & Dekovic, 2003). As for juvenile delinquency, the same legal term can be used but then it refers to individuals who are minors, commonly under the age of 18 (Smith, 2008). How is it that one individual follows a path of delinquent behavior while another manages to stay on the right path?
Many studies researched delinquency among youth. For example, back in 1932, Piaget already saw possibilities for clarifying delinquency by studying the moral development of juveniles (Piaget, 1932). Thereafter, several studies began investigating the link between moral development and delinquency. Noteworthy is that several studies found age differences in both moral development (Kohlberg, 1984) and delinquency (Modecki, 2008). Moreover, as seen in the real world, early adolescents engage in less risk behavior than mid- or late adolescents (Johnson & Malow-Iroff, 2008). However, meta-analysis from Defoe, Dubas, Figner and Van Aken (2015) found that early adolescents took more risks than mid/late adolescents on an experimental risk-taking task. A possible explanation for these discrepancies is that different age groups are exposed to various degrees of access to risk conducive situations (i.e., physical risk exposure; Defoe et al., 2015). This physical risk exposure could subsequently strengthen or weaken the link between moral development and delinquency and then functions as a moderator. For this reason, current study aims to investigate the moderation effect of physical risk exposure on the link between moral development and delinquency in mid- and late adolescents.
Moral Development
Moral development is a broad concept that can be divided in both moral emotions and moral cognition. Moral emotions consist of empathy as well as guilt and shame which are self-conscious emotions that provide an individual with instant feedback of certain behavior (Tangney, Stuewig, & Mashek, 2007). Moral cognition consists of moral judgment (Van Vugt et al., 2012). Moral judgment is a cognitive process that provides individuals with the possibility to make decisions with reference to a particular situation, making that decision morally acceptable (Romeral, Fernández, & Fraguela, 2018). An individual’s behavior can solely be considered as moral if it emerges from moral judgment (Gibbs, 2019). The current study only focused on moral cognition (i.e., moral judgment), since that is the part of moral development which is used to measure the construct. Instruments such as the Moral Judgment Interview (MJI; Colby & Kohlberg, 1987) and the Sociomoral Reflection Measure-Short Form (SRM-SF; Gibbs, Basinger, & Fuller, 1992) namely determine an individual’s level of moral development by measuring moral cognition based on Kohlberg’s three levels of moral judgment (Kohlberg, 1984). These levels will be discussed in the next paragraph.
It is Kohlberg (1984) that was the first to describe moral judgment as three different levels, consisting of six chronologically ranked stages. The first level is the preconventional level, which consists of stage one i.e., obedience and punishment and stage two i.e., self-interest (instrumental orientation). In this level, behavior is mostly judged through reward and punishment and an individual is mostly acting out of self-interest. The second level is the conventional level, which consists of stage three i.e., interpersonal agreement (relationships) and stage four i.e., social system orientation (laws and regulations). In this level, interpersonal agreement becomes important and a person behaves according to the law and regulations. The third level is the postconventional level, which consists of stage five i.e., social contracts and stage six i.e., universal ethical principles. In this level, the self is differentiated from rules and
expectations and social contracts and universal ethical principles become of more value. Individuals advance through these levels and stages during life, whereas the first level is accomplished at the age of 9, the second level is accomplished at adolescence (until the age of 18) and the third level is accomplished in adulthood (Kohlberg, 1984). Thus, based on these levels and stages, an individual’s level of moral development is measured. The third and fourth stages are particularly of interest in adolescence.
The Link Between Moral Development and Adolescent Delinquency
Adolescence is often described in the literature as a distinct stage of life in which individuals have an increased chance of risk-taking behavior (Steinberg, 2007; Steinberg, 2008; Johnson & Malow-Iroff, 2008; Zhang, Zhang, & Shang, 2016). This risk-taking behavior causes adolescents to engage in more delinquent behavior. Research showed that delinquency generally increased during adolescence, also early pubertal timing was related to higher levels of delinquency (Negriff, Susman, & Trickett, 2011). An increase in delinquency during adolescence could be attributed to environmental circumstances, peers and spare time activities (Kretschmer, Oliver, & Maughan, 2014). However, it could also be attributed to biological changes and the underdeveloped adolescent brain (Barbot & Hunter, 2012).
Kohlberg’s theory also suggests that delinquency is dependent on age (Kohlberg, 1984). According to this theory, late adolescents will engage less in risk behavior such as delinquency than mid-adolescents because they have a more advanced degree of moral judgment. Namely, the fourth stage (i.e., social system orientation) in contrast to the third stage (i.e., interpersonal agreement; Kohlberg, 1984). Hence, they are further developed in the conventional level. A more advanced degree of moral judgment also includes compassion for the victim, respect for the law, obedience to authority and respect for human rights (Raaijmakers, Engels, & Van Hoof, 2005). Adolescents are expected to move from the first and second stage (i.e., the preconventional level, age 0-9) to the third or fourth stage (i.e., the
conventional level, age 10-18). This development generally takes three to six years, until the age of 18 (Kohlberg, 1984). Adolescents engaging in delinquency could thus be understood as still reasoning on the second stage. Since the transition to the new stage takes three to six years and is different for every individual, both in onset and length (Raaijmakers et al., 2005). A solid view of moral development and its relation to delinquency is part of Kohlberg’s theory of moral judgment (Kohlberg, 1984). More recent views of the moral judgment stages diminish the importance of the three levels and instead highlight the superficial and strong egocentric bias of the infantile stages (one and two) as risk factors for antisocial behavior in adolescence (Stams et al., 2006). Also, research from Kuther and Higgins-D'Alessandro (2000) found that moral judgment was consistently related to delinquency among students from grade 10 to grade 12 (i.e., age 15 to age 17). When students considered delinquency to be a moral issue, engagement in such activity was related to lower levels of postconventional reasoning (Kuther & Higgins-D'Alessandro, 2000). Other studies found similar results (Palmer, 2003; Raaijmakers et al., 2005). Moreover, meta-analysis from Romeral et al. (2018) found more delinquency among youth with lower levels of moral development. Another meta-analysis (Stams et al., 2006) concluded that developmentally delayed moral judgment was strongly associated with juvenile delinquency. Finally, lower levels of moral development were not merely associated with delinquency, but also with recidivism as meta-analysis of Van Vugt et al. (2011) concluded.
Risk Exposure
Contrary to neurological findings and the assumption of Kohlberg’s theory of moral judgment, meta-analysis of Defoe et al., (2015) found that adolescents did not always engage in more taking than children, and that mid/late adolescents did not engage in more risk-taking than early adolescents. These findings are not in line with what is seen in the real world. Namely, early adolescents (age 11-13) engage in less risk behavior than mid/late
adolescents (age 14-19) in real life (Defoe et al., 2015). The question that arises is how the discrepancies of age differences and experimental tasks versus real life can be explained. A possible explanation is that older adolescents are exposed to more risks in real life than younger adolescents. Defoe et al. (2015) namely explained how opportunity factors (i.e., risk exposure) clearly play a role in risky choices that adolescents make. In the real world are age differences in access to risk conducive situations (i.e., physical risk exposure), but in the lab-setting are equal risk opportunities for all age groups (Defoe et al., 2015).
Risk exposure can be distinguished in both social and physical risk exposure (Defoe, Dubas, & Romer, 2019). Social risk exposure refers to heightened risk taking through social influences (e.g., deviant peers, exposure to media that positively represents risk or influences from parents), as mentioned above, physical risk exposure refers to access to risk conducive situations (e.g., physical availability of alcohol at home or no surveillance at a store; Defoe, 2016; Defoe et al., 2015). Other examples of risk exposure are the laws and regulations that provide youth older than eighteen with more opportunities such as legal alcohol age or getting a driver’s license. These examples can be considered as cultural factors, i.e., local government policies (Defoe, 2016). In real life, higher age has greater access to risk conducive situations. The current study will focus on this physical aspect of risk exposure.
A proxy for physical risk exposure could be “disadvantaged neighborhood” or “absence of a police station”. Wikström and Loeber (2000) namely explained how youth living in a disadvantaged neighborhood were more likely to be involved in serious crime because of greater physical exposure to behavior settings involving risk. Examples of this are low-quality homes, vandalized streets and pubs. Other researches pointed out that there was a correlation between disadvantaged neighborhood and youth delinquency (Gold & Nepomnyaschy, 2018; Pinkster & Fortuijn, 2009). Additionally, presence of a police station can have a deterrent effect (Marco, Gracia, & López-Quílez, 2017). So, absence of a police
station could be considered as physical risk exposure because there is less surveillance and more access to risk conducive situations compared to areas that do have a police station. Namely, research showed that theft (i.e., delinquency) corresponded to the location and distance of a police station (Escobar, Rodriguez, & Vasquez, 2019).
The Link between Moral Development, Risk Exposure and Delinquency
A theory that describes the link between moral development, risk exposure and delinquency is the Situational Action Theory (SAT). This theory suggests that an individual’s actions are situational and that the specific action happens as a result from the specific context and individual with its own characteristics (Wikström, 2014). The action is a combination from the interaction between a person’s crime propensity and criminogenic exposure converged in time and space (Wikström, 2019). Crime propensity is the overall tendency of an individual to perceive and opt for crime as an action alternative. This crime propensity is dependent on an individual’s morality (i.e., moral rules and moral emotions) and self-control (Wikström, 2014). Since self-control is the process that manages an individual’s morality (Wikström, 2014), it will not be focused on in the current study. Criminogenic exposure is a setting that is conducive to crime which is determined by the moral context (Wikström, 2014). Examples of this are settings with criminogenic others such as deviant peers or settings with low detection and supervision (Hirtenlehner & Hardie, 2016). So, the criminogenic exposure concept is similar to (physical) risk exposure as mentioned in Defoe et al. (2015).
According to SAT, individuals with poor personal morality and little self-control are more likely to engage in delinquency because they consider crime as an option more often. Likewise, SAT implies that an individual’s level of exposure to settings with a moral context that encourages crime, plays an important part in crime causation (Wikström, 2014). Research of Wikström (2014) investigated these links and found strong support for both implications. Research of Wikström, Mann and Hardi (2018) investigated crime as an outcome of a given
crime propensity, when criminogenic exposed or not. That study found strong support that higher criminogenic exposure caused larger effects of crime propensity on crime. These findings are based on an ongoing longitudinal study that follows a random sample of 716 individuals from age 12 through adolescence and into young adulthood. Participants filled in a questionnaire to measure their level of crime involvement. To measure criminogenic exposure, an index of two scales was used based on the same data. One scale measured time spent in criminogenic places, the other scale measured association with criminogenic peers (detailed information on the study can be found in Wikström et al. (2018)). The current study will not focus on time spent in criminogenic places as measurement of physical risk exposure, but on neighborhood quality and availability of a police station.
Current Study
Little research has been done to investigate the impact of physical risk exposure and its relation to age differences in real life risk behavior (Defoe et al., 2015; 2019). As SAT suggests; risk exposure and moral development are important predictors for delinquency (Wikström, 2014). Hence, the research question for current study is: “Does moral development predict delinquency and does physical risk exposure moderate this link?”. Physical risk exposure will hereby be assessed through neighborhood quality and availability of a police station (Wikström & Loeber, 2000; Escobar et al., 2019). The hypothesis for current study is that adolescents with lower levels of moral development will more likely engage in delinquency, especially when they are exposed to more risk conducive situations (i.e., higher levels of physical risk exposure; Stams et al., 2006; Tarry & Emler, 2007). This is also what SAT suggests (Wikström, 2014). However, SAT includes both social and physical risk exposure. The current study will solely focus on physical risk exposure. The current study will also account for age by controlling for it, since age is shown to predict risk exposure (Defoe et al., 2015), delinquency (Modecki, 2008) and moral development (Kohlberg, 1984).
Method
The current study used data of a previous prospective 3-year longitudinal study in the Netherlands called “The Adolescent Risk Taking (ART) Project”, which started in 2012 (Defoe, Dubas, Somerville, Lugtig, & Van Aken, 2016; Defoe, Dubas, Dalmeijer, & Van Aken, 2019). The ART Project was a research project on adolescent risk-taking in multiple domains. For current study, only data from the second wave was used (detailed information on The ART Project can be found in Defoe et al. (2016)).
Participants
In Wave 2, the overall sample consisted of N = 582 (N = 578 valid cases) Dutch adolescents of which 45.6% was female and 54.4% was male. The range of age was from 12 to 18 years old (M = 14.60; SD = 1.30). The age groups were represented with sample sizes of respectively N = 2, N = 173, N = 80, N = 164, N = 122, N = 35, N = 2 from youngest to oldest (Defoe et al., 2016). The adolescents were either in the second or fourth year of “preparatory middle-level applied education” (i.e., VMBO) or “higher general continued education” (i.e., HAVO). They completed the questionnaire and cognitive tasks during school hours at their schools (Defoe et al., 2016; Defoe et al., 2019). At the start of the study, 93.2% of the adolescents reported that they were born in the Netherlands. Of this percentage, 61.6% identified themselves as Dutch, 9.3% as Turkish of Turkish-Dutch, 7.4% as Surinamese or Surinamese-Dutch, 5.5% as Moroccan or Moroccan-Dutch, and the remaining 16.2% identified with several other ethnicities (Defoe et al., 2016; Defoe et al., 2019). The parents of the adolescents were mostly married or living together (68.4%) and the minority of parents were divorced or separated (24.8%; Defoe et al., 2016; Defoe et al., 2019). At the start of the study, about half of the adolescents were uninformed about the highest degree of completed education of their parents, respectively 44.9% for fathers and 46.5% for mothers. Of all reported educational degrees, 6.7% of mothers and 6.4% of fathers did not complete a
secondary educational degree at all. As for lower- or middle-degree vocational education, 35.8% of mothers and 28.0% of fathers completed this. Finally, 3.8% of mothers and 10.5% of fathers completed a university degree (Defoe et al., 2016; Defoe et al., 2019).
Procedure
Participants were collected through select cluster sampling. Seven high-schools in the Netherlands were recruited during the Fall and Winter of 2013. The schools were divided over six different regions and were ethnically diverse (Defoe et al., 2016; Defoe et al., 2019). The researchers first e-mailed the schools and subsequently called them (Defoe et al., 2016; Defoe et al., 2019). The parents of the adolescents received an information letter about the research project alongside a dissent letter that could be sent back to the schools if parents disapproved of participation (i.e., the researchers used passive consent letters; Defoe et al., 2016; Defoe et al., 2019). Approximately 810 potential participants were available at the beginning of the study. Of these potential participants, 9.75% did not participate because of parental disapproval. The remaining adolescents who did not participate refused to participate because of their own willingness or were not present at the times of the data collection (Defoe et al., 2016; Defoe et al., 2019). Trained research assistants collected the data at schools. They gave both written and verbal instructions to the participants. As a thank you gesture, participants could choose to obtain a chocolate candy worth two euros or have their name entered in a lottery for a chance to win a 50-euro gift card (Defoe et al., 2016; Defoe et al., 2019).
Instruments
Delinquency. Delinquency was measured through four items which were derived from
the International Self-Reported Delinquency Questionnaire (ISRD; Junger-Tas, Terlouw, & Klein, 1994; Junger-Tas, Marshall, & Ribeaud, 2003). The four items covered property crime related to theft (Defoe et al., 2016; Defoe et al., 2019). An example of a question is: “Have you ever stolen a bicycle, scooter or motorcycle?”. The answer-categories for all four items
on this scale were 0 = Never; 1 = Yes, but that was longer than twelve months ago; 2 = Yes, once in the past twelve months; 3 = Yes, twice in the past twelve months; 4 = Yes, three times or more during the past twelve months (Defoe et al., 2019). The mean score for this scale was calculated with a Cronbach’s alpha of .82, which indicated an adequate reliability with a higher mean reflecting higher levels of delinquency.
Moral Development. Moral development was measured through one question, derived
from the Youth Decision Making Questionnaire (YDMQ; Ford, Wentzel, Wood, Stevens, & Siesfeld, 1990). The question concerned a hypothetical dilemma: “Tonight, you are going into
town with your friends. You have already spoken to your friends and everyone has bought something special for this evening. You would also like something nice, but you do not have anything nice to wear. You really do not have any money to buy something new and even if you look at your mother sweetly, she does not want to pay for a new shirt. Nevertheless, you are going to have a look at H&M. There you see a shirt that is perfect for this evening. You are trying it on to see how it fits. Once in the fitting room, the shirt appears to fit perfectly. Then you notice the shirt has no alarm. If you want, you could just keep the shirt on underneath your own clothes and walk out of the store. Then you would have a great shirt for tonight. How likely is it that you will steal the shirt?” Answer categories were formulated as: I
will: 1 = Definitely steal the shirt; 2 = Probably steal the shirt; 3 = Probably not steal the shirt; 4 = Definitely not steal the shirt, with a higher score reflecting higher levels of moral development. This dilemma is a well-defined scenario of social acceptability of right and wrong because the decision may hypothetically result in negative outcomes such as police arrest (Centifanti, Modecki, MacLellan, & Gowling, 2016; Magar, Phillips, & Hosie, 2008). Hence, the participants had to outweigh the possible risks which includes a degree of conflict between opting to socially moral act and the socially careless course of action and can thus be considered as measurement of moral development (Centifanti et al., 2016; Magar et al., 2008).
Physical Risk Exposure. Physical risk exposure was measured through “disadvantaged
neighborhood” and “absence of a police station”. The neighborhood variable was computed from postal codes that the youth participants provided during the start of the study. For each participant, the postal code of the home address was looked up on the internet to observe which neighborhood it is part of. Subsequently, with the website of “Leefbarometer” the extent in which the area could be considered as “disadvantaged” was determined. The website indicated if an area was advantaged or disadvantaged and to what extent (i.e., weak, insufficient or largely insufficient). Advantaged neighborhood was coded as 0 = Advantaged, disadvantaged neighborhood was coded as 1 = Weak, 2 = Insufficient, 3 = Largely Insufficient. For the analysis, neighborhood quality was dummy coded first which was automatically done with a dummy coding tool in SPSS (see details below). Similar strategies were conducted in other researches where data from surveys were paired with neighborhood census-based data using the participants’ postal codes (Blair, Gariépy, & Schmitz, 2015; Slutske, Deutsch, Statham, & Martin, 2015). Moreover, the areas were examined on absence of a police station in the area as an indication of physical risk exposure. Presence of a police station was coded as 0 = Present, absence of a police station was coded as 1 = Absent.
Data Analysis
All statistical analyses were conducted in SPSS Version 26 (IBM, Inc.). First, descriptive statistics of the participants were calculated to give a description of the sample. Second, the regression assumptions for linearity, homoscedasticity, multicollinearity and normality were examined to decide whether regression would be statistically allowed (Agresti & Franklin, 2014). The data met the assumptions for homoscedasticity (based on the median; Brown & Forsythe, 1974) and multicollinearity. The data was not normally distributed but since the sample size was bigger than ten observations per parameter, the results were not impacted and thus still valid (Schmidt & Finan, 2018). Besides, linear regression models are
often robust to assumption violations (Hoffmann & Shafer, 2015). So, despite the violation of the linearity assumption, the analyses were conducted. Then third, a hierarchical multiple linear regression analysis with two-way categorical by continuous interaction was conducted to predict delinquency as a function of moral development and physical risk exposure, controlled for age. This control variable (i.e., age) was first entered into the regression model to determine the proportion of explained variance. By adding the other variables into the model, it could be determined if the variance significantly increased above and beyond the effect of age. In doing so, the control variable was held constant so that an effect could not be attributed to age (Stams et al., 2006) and the risk of a Type I error would decrease (Tarry & Emler, 2007). Subsequently, physical risk exposure and moral development were entered into the model. A possible moderation effect by physical risk exposure on the link between moral development and delinquency was tested by including a two-way categorical by continuous interaction between moral development and physical risk exposure into the model at last (Walters, Kremser, & Runell, 2020; Woerner, Ye, Hipwell, Chung, & Sartor, 2019). This moderation variable was computed after examining for the regression assumptions. Moderation is a statistical technique used in regression that aims to examine how the influence of an independent variable on the dependent variable may alter depending on the value of another, moderating variable (Hayes, 2017). This analysis was conducted two times; one time with neighborhood quality and one time with availability of a police station as a proxy for physical risk exposure. Neighborhood quality was hereby dummy coded with advantaged neighborhood as the reference category (i.e., coded as 0). Additionally, the insufficient and largely insufficient disadvantaged neighborhood categories were joined together as “Insufficient-Largely Insufficient” because the latter category merely contained eight participants (Siebert & Siebert, 2017). Regarding availability of a police station, presence of a police station was set as the reference category (i.e., coded as 0).
Results Descriptive Statistics
The mean score for moral development was 3.59 (SD = .82) and the mean scores for physical risk exposure were .29 (SD = .61) for neighborhood quality and .64 (SD = .48) for availability of a police station. The mean score for delinquency was .18 (SD = .52). The correlations between the variables of interest are shown in Table 1 and Table 2.
Table 1
Correlations of Variables of Interest with Neighborhood Quality
1 2 3 4 5 1. Age - 2. Moral Development -.007 - 3. Neighborhood – W -.022 -.062 - 4. Neighborhood – I-LI .152** .032 -.117* - 5. Delinquency .107* -.262*** .016 .056 -
Note. W = Weak, I-LI = Insufficient-Largely Insufficient; * p < .05, ** p < .01, *** p < .001.
Table 2
Correlations of Variables of Interest with Availability of a Police Station
1 2 3 4 1. Age - 2. Moral Development -.007 - 3. Police Station .022 -.107* - 4. Delinquency .107* -.262** .023 - Note. * p < .05, ** p < .001.
Moral Development by Neighborhood Quality Interaction
For the first block of the hierarchical linear regression analysis, the control variable age was analyzed (see Table 3). The results of the first block revealed the model to be statistically significant with a significant proportion of explained variance (R² = .011, p = .034). These results indicate that age accounts for 1.1% of the variation in delinquency.
Table 3
Summary of Hierarchical Linear Regression Analysis for Variables predicting Delinquency with the Moral Development by Neighborhood Quality Interaction Effect
Model, Step and Predictor Variable B β t sr² R² ∆R² ∆F
Model 1 .011 .011 4.52* 1. Age .044 .107* 2.13 .01 Model 2 .082 .071 9.89*** 1. Age .040 .098* 1.98 .01 2. Moral Development -.161 -.262*** -5.36 .07 3. Neighborhood – W .012 .008 .16 .00 4. Neighborhood – I-LI .091 .051 1.02 .00 Model 3 .098 .016 3.32* 1. Age .039 .095 1.93 .01 2. Moral Development -.120 -.196** -3.47 .03 3. Neighborhood – W -.015 -.010 -.19 .00 4. Neighborhood – I-LI .092 .051 1.04 .00
5. Moral Development * Neighborhood – W -.201 -.141* -2.57 .02 6. Moral Development * Neighborhood – I-LI -.056 -.025 -.49 .00
For the second block analysis, the predictor variables moral development, weak disadvantaged neighborhood and insufficient-largely insufficient disadvantaged neighborhood were entered into the model. Both age and moral development were significant predictors for delinquency. The results of the second block revealed the model to be statistically significant with a significant proportion of explained variance (R² = .082, p < .001). The R² change value was .071, which indicates that the addition of the variables moral development and neighborhood to the first block model accounts for 7.1% of the variation in delinquency.
For the third block analysis, the interaction variables between moral development and weak disadvantaged neighborhood and insufficient-largely insufficient disadvantaged neighborhood were entered into the model (Boisclair Demarble, Fortin, D’Antono, & Guay, 2020; Fusilier & Durlabhji, 2008). These results only identified weak disadvantaged neighborhood as a significant moderator (β = -.141, 95% C.I. (-.354, -.047) p = .010), which indicates that adolescents with lower levels of moral development engaged in higher levels of delinquency, especially when they lived in a weak disadvantaged neighborhood. These results did not identify insufficient-largely insufficient disadvantaged neighborhood as a significant moderator. However, the interaction term was left in the model because it was part of the hypothesis. Namely, based on a well-founded theoretical expectation (i.e., SAT) an interaction effect was expected to be present. In this case, the insignificant interaction term is informative and contributes to answering the research question (Aguinis & Gottfredson, 2010; Osborne, 2008). Overall, the results of the third block revealed the model to be statistically significant (R² = .098, p = .037). The R² change value was .016, which indicates that the addition of the interaction variables accounts for 1.6% of the variation in delinquency, in addition to the first and second block models.
To interpret the significant interaction effect between moral development and weak disadvantaged neighborhood, a multiple regression analysis was conducted for all three
neighborhood categories separately (Table 4). The results showed that the negative link between moral development and delinquency was strongest for weak disadvantaged neighborhood compared to the other categories.
Table 4
Summary of Regression Analyses for Advantaged, Weak and Insufficient-Largely Insufficient Disadvantaged Neighborhood Separately
Delinquency
Advantaged Weak I-LI
Predictor Variable B β R² ∆F B β R² ∆F B β R² ∆F
Model 1 .050 7.60* .355 12.64* .046 .75
1. Age .033 .087 .060 .137 .069 .095
2. MD -.120 -.205* -.319 -.670* -.173 -.184
Note. MD = Moral Development, I-LI = Insufficient-Largely Insufficient; ∆F is change in F; * p <.001.
Moral Development by Availability of a Police Station Interaction
To assess whether moral development predicts delinquency and whether the availability of a police station moderates this link, a second hierarchical linear regression analysis was conducted. For the first block analysis, the control variable age was again analyzed with the same statistical values as in the first regression analysis.
For the second block analysis, the predictor variables moral development and availability of a police station were entered into the model. Table 5 shows that both age and moral development were significant predictors for delinquency. Overall, the results of the second block revealed the model to be statistically significant with a significant proportion of explained variance (R² = .080, p < .001). The R² change value was .068, which indicates that
the addition of the variables moral development and availability of a police station to the first block model accounts for 6.8% of the variation in delinquency.
Table 5
Summary of Hierarchical Linear Regression Analysis for Variables predicting Delinquency with the Moral Development by Availability of a Police Station Interaction Effect
Model, Step and Predictor Variable B β t sr² R² ∆R² ∆F
Model 1 .011 .011 4.52* 1. Age .044 .107* 2.13 .01 Model 2 .080 .068 14.33** 1. Age .044 .106* 2.16 .01 2. Moral Development -.161 -.262** -5.34 .07 3. Police Station -.008 -.008 -.16 .00 Model 3 .091 .011 4.85* 1. Age .043 .103* 2.12 .01 2. Moral Development -.032 -.052 -.48 .00 3. Police Station .001 .001 .024 .00
4. Moral Development * Police Station -.163 -.235* -2.20 .01
Note. ∆R² is change in R², ∆F is change in F; * p < .05, ** p <.001.
For the third block analysis, the interaction variable between moral development and availability of a police station was entered into the model. The interaction variable was found to be statistically significant (β = -.235, 95% C.I. (-.308, -.017) p = .028). The results of the third block revealed the model to be statistically significant with a significant proportion of explained variance (R² = .091, p = .028). The R² change value was .011, which indicates that
the addition of the interaction variable accounts for 1.1% of the variation in delinquency, in addition to the first and second block models. Taken together, consistent with the hypothesis, these results showed that adolescents with lower levels of moral development engaged in higher levels of delinquency, especially when they did not have a police station present in their neighborhood.
To interpret the significant interaction effect between moral development and availability of a police station, a multiple regression analysis was conducted for both presence and absence of a police station separately (Table 6). The results showed that the negative link between moral development and delinquency was strongest for absence of a police station.
Table 6
Summary of Regression Analyses for Presence and Absence of a Police Station Separately
Delinquency Presence Absence Predictor Variable B β R² ∆F B β R² ∆F Model 1 .045 2.97 .117 16.98** 1. Age .081 .208* .023 .054 2. Moral Development -.034 -.046 -.195 -.336** Note. ∆F is change in F; * p <.05, ** p <.001. Discussion
The aim of current study was to investigate the moderation effect of physical risk exposure on the link between moral development and delinquency among mid- and late adolescents. Physical risk exposure was hereby assessed through the proxies neighborhood quality and availability of a police station. These two physical risk exposure proxies were
measured in separate regression analyses. Based on SAT and other researches, the hypothesis for current study was that adolescents with lower levels of moral development will more likely engage in delinquency, especially when they are exposed to more risk conducive situations (i.e., higher levels of physical risk exposure; Stams et al., 2006; Tarry & Emler, 2007; Wikström, 2014). The current study controlled for age in order to ensure that significant findings could not be attributed solely to age, which is typically a predictor for delinquency during adolescence (Stams et al., 2006). In addition to age being a significant predictor, the current study showed that moral development was also a significant predictor for delinquency. The results further showed that for neighborhood quality, only weak disadvantaged neighborhood was a significant moderator. Insufficient-largely insufficient disadvantaged neighborhood was not significant and therefore did not moderate the link between moral development and delinquency. These categories were compared to the reference category advantaged neighborhood. Availability of a police station was also a significant moderator. These results signify that adolescents with lower levels of moral development engaged in higher levels of delinquency, especially when they were exposed to more risk conducive situations i.e., weak disadvantaged neighborhood or absence of a police station.
Neighborhood Quality as a Moderator
Considering neighborhood quality as a proxy for physical risk exposure, the findings are in line with the definition (lack of) of exposure to risk conducive situations (Defoe et al., 2019). Although, neighborhood quality on its own was not a significant predictor for delinquency in the current study. Other research showed how low neighborhood quality (i.e., disadvantaged neighborhood) did predict adolescent delinquency (Gold & Nepomnyaschy, 2018; Kim & Glassgow, 2018; Wikström & Loeber, 2000). For instance, research of Gold and Nepomnyaschy (2018) found that physical neighborhood disorder (e.g., garbage, broken glass and graffiti) predicted early adolescent delinquency. Moreover, that study found that
experiencing physical neighborhood disorder for a longer time (i.e., in multiple waves because data was longitudinal) was associated with increased early adolescent delinquency. Despite not finding a main effect for neighborhood quality on delinquency in current study, the first finding from research of Gold and Nepomnyaschy (2018) could be compared to the current study. Few studies were found that investigated the specific same moderation effect.
The finding of current study that weak disadvantaged neighborhood moderated the link between moral development and delinquency further demonstrates that adolescents with lower levels of moral development engaged in higher levels of delinquency, especially when they lived in a weak disadvantaged neighborhood. This finding is in line with SAT. For example, research investigating SAT showed that social and physical risk exposure strengthened the link between adolescents with lower levels of moral development and higher levels of delinquency (Wikström, 2014; Wikström et al, 2018). This was measured through time spent in criminogenic places such as areas with poor collective efficacy, a city or local centre area and the engagement in unstructured and unsupervised activities with peers (Wikström, 2014; Wikström et al, 2018). SAT emphasizes that crime is always an outcome of an individual’s crime propensity and criminogenic exposure. It is hereby crucial that the higher an individual’s crime propensity, the more vulnerable that individual is to criminogenic exposure (Wikström et al, 2018). Current study is of added value to studies that researched SAT, because SAT does not differentiate between social and physical risk exposure.
However, it is important to note that these findings suggest that only weak disadvantaged neighborhood (but not insufficient-largely insufficient disadvantaged neighborhood) functions as physical risk exposure, since only that type of neighborhood moderated the link between moral development and delinquency. The following could perhaps explain this discrepancy. When adolescents live in a weak disadvantaged neighborhood, the neighborhood is often fine to live in except for some services or qualities
(Haynie, Silver, & Teasdale, 2006). However, the circumstances in insufficient and largely insufficient disadvantaged neighborhoods are often worse (Haynie et al., 2006). For instance, the degree of safety which can be related to delinquency, is often low in such neighborhoods (Gold & Nepomnyaschy, 2018). Because delinquency is a lot more common in those neighborhoods, the discrepancy between the current living situation and physical risk exposure is limited. Physical risk exposure is then not a self-contained characteristic of the neighborhood, but it identifies the neighborhood as a whole (Mair, Kaplan, & Everson-Rose, 2012). Therefore, perhaps this is the reason why insufficient-largely insufficient disadvantaged neighborhood did not interact with moral development and delinquency. However, the above-mentioned statement is speculative and should be interpreted with caution. The current study is of added value to the aforementioned studies (Gold & Nepomnyaschy, 2018; Kim & Glassgow, 2018; Wikström & Loeber, 2000) because it measured neighborhood quality as a moderating predictor for delinquency instead of a self-contained predictor. Hence, it is important that neighborhood quality is taken into account when measuring the link between a predictor and delinquency.
Availability of a Police Station as a Moderator
Considering availability of a police station as a proxy for physical risk exposure, the findings are also in line with the definition (lack of) of exposure to risk conducive situations (Defoe et al., 2019). Although, availability of a police station on its own was not a significant predictor for delinquency in the current study. Other research showed how police presence did predict delinquency (Alvarado Laguna, Valencia Sandoval, & Iturralde Mota, 2019; Escobar et al., 2019; Marco et al., 2017; Sohn, 2016). Research namely showed that locations closest to a police station were less likely to be the scene of crimes because presence of a police station in the neighborhood can have a deterrent effect (Marco et al., 2017) and discourages people from committing a crime (Sohn, 2016). So, absence of a police station can be in line
with the (fewer levels of) physical risk exposure concept as described in Defoe et al. (2019), because there is less surveillance and more access to risk conducive situations compared to areas that do have a police station. Moreover, it can be in line with criminogenic exposure from SAT, because the environment enforces the breaking of some rule(s) of law more compared to areas that do have a police station (Wikström, 2014). Hence, these findings could explain why not having a police station present in the neighborhood strengthened the link between moral development and delinquency. The finding that availability of a police station moderated the link between moral development and delinquency means that adolescents with lower levels of moral development engaged in higher levels of delinquency, especially when they did not have a police station present in their neighborhood.
The current study is of added value to the aforementioned studies (Alvarado Laguna et al., 2019; Escobar et al., 2019; Marco et al., 2017; Sohn, 2016), because current study measured availability of a police station as a moderator on the link between moral development and delinquency. No similar studies were found that investigated this particular moderation effect. This effect indicates that presence of a police station can have a deterrent and discouraging effect on people with lower levels of moral development which withholds them from committing a crime, because the likelihood and fear of getting caught are bigger with presence of a police station (Marco et al., 2017). This way, the availability of a police station can strengthen or weaken the effect of moral development on delinquent behavior. Strengths, Limitations and Future Directions
The findings of current study give new insights into the moderating role of physical risk exposure on the link between moral development and delinquency. However, current study did not measure physical risk exposure directly, but proxies were used to measure the construct. Using proxies was needed because the data for current study could not be collected due to the circumstances around COVID-19. For this reason, current study used data of an
existing dataset that was not necessarily designed to answer the research question. Using these proxies for physical risk exposure can be considered as a methodological limitation, since disadvantaged neighborhood and absence of a police station are quite general and broad concepts. However, disadvantaged neighborhood and absence of a police station are closely related to physical risk exposure as Defoe et al. (2019) concluded and both proxies were found to be a significant moderator in current study. At the same time, as for future research, it is recommended to measure physical risk exposure more directly without using proxies for it. This way, the research validity and reliability will increase (Agresti & Franklin, 2014).
Another methodological limitation from current study is that moral development was measured through one question. This was also again a consequence of using an existing dataset that was not designed to answer the current research question. Using just one question to measure a construct makes it harder to draw any conclusions and to generalize it, which is at the expense of the validity of research (Agresti & Franklin, 2014). Moreover, the question did not directly measure moral development as measurements like the MJI (Colby & Kohlberg, 1987) and the SRM-SF (Gibbs et al., 1992) do. However, as mentioned earlier, the dilemma could be considered as measurement of moral development because the participants had to outweigh the possible risks which includes a degree of conflict between opting to socially moral act and the socially careless course of action (Centifanti et al., 2016; Magar et al., 2008). Besides, despite using just one question, a significant effect for moral development was found. Nevertheless, for future research, it is recommended to use more questions to measure moral development and to use questions that directly measure the construct. By doing this, the research validity and reliability will increase (Agresti & Franklin, 2014).
Lastly, a methodological limitation from current study is the violation of the linearity assumption. Although the results of current study can be used to draw conclusions about the adolescents in the sample, this regression model cannot be generalized to adolescents beyond
the sample (Field, 2013). Hence, any statements made about the moderating effect of physical risk exposure on the link between moral development and delinquency in general, should be interpreted with caution. The current study did not correct for non-linearity because the possibilities for correcting for violated assumptions are somewhat restricted and also have their own limitations (Field, 2013).
Notwithstanding, despite these limitations, current study focused on a unique construct (i.e., physical risk exposure) since little research has been done to test this construct. For instance, SAT suggests that delinquency is a result of the interaction between a person’s crime propensity (i.e., morality and self-control) and criminogenic exposure (i.e., risk exposure; Wikström, 2019). However, SAT does not differentiate between social risk exposure and physical risk exposure. The current study only focused on physical risk exposure, which is theorized to be of added value to social risk exposure (Defoe et al., 2019). For example, it is already well-established that social risk exposure is an important predictor for delinquency (Brauer & De Coster, 2015; Ingram, Patchin, Huebner, McCluskey, & Bynum, 2007; Patacchini & Zenou, 2012). This is especially the case in adolescence, since adolescents become more independent from their parents and are more socialized through friends (Collins & Laursen, 2004). Likewise, older individuals have more access to media that positively represents risk (Strasburger, Wilson, & Jordan, 2013). By making a distinction in social and physical risk exposure, problems can be addressed in a more specific manner. This way, it will be easier to find or determine the cause of certain problems. Thereafter, prevention and intervention programs for juvenile delinquents can be specialized and more focused on particular aspects of causes of delinquency or how to address it. This is a purpose future researchers can build on. Moreover, current study used a large sample size of N = 582 participants that was ethnically diverse. These characteristics make the results more reliable and better generalizable, which are important requirements for research (Kothari, 2004).
Conclusion
All in all, the current study showed that in general, adolescents with lower levels of moral development engage in higher levels of delinquency, especially when they are exposed to more risk conducive situations (i.e., higher levels of physical risk exposure), even while controlling for age. Thus, in line with SAT, risk exposure strengthens the link between moral development and delinquency (Wikström, 2014). However, it is important to build on the findings of current study since little is known about physical risk exposure compared to social risk exposure (cf. Defoe et al., 2019). Figuring out if physical risk exposure is as important as social risk exposure in predicting delinquency, can specify the needs for adolescents in preventing or reducing crime. In doing so, problems can be dealt with at its source. Furthermore, consistent with more studies, moral development is a predictor for delinquency. Taking this into account, youth with lower levels of moral development are an at-risk group in society to become delinquent, particularly when they experience higher levels of physical risk exposure. Hence, more attention must be paid to them specifically so that prevention programs can be used for those youths to prevent them from becoming delinquent. In conclusion, these findings are useful for guiding the development of prevention and intervention programs for juvenile delinquents or at least be a starting point which future research can build on. It is important to pay attention to and care for youth who engage in delinquency because they are one’s children and therefore one’s future.
References
Agresti, A., & Franklin, C. (2014). Statistics: The Art and Science of Learning from Data. (New International Edition) Boston, MA: Pearson Education.
Aguinis, H., & Gottfredson, R. K. (2010). Best‐Practice Recommendations for Estimating Interaction Effects using Moderated Multiple Regression. Journal of Organizational
Behavior, 31(6), 776-786. doi:10.1002/job.686
Alvarado Laguna, E., Valencia Sandoval, K., & Iturralde Mota, O. (2019). Impact of the Social Environment on the Theft of Microenterprises in the Metropolitan Area of Monterrey. Economia Socieda y Territorio, 19(59), 1035-1058. doi:10.22136/est20191271
Barbot, B., & Hunter, S. R. (2012). Developmental Changes in Adolescence and Risks for Delinquency. In Handbook of Juvenile Forensic Psychology and Psychiatry (pp. 11-34). Boston, MA: Springer.
Blair, A., Gariépy, G., & Schmitz, N. (2015). The Longitudinal Effects of Neighbourhood Social and Material Deprivation Change on Psychological Distress in Urban, Community-Dwelling Canadian Adults. Public Health, 129(7), 932-940. doi:10.1016/j.puhe.2015.05.011
Boisclair Demarble, J., Fortin, C., D’Antono, B., & Guay, S. (2020). Gender Differences in the Prediction of Acute Stress Disorder from Peritraumatic Dissociation and Distress among Victims of Violent Crimes. Journal of Interpersonal Violence, 35(5-6), 1229-1250. doi:10.1177/0886260517693000
Brauer, J. R., & De Coster, S. (2015). Social Relationships and Delinquency: Revisiting Parent and Peer Influence during Adolescence. Youth & Society, 47(3), 374-394. doi:10.1177/0044118X12467655
Brown, M. B., & Forsythe, A. B. (1974). Robust Tests for the Equality of Variances. Journal
of the American Statistical Association, 69(346), 364-367.
doi:10.1080/01621459.1974.10482955
Centifanti, L. C. M., Modecki, K. L., MacLellan, S., & Gowling, H. (2016). Driving Under the Influence of Risky Peers: An Experimental Study of Adolescent Risk Taking.
Journal of Research on Adolescence, 26(1), 207-222. doi:10.1111/jora.12187
Colby, A., & Kohlberg, L. (1987). The Measurement of Moral Judgement (Vol.1). Cambridge, England: Cambridge University Press.
Collins, W. A., & Laursen, B. (2004). Changing Relationships, Changing Youth: Interpersonal Contexts of Adolescent Development. The Journal of Early
Adolescence, 24(1), 55-62. doi:10.1177/0272431603260882
Defoe, I. N. (2016). The Puzzle of Adolescent Risk Taking: An Experimental-Longitudinal
Investigation of Individual, Social and Cultural Influences (Doctoral Dissertation,
Utrecht University). Retrieved from https://dspace.library.uu.nl/handle/1874/342331 Defoe, I. N., Dubas, J. S., & Romer, D. (2019). Heightened Adolescent Risk-Taking? Insights
from Lab Studies on Age Differences in Decision-Making. Policy Insights from the
Behavioral and Brain Sciences, 6(1), 56-63. doi:10.1177/2372732218801037
Defoe, I. N., Dubas, J. S., Dalmaijer, E. S., & Van Aken, M. A. (2019). Is the Peer Presence Effect on Heightened Adolescent Risky Decision-Making Only Present in Males?
Journal of Youth and Adolescence, 49(3), 693-705. doi: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. doi: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. doi:10.1037/dev0000198
Escobar, C. A. M., Rodriguez, C. M. C., & Vasquez, A. R. (2019). Spatial and Temporal Analysis of Cell Phone Theft, Pereira, Risaralda, 2018. Revista Logos Ciencia &
Tecnología Journal, 11(2), 167-175. doi:10.22335/rlct.v11i2.810
Field, A. (2013). Discovering Statistics using IBM SPSS Statistics. (4th ed.). Washington, DC: Sage.
Ford, M. E., Wentzel, K. R., Wood, D., Stevens, E., & Siesfeld, G. A. (1990). Processes Associated with Integrative Social Competence: Emotional and Contextual Influences on Adolescent Social Responsibility. Journal of Adolescent Research, 4(4), 405-425. doi:10.1177/074355488944002
Fusilier, M., & Durlabhji, S. (2008). Predictors of Student Internet Use: Data from Three Developing Countries. Learning, Media and Technology, 33(1), 59-69. doi:10.1080/17439880701868903
Gibbs, J. C. (2019). Moral Development and Reality: Beyond the Theories of Kohlberg,
Hoffman, and Haidt. (4th ed.). New York City, NY: Oxford University Press.
Gibbs, J.C., Basinger, K.S., & Fuller, D. (1992). Moral Maturity: Measuring the Development
of Sociomoral Reflection. New Jersey, NY: Erlbaum.
Gold, S., & Nepomnyaschy, L. (2018). Neighborhood Physical Disorder and Early Delinquency among Urban Children. Journal of Marriage and Family, 80(4), 919-933. doi:10.1111/jomf.12487
Hayes, A. F. (2017). Introduction to Mediation, Moderation, and Conditional Process
Analysis: A Regression-Based Approach. New York City, NY: The Guilford
Haynie, D. L., Silver, E., & Teasdale, B. (2006). Neighborhood Characteristics, Peer Networks, and Adolescent Violence. Journal of Quantitative Criminology, 22(2), 147-169. doi:10.1007/s10940-006-9006-y
Hirtenlehner, H., & Hardie, B. (2016). On the Conditional Relevance of Controls: An Application of Situational Action Theory to Shoplifting. Deviant Behavior, 37(3), 315-331. doi:10.1080/01639625.2015.1026764
Hoffmann, J. P., & Shafer, K. (2015). Linear Regression Analysis: Assumptions and
Applications. Provo, UT: NASW Press.
Ingram, J. R., Patchin, J. W., Huebner, B. M., McCluskey, J. D., & Bynum, T. S. (2007). Parents, Friends, and Serious Delinquency: An Examination of Direct and Indirect Effects among At-Risk Early Adolescents. Criminal Justice Review, 32(4), 380-400. doi:10.1177/0734016807311436
Johnson, P. B., & Malow-Iroff, M. S. (2008). Adolescents and Risk: Making Sense of
Adolescent Psychology. Westport, CT: Praeger Publishers.
Junger, M., & Dekovic, M. (2003). Crime as Risk-Taking: Co-Occurrence of Delinquent Behavior, Health Endangering Behaviors and Problem Behaviors. In C. L. Britt & M. R. Gottfredson (Eds.), Control Theories of Crime and Delinquency (pp. 213-248). New Brunswick, NY: Transaction Publishers.
Junger-Tas, J., Marshall, I. H., & Ribeaud, D. (2003). Delinquency in an International
Perspective: The International Self-Reported Delinquency Study (ISRD). The Hague,
The Netherlands: Kugler Publications.
Junger-Tas, J., Terlouw, G. J., & Klein, M. W. (Eds.). (1994). Delinquent Behavior among
Young People in the Western World: First Results of the International Self-Report Delinquency Study. Amsterdam, The Netherlands: Kugler Publications.
Kim, S., & Glassgow, A. E. (2018). The Effect of Father's Absence, Parental Adverse Events, and Neighborhood Disadvantage on Children’s Aggression and Delinquency: A Multi-Analytic Approach. Journal of Human Behavior in the Social Environment, 28(5), 570-587. doi:10.1080/10911359.2018.1443866
Kohlberg, L. (1984). Essays on Moral Development: Vol. 2. The Psychology of Moral
Development. San Francisco, CA: Harper & Row.
Kothari, C. R. (2004). Research Methodology: Methods and Techniques. New Delhi, India: New Age International.
Kretschmer, T., Oliver, B. R., & Maughan, B. (2014). Pubertal Development, Spare Time Activities, and Adolescent Delinquency: Testing the Contextual Amplification Hypothesis. Journal of Youth and Adolescence, 43(8), 1346-1360. doi:10.1007/s10964-013-0074-7
Kuther, T. L., & Higgins-D'Alessandro, A. N. N. (2000). Bridging the Gap between Moral Reasoning and Adolescent Engagement in Risky Behavior. Journal of Adolescence,
23(4), 409-422. doi:10.1006/jado.2000.0328
Magar, E. C., Phillips, L. H., & Hosie, J. A. (2008). Self-Regulation and Risk-Taking.
Personality and Individual Differences, 45(2), 153-159.
doi:10.1016/j.paid.2008.03.014
Mair, C., Kaplan, G. A., & Everson-Rose, S. A. (2012). Are there Hopeless Neighborhoods? An Exploration of Environmental Associations between Individual-Level Feelings of Hopelessness and Neighborhood Characteristics. Health & Place, 18(2), 434-439. doi:10.1016/j.healthplace.2011.12.012
Marco, M., Gracia, E., & López-Quílez, A. (2017). Linking Neighborhood Characteristics and Drug-Related Police Interventions: A Bayesian Spatial Analysis. ISPRS International
Modecki, K. L. (2008). Addressing Gaps in the Maturity of Judgment Literature: Age Differences and Delinquency. Law and Human Behavior, 32(1), 78-91. doi:10.10007/s10979-007-9087-7
Mowen, T. J., Brent, J. J., & Bares, K. J. (2017). How Arrest Impacts Delinquency over Time Between and Within Individuals. Youth Violence and Juvenile Justice, 16(4), 358-377. doi:10.1177/1541204017712560
Negriff, S., Susman, E. J., & Trickett, P. K. (2011). The Developmental Pathway from Pubertal Timing to Delinquency and Sexual Activity from Early to Late Adolescence.
Journal of Youth and Adolescence, 40(10), 1343-1356.
doi:10.1007/s10964-010-9621-7
Osborne, J. W. (Ed.). (2008). Best Practices in Quantitative Methods. Washington, DC: Sage Publications.
Palmer, E. J. (2003). An Overview of the Relationship between Moral Reasoning and
Offending. Australian Psychologist, 38(3), 165-174.
doi:10.1080/00050060310001707177
Patacchini, E., & Zenou, Y. (2012). Juvenile Delinquency and Conformism. The Journal of
Law, Economics, & Organization, 28(1), 1-31. doi:10.1093/jleo/ewp038
Piaget, J. (1932). The Moral Judgment of the Child. Glencoe, Scotland: Free Press.
Pinkster, F. M., & Fortuijn, J. D. (2009). Watch Out for the Neighborhood Trap! A Case Study on Parental Perceptions of and Strategies to Counter Risks for Children in a Disadvantaged Neighborhood. Children's Geographies, 7(3), 323-337. doi:10.1080/14733280903024498
Raaijmakers, Q., Engels, R., & Van Hoof, A. (2005). Delinquency and Moral Reasoning in Adolescence and Young Adulthood. International Journal of Behavioral
Romeral, L. F., Fernández, J. S., & Fraguela, J. A. G. (2018). Moral Reasoning in Adolescent Offenders: A Meta-Analytic Review. Psicothema, 30(3), 289-294. doi:10.7334/psicothema2017.378
Schmidt, A. F., & Finan, C. (2018). Linear Regression and the Normality Assumption.
Journal of Clinical Epidemiology, 98, 146-151. doi:10.1016/j.jclinepi.2017.12.006
Siebert, C. F., & Siebert, D. C. (2017). Data Analysis with Small Samples and Non-Normal
Data: Nonparametrics and Other Strategies. New York, NY: Oxford University Press.
Slutske, W. S., Deutsch, A. R., Statham, D. J., & Martin, N. G. (2015). Local Area Disadvantage and Gambling Involvement and Disorder: Evidence for Gene-Environment Correlation and Interaction. Journal of Abnormal Psychology, 124(3), 606. doi:10.1037/abn0000071
Smith, C. A. (2008). Juvenile Delinquency: An Introduction. Prevention Researcher, 15(1),
3-7. Retrieved from
https://web-a-ebscohost- com.proxy.uba.uva.nl:2443/ehost/pdfviewer/pdfviewer?vid=1&sid=04db0c43-3b84-4eee-b0f7-80bcc455c209%40sessionmgr4007
Sohn, D. W. (2016). Residential Crimes and Neighbourhood Built Environment: Assessing the Effectiveness of Crime Prevention through Environmental Design (CPTED).
Cities, 52, 86-93. doi:10.1016/j.cities.2015.11.023
Stams, G. J., Brugman, D., Deković, M., Van Rosmalen, L., Van Der Laan, P., & Gibbs, J. C. (2006). The Moral Judgment of Juvenile Delinquents: A Meta-Analysis. Journal of
Abnormal Child Psychology, 34(5), 692-708. doi:10.1007/s10802-006-9056-5
Steinberg, L. (2007). Risk Taking in Adolescence: New Perspectives from Brain and Behavioral Science. Current Directions in Psychological Science, 16(2), 55-59. doi:10.1111/j.1467-8721.2007.00475.x
Steinberg, L. (2008). A Social Neuroscience Perspective on Adolescent Risk-Taking.
Developmental Review, 28(1), 78-106. doi:10.1016/j.dr.2007.08.002
Strasburger, V. C., Wilson, B. J., & Jordan, A. B. (2013). Children, Adolescents, and the
Media. Washington, DC: Sage Publications.
Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). Moral Emotions and Moral Behavior.
Annual Review of Psychology, 58, 345-372.
doi:10.1146/annurev.psych.56.091103.070145
Tarry, H., & Emler, N. (2007). Attitudes, Values and Moral Reasoning as Predictors of Delinquency. British Journal of Developmental Psychology, 25(2), 169-183. doi:10.1348/026151006X113671
Van Vugt, E., Asscher, J., Hendriks, J., Stams, G. J., Bijleveld, C., & Van der Laan, P. (2012). Assessment of Moral Judgment and Empathy in Young Sex Offenders: A Comparison of Clinical Judgment and Test Results. International Journal of Offender Therapy and
Comparative Criminology, 56(7), 986-996. doi:10.1177/0306624X11420083
Van Vugt, E., Gibbs, J., Stams, G. J., Bijleveld, C., Hendriks, J., & Van der Laan, P. (2011). Moral Development and Recidivism: A Meta-Analysis. International Journal of
Offender Therapy and Comparative Criminology, 55(8), 1234-1250.
doi:10.1177/0306624X11396441
Walters, G. D., Kremser, J., & Runell, L. (2020). Moderating the Bullying Perpetration– Delinquency Relationship with Parental Support and Knowledge: A Prospective Analysis of Middle School Students. Aggressive Behavior, 46(3), 1-10. doi:10.1002/ab.21885
Wikström, P. O. H. (2014). Why Crime Happens: A Situational Action Theory. Analytical
Sociology: Actions and Networks, 74-94. Retrieved from
Wikström, P. O. H. (2019). Situational Action Theory: A General, Dynamic and Mechanism-Based Theory of Crime and its Causes. In Handbook on Crime and Deviance (pp. 259-281). 2nd Edition. New York, NY: Springer.
Wikström, P. O. H., & Loeber, R. (2000). Do Disadvantaged Neighborhoods Cause Well‐Adjusted Children to Become Adolescent Delinquents? A Study of Male Juvenile Serious Offending, Individual Risk and Protective Factors, and Neighborhood Context. Criminology, 38(4), 1109-1142. doi:10.1111/j.1745-9125.2000.tb01416.x Wikström, P. O. H., Mann, R. P., & Hardie, B. (2018). Young People’s Differential
Vulnerability to Criminogenic Exposure: Bridging the Gap between People-and Place-Oriented Approaches in the Study of Crime Causation. European Journal of
Criminology, 15(1), 10-31. doi:10.1177/1477370817732477
Woerner, J., Ye, F., Hipwell, A. E., Chung, T., & Sartor, C. E. (2019). Relational Peer Victimization Interacts with Depression Severity to Predict the Timing of Alcohol Use Initiation in Adolescent Girls. Alcoholism: Clinical and Experimental Research, 44(1), 255-263. doi:10.1111/acer.14241
Zhang, L., Zhang, C., & Shang, L. (2016). Sensation-Seeking and Domain-Specific Risk-Taking Behavior among Adolescents: Risk Perceptions and Expected Benefits as Mediators. Personality and Individual Differences, 101, 299-305. doi:10.1016/j.paid.2016.06.00
