Abstract
Adolescent intentional and unintentional injury is commonly related to involvement in violence and transportation behaviors. While many risk and promotive factors have been identified, a cumulative assessment of such factors is less common, and this has rarely been undertaken with transportation behaviors. The study involved Australian adolescents from high schools with greater than 75% of students from low socioeconomic status (SES) backgrounds, aged 13 to 14 years (n = 826). Findings showed the presence of risk factors increased the odds of engagement in unintentional and intentional injury-risk behavior and the presence of promotive factors decreased the odds, supporting a compensatory model of resiliency. An interaction term of cumulative risk by promotive factors was a significant predictor in logistic regression analyses suggesting a protective-factor model of resiliency also applies. The research has implications for resiliency theory in the field of transportation and adds to the research on the value of compensatory and protective-factor models of resiliency.
Fatal and nonfatal injuries represent a major public health concern. Both have a significant impact on communities, with injuries estimated to represent the reason for 1.8 million emergency department (ED) visits and over 12,000 fatalities in Australia (Australian Institute of Health and Welfare [AIHW], 2015, 2017). For adolescents, violence and transportation are among the more common intentional and unintentional injury-risk experiences. Findings from the U.S. Youth Risk Behavior Surveillance System survey (Kann et al., 2016) of nationally representative high school students showed 20% had ridden in a car driven by someone who had been drinking alcohol in the past 30 days and 22% had been in a physical fight in the preceding 12 months. In Australia, such behaviors are among the leading causes of injury hospitalizations among youth, with the top three reasons related to transportation, falls, and assaults (AIHW, 2012).
This study takes a less commonly used perspective to understand early adolescent injury, examining cumulative or an aggregate of risk and promotive factors (at one time point). The approach suggests an accumulation of factors affects health outcomes and has rarely been studied with regard to transportation behaviors. We include a focus on factors that increase the likelihood of injury-risk behavior but, as is less commonly undertaken, we seek to understand the role of adolescent resources and assets (promotive factors). The term promotive factors is used, in contrast to protective factors, to reflect strengths and resources rather than just absence of risk. Risk factors are those associated with a higher likelihood of a negative outcome, and promotive factors enhance healthy development (Zimmerman et al., 2013).
Low Socioeconomic Status (SES) and Engagement in Violence and Transportation Risks
Promotive factors can be considered as resources that support positive outcomes despite the presence of adversity, such adversity that might be evident in communities with low SES. Adolescents growing up in areas of low SES are at greater risk of poor health outcomes, including injury-related deaths (Lawlor, Sterne, Tynelius, Davey Smith, & Rasmussen, 2006), and they are likely to experience greater adverse childhood experiences (ACE). Hughes et al. (2017) described ACE as direct individual harms (e.g., abuse) and indirect harms experienced in the adolescent’s environment (e.g., family conflict).
Leventhal and Brooks-Gunn (2000) reviewed links between neighborhood residence and adolescent outcomes, and report associations between low SES and experience of violence. Similarly, various measures of SES have been associated with involvement in transportation risk behaviors. Low SES is also associated with a greater likelihood of motor vehicle crashes and greater severity of a crash (Hanna, Taylor, Sheppard, & Laflamme, 2006), with one consideration that the vehicles used in low SES areas may have fewer safety features. Pickett et al. (2012) found that among a representative sample of Canadian adolescents who had not reached licensing age (M age = 13 years), 21% reported riding with a driver who had used alcohol or other drugs and 10% reported driving after drinking or using other drugs. This behavior was associated with being male, living in a rural area, and low SES (self-defined by adolescents with regard to how well-off they believed their family to be). Another study using an alternate measure of SES (i.e., receipt of public assistance) showed an association with likelihood of riding with a drinking driver (Buckley et al., 2017). Furthermore, low SES is also associated with greater cycling and pedestrian injuries (Williams, Currie, Wright, Elton, & Beattie, 1997).
Cohen et al.’s review of ACE highlights individual and school conditions, such as physical exposure (e.g., building conditions) and social experiences (e.g., social capital, crime) that may be associated with low SES and provide the conditions for pathways to later harm. Adolescents who experience adversity inherent in the neighborhood and schools in which they live thus may have a complexity of factors that affect their health behavior. More recent interest in understanding ACE and the complexity of health promotion suggests it is timely to consider both risk and promotive factors and continue to understand the role of cumulative factors associated with unintentional and intentional injuries in areas of adversity.
Resiliency Theory
Resiliency theory provides one method of theoretically integrating clusters of explanatory factors. Resilience can operate to interrupt the trajectory of risk (e.g., low SES) on poor outcomes (e.g., engagement in injury-risk behaviors; Zimmerman et al., 2013). The framework has been proposed to explain the cumulative impact of factors on adolescents’ behavior and, ultimately, their health and well-being. It aligns with other theories, such as general strain theory that recognizes multiple factors at the individual- and contextual-level affect behavior. General strain theory suggests there are multiple forms of stress or strain, including failure to achieve positive goals, presentation of negative stimuli, and removal of positive stimuli that may be managed by an adolescent as criminal behavior (Agnew, 1992). Resiliency theory explicitly makes consideration of resources or assets that promote health despite risk or adversity. Testing the cumulative role of risk and promotive factors rather than a single construct enables a move beyond cataloging factors to identify mechanisms for change. A benefit to understanding a cumulative role is that it has the potential to recognize complexity of an individual’s ecology across domains. Indeed, Ostaszewski and Zimmerman (2006) described research highlighting the intricacies and interactions of risks and resources.
Fergus and Zimmerman (2005) discussed two models of resiliency, including the compensatory model, which suggests that promotive factors counteract (through direct effects) the impact of risk factors. In this model, additional variance in the injury-risk outcome would be explained by the inclusion of promotive factors as predictors in a model that already includes risk factors. In contrast, the protective-factor model suggests that promotive factors moderate or reduce the effects of risk factors on poor adolescent outcomes, and in this case, additional variance would be explained with the inclusion of a risk by promotive factor interaction over and above the variance explained by risk or promotive alone.
Cumulative Risk and Promotive Factors for Violence and Transportation Risks
Stoddard and colleagues (2013) have been one of the few to examine the cumulative effects of risk and promotive factors across multiple domains (individual, peers, family, school, and community), and this study builds on their research. The results of Stoddard et al. (2013) supported the protective-factor model of resiliency, whereby a cumulative measure of promotive factors (including self-efficacy, attitudes, religious involvement, positive peer behavior, living with a parent, and parental monitoring) was found to moderate a cumulative measure of risk (including substance use, delinquency, school failure, gang involvement, negative peer behavior, family conflict, and exposure to community violence) in predicting violent behavior. They also found support for the compensation model of resiliency. While their research was novel, they focused on an older cohort of adolescents and only examined violent behavior, while other adolescent behaviors that have the potential to negatively impact health and well-being, including risky road use, are rarely examined from this perspective.
The current study sought to extend upon such research by examining the impact of cumulative risk and promotive factors on violent behavior and transportation behaviors. Early adolescence represents a developmental period that may be relevant for prevention efforts in being prior to a peak of injury experiences, with those aged 15 to 24 years the most likely to present to the ED (AIHW, 2015). However, there is engagement and experience with injury-risk behaviors in early adolescence. In an Australian sample, Chapman et al. (2011) found 43% of 13- to 14-year-olds had been injured in a fight, 43% injured riding a bicycle, 13% injured as a passenger, and 5% injured driving a car. At this earlier stage of adolescence, risky transportation behaviors may thus reflect cycling, passenger experiences, and, to a lesser extent, driving. There are, however, still clear injury prevention needs around transportation and violence.
We also sought to build on the work of those such as Haegerich, Shults, Oman, and Vesely (2016) who focus on a breadth of promotive factors. Their study examined multiple assets associated with driving after drinking and riding with a driver who has been drinking. While they did not use a cumulative approach or examine risks, they found parental, peer, and school assets were associated with a lower likelihood of engagement in such behaviors. Another example of the examination of cross-domain factors involves a study by Jessor and colleagues (2003), although it should be noted that this study also did not consider factors as cumulative constructs. They looked to explain adolescent’s involvement in a collection of health compromising behaviors focusing on substance use and delinquency (e.g., marijuana use, theft, and aggression). The risk factors were vulnerability (e.g., depression), opportunity (e.g., availability of alcohol at home), and role models (e.g., friends who engage in risk behaviors). Promotive factors included support (e.g., parents interest in their child), role models (e.g., friends who behave prosocially), and social controls (e.g., parent’s disapproval). While not using cumulative measures, a cross-domain collection of risk and promotive factors predicted overall health compromising behavior (compensatory model). In addition, the promotive factors mediated the relationship between risk factors and behavior as evidenced from significant interactions between risks and promotive factor items, also supporting a protective-factor model (Jessor et al., 2003).
As noted, adolescents are not singularly exposed to a single promotive or risk factor and may, for example, have multiple intrapersonal assets and/or may have few social and contextual resources (Zimmerman et al., 2013). Despite this, a number of separate risk and promotive factors have been identified with regard to violence and transportation injury-risk behaviors (e.g., Fagan & Catalano, 2013; Shope & Bingham, 2008; Stone, Becker, Huber, & Catalano, 2012). Research has demonstrated risk factors for these behaviors including poor family dynamics (Espelage, Low, Rao, Hong, & Little, 2014; Nargiso, Friend, & Florin, 2013), having delinquent friends (Goodearl, Salzinger, & Rosario, 2014; Nargiso et al., 2013), limited efficacy (Caprara, Regalia, & Bandura, 2002; Scheier, Botvin, Diaz, & Griffin, 1999), lower self-control (Brody & Ge, 2001; Buckley, Sheehan, & Chapman, 2009; Higgins, Jennings, Tewksbury, & Gibson, 2009), attending a school in which they feel that they do not belong (Aspy et al., 2012; Wilson, 2004), and residing in neighborhoods of inequality (Pabayo, Molnar, & Kawachi, 2014; Topalli et al., 2014). In contrast, research shows health benefits associated with having parents who demonstrate warmth, model safe behavior, monitor their children, and have consistent and clear communication (Brendgen, Vitaro, Tremblay, & Lavoie, 2001; Shope & Bingham, 2008); having prosocial friends (Foshee et al., 2015); feeling a part of a connected and supportive school (Chapman et al., 2011); future orientation; and having positive perceptions of the neighborhood (Stoddard, Zimmerman, & Bauermeister, 2011). This study seeks to address a research gap examining a systematic conceptualization of cumulative risk and promotive factors in unintentional and intentional injury-risk behaviors.
Aims
We seek to identify associations among specific injury-risk behaviors (violence and transportation) with experience of higher cumulative risk factors and lower cumulative promotive factors (testing the compensatory model of resiliency). After accounting for these effects, and in line with our test of the protective-factor model of resiliency, we also examine whether the cumulative promotive factors would moderate the negative effects of cumulative risks on the injury-risk behaviors of violence and transportation.
Method
Research Design
We used baseline data collected from a randomized control trial (RCT, with randomization at the school-level) of a curriculum-based injury prevention program (Skills for Preventing Injury in Youth). Schools were state-funded high schools in urban and regional areas of Queensland. To focus on a high-risk sample, we analyzed data from schools in our RCT in the bottom SES quartile nationally. We use nationally available data reported on schools about students’ SES. The measure is created from enrolled student’s family background (e.g., parents’ education) and school factors (e.g., remoteness; Australia Curriculum, Assessment and Reporting Authority, 2016).
Participants
The sample was 826 students enrolled in Grade 9 from state-funded high schools (n = 11 schools), 47% female, with a mean age of 13.47 years (SD = 0.54). Students predominantly identified as White/Caucasian (70%); 7.7% identified as being of Aboriginal background or Torres Strait Islander background with no students identifying as both. The national average across all ages is 3% (Australian Bureau of Statistics, 2013).
Measures
Injury-risk behaviors
Violence and transportation behaviors were assessed using the Australian Self-Report Delinquency Scale (Mak, 1993), with adjustments made by Western, Lynch-Blosse, and Ogilvie (2003). Participants indicated whether or not they had participated in any of three violence-related behaviors (e.g., taken part in a physical fight, a group fight, or threatened someone) or four transport-related risk behaviors (e.g., ridden in a car with someone who has been drinking, ridden in a car with a dangerous driver, driven a car, driven a bicycle after drinking) in the previous 3 months. Of note, the minimum licensing age for solo driving is 17 years. Responses were dichotomized reflecting engagement in any item.
Risk and promotive factors
Factors included individual characteristics, peer and family factors, and wider school-level influences. Table 1 shows the six risk and seven promotive factors used in the current study, including a sample item and descriptive statistics (mean, standard deviation, and Cronbach’s alpha).
Summary of Risk and Promotive Factors and Behavior Measures.
p < .05. **p < .01. ***p < .001.
Similar to previous research (e.g., Stoddard et al., 2013; Bowen & Flora, 2002), risk and promotive composite factor indices were created by standardizing the initial measures. The upper 16% of the distribution of each scale (>1 standard deviation from the mean) was given a score of 2 and signified a “high” level of the risk or promotive factor, the middle 68% was given a score of 1 and signified an “average” level of the factor, and the lower 16% of the distribution (<1 standard deviation from the mean) was given a score 0 and signified low levels of risk or promotion. The dichotomous variable of household composition was coded with a score of 1 for those who responded “yes” (i.e., live with both biological parents) and a score of 0 for those who responded “no.” The standardized and recoded measures were summed, creating a risk composite score with a potential range of 0 to 12, and a promotive composite score with a potential range of 0 to 13. Thus, each scale or individual measure of risk or promotion as identified in Table 1 equally contributed to the respective total risk or promotion score.
Procedure
Approval for the conduct of this research was obtained from the University Human Research Ethics Committee and Department of Education. Following written principal approval, active parental consent was obtained by having students take an information sheet and consent form home to their parent/guardian. To invite student’s participation in the research, a parent/guardian was required to sign the form and return it to the school with their child. Students were provided with the opportunity to go into a class draw of a gift voucher for returning the sheet (regardless of consent), no other incentives were offered. Written consent for research participation was obtained from 42% of parents, similar to research in other schools in the same jurisdiction in recent years (Hasking, Tatnell, Martin, 2015). Only students available on the day of data collection participated.
Questionnaires were administered during approximately 45-minute classes. Prior to completing the questionnaire, information sheets and consent forms were provided to the students and they provided consent. A set of standardized instructions was read aloud by the researcher. The researcher remained available throughout students’ completion of the questionnaire to answer any student questions. Teachers remained in the classroom but were not involved in administering the questionnaire.
Data Analysis
We ran two sets of four stepwise logistic regression analyses with a different dependent variable (DV) in each set. DVs included (a) violence and (b) transportation injury-risk behaviors. The violence and transportation variables reflect a dichotomized item of reported engagement in at least one of the behaviors (compared with none). The independent variables (IVs) were consistent across the sets of analyses. Participant’s sex (male compared with female) and identification as Aboriginal or Torres Strait Islander Australians (compared with identifying as any other background) were entered in logistic regression of Model 1. The cumulative risk factor was added to these variables in Model 2, and in Model 3, the cumulative promotive factor was included. Model 4 included all of the aforementioned variables along with the interaction term of cumulative risk by cumulative promotive factors. Findings in Model 3 represent a test of the compensatory resilience model and findings in Model 4 test the protective-factor resiliency model. As recommended by Aiken, West, and Reno (1991), the cumulative risk and cumulative promotive factors were centered prior to the interaction term being created.
Results
Descriptive Results
Participants’ mean score on the cumulative risk factor was 6.32 (SD = 1.85, range = 0-12) and on the cumulative promotive factor was 5.37 (SD = 2.51, range = 0-13). With regard to overall injury-risk behavior, 33.9% of students reported engaging in at least one violent behavior (17% in a physical fight, 22% in a group fight, 14% threatening someone). Furthermore, 35.6% of students reported engaging in at least one of the transport-related behaviors (16% riding with a dangerous driver, 17% riding with a driver who had been drinking, 5% cycling after drinking, 22% driving) in the previous 3 months.
Multivariate Models
Results for each set of analyses are presented in Table 2. Each set of analyses has four regression models, with the first model including IVs of sex and Indigenous Australian background. We also ran analyses separately for each specific violence and transportation item. That is, we had separate sets of models with specific behavioral DVs (e.g., of fights, as a passenger of driver who had been drinking). In these models, we did not include Indigenous Australian background (due to some small cell sizes) and we found the same pattern of predictors as observed with the behavioral clusters, of violence and transportation injury risks. There was one exception with regard to the model of riding a bicycle after drinking whereby the interaction term was nonsignificant. Only the overall analyses of violence and transportation injury-risk behaviors are reported.
Sets of Logistic Regression Analyses for Injury-Risk Behaviors: Models of Risk and Promotive Factors.
Note. OR = odds ratio; CI = confidence interval.
p < .05. **p < .01. ***p < .001.
In the set of analyses predicting violence experience, neither demographic was a significant predictor. However, in the set of analyses predicting transport injury-risk, identification as either Aboriginal or Torres Strait Islander Australian was significant in Models 1 and 2 but did not remain so with the addition of promotive factors in the models.
Violence
Model 2 in each set examined the addition of the cumulative risk factor to demographics. Cumulative risk was associated with greater odds of violence (adjusted odds ratio [AOR] = 1.22, p < .001). Model 3 tested the compensatory model and the main effect of the cumulative promotive factor after taking demographics and the cumulative risk factor into account. Having a greater cumulative promotive factor score was associated with lower odds of violent behavior after accounting for the experience of cumulative risk factors and demographics (AOR = .82, p < .001). Model 4 tested the protective-factor model of resiliency, by examining the effect of the cumulative risk by cumulative promotive interaction term after the other factors had been accounted for. The cumulative risk by cumulative promotive interaction term was associated with increased odds of violent behavior (AOR = 1.04, p < .01) after accounting for the effects of demographics, cumulative risk factor, and cumulative promotive factor scores.
Transportation Injury-Risk Behavior
Model 2 again examined the addition of the cumulative risk factor to demographics. The cumulative risk factor was associated with greater odds of a transportation injury-risk behavior (AOR = 1.10, p < .001). Model 3 tested the compensatory model, as per the results for violent behavior; the cumulative promotive factor was also related to transportation injury-risk behavior after accounting for the cumulative risk factor and demographics (AOR = .77, p < .001). Finally, Model 4 tested the protective-factor model of resiliency, by examining the effect of the cumulative risk by cumulative promotive interaction term after the other factors had been accounted within the model. The cumulative risk by cumulative promotive interaction term was associated with increased odds of transport risk behavior (AOR = 1.07, p < .01) after accounting for the effects of the demographics, cumulative risk factor, and cumulative promotive factor.
Discussion
This study explored the relationship between cross-domain cumulative risk and cumulative promotive factors on the injury-risk behavior of a sample of early adolescents attending schools characterized by low SES (among the bottom quartile in the country). Thus, the study provides a further step in understanding the relationship between cumulative risk and promotive factors across multiple domains, in contrast to previous research that more commonly has focused on single factors or single domains to predict health (Zimmerman, 2013). In alignment with previous research, higher levels of cumulative risk factors were found to be associated with greater odds of injury-risk behavior. We also found a direct effect of cumulative promotive factors for violence and transportation behaviors after the effect of the cumulative risk factors was considered, with the interaction terms accounting for further variance.
Resiliency Theory and Cumulative Risk and Promotive Factors
The results of this study support both the compensatory and protective-factor models of resiliency. We found stronger support for the compensatory model after observing little extra (although significant) variance in the models containing the interaction term. Findings thus suggest that there is value in considering the promotive factors as well as the risk context in attempts to reduce violence and transportation risks. The approach of focusing on cumulative factors suggests that regardless of which specific risk or promotive factors present for an adolescent, the compounding or cumulative nature of such factors even at a single time point is pertinent to engagement in injury-risk behavior. Furthermore, the findings show that risk and promotive factors were of greater relevance in predicting injury-risk behaviors than factors such as sex and background that are not amenable to change.
Identifying as Aboriginal or Torres Strait Islander peoples was significant in predicting transportation risk behaviors, including in the model with cumulative risk factors. However, with the inclusion of promotive factors in the model, this was non-significant (although the lower end of the confidence interval approached 1, at .99). We also ran models testing interaction effects of Indigenous Australian by cumulative factor; however, such terms did not add variance.
There is considerable inequity among injury rates more broadly, with fatal injuries more than 3 times greater as a car passenger and 5 times greater as a pedestrian for Indigenous Australians compared with non-Indigenous Australians (Hanlan & Harrison, 2013). This presents clear research needs; however, there is little knowledge of Indigenous youth’s experience of risk and promotive factors including specific factors and understanding validity and reliability of measurement. Furthermore, early adolescence is often a neglected period for road safety, with policy factors typically the focus closer to driving age (Haegerich et al., 2016).
Previous research in other areas of Indigenous adolescents’ well-being speaks to the complexity. For example, Redmond et al. (2016) suggested differing definitions of family and community among Indigenous adolescents. Tomyn, Cummins, and Norrish’s (2015) work suggests a need to understand what is a risk or promotive factor. They compared reports of happiness among at-risk Indigenous adolescents, at-risk non-Indigenous (who had or were likely to disconnect from school), and mainstream school students. At-risk Indigenous adolescents were the happiest with their relationships and were only less likely to be happy than the mainstream group regarding “things that they have” and what may “happen later on in life” (with no difference of happiness with health, achievement, safety, community). In all domains, they were more likely happier than the non-Indigenous adolescents who were also at-risk. Furthermore, Hopkins, Zubrick, and Taylor (2014) found variation in relevant risk and promotive factors depending upon the experience of potential family stressors for Indigenous adolescents. This study highlights the complexity of defining risk and promotive factors within the at-risk context, that is, consideration should be given to defining the adverse events that potentially create vulnerability for adolescents.
There are also likely risk and promotive factors that are not captured in the current study but that may be of greater relevance for Indigenous youth. For example, Homel, Lincoln, and Herd (1999) suggested for violence considering cumulative risk that includes forced removal, dependence, racism, cultural features, and substance use as well as promotive factors that includes cultural resilience, self-esteem, and family and community bonds.
Future research might generally tease out different patterns of experience of risk and promotive factors and test the interactions of such factors across development. A cascade analysis would thus be warranted to examine how ecological levels influence each other over time to inform more tailored interventions (Masten & Cicchetti, 2010). Cumulative risk and promotive factors may also be considered with regard to direct or indirect influence on behavior and could have important implications for policy and allocating limited resources.
Implications for Intervention
The value of a theory, such as resiliency theory, is that it enables program designers and public health educators to develop resources under a unifying theme to study assets and resources in a coherent and comprehensive manner (Zimmerman, 2013). Primarily the findings highlight that there is a need to consider the interaction and accumulation of risk and promotive factors. Such approaches are likely multi-pronged, and in the violence context, reviews suggest that the strongest evidence for behavior change is achieved when designers focus on multiple components (see Cox et al., 2016, for a review of Australian programs). While such approaches may also be initially more expensive than focusing on a single promotive factors or single risk factors, there is potential that they have greater benefit in the longer term. This also speaks to a need for cost-effectiveness research that is aligned with program theory. Given the potential complexity of risk and promotive factors associated with violence and transportation risks, intervention efforts will need to consider how they can be tailored to the relevant mix of specific factors at an appropriate developmental period.
The findings suggest that intervention efforts build and support promotive factors at both an individual and contextual level. This is an approach rarely considered in efforts to reduce transportation risks that more commonly focus on skills training or parenting around learning to drive (Haegerich et al., 2016). Programs such as the Skills for Preventing Injury in Youth program are one of the few exceptions, targeting school connectedness, individual resources, and positive peer behavior and finding reduced passenger harms and few cycling-related injuries (Buckley et al., 2009; Chapman, Buckley, & Sheehan, 2012). More commonly, little focus is provided to preventing transportation harms in early adolescence; our data, however, suggests that there are potential harms and need for intervention.
Limitations and Avenues for Future Research
The research must, however, be viewed in light of study limitations. The research used cross-sectional data and while it covered multiple domains, there is always scope for additional factors to be considered. Additional examples of promotive factors may, for example, include attachment to neighborhood (Park, Lee, Bolland, Vazsonyi, & Sun, 2008) and individual identity within the neighborhood (Caldwell, Kohn-Wood, Schmeelk-Cone, Chavous, & Zimmerman, 2004). Perhaps a more relevant future approach will be to weight risk and promotive factors to create a more nuanced cumulative picture. A longitudinal assessment would provide further clarity and would also better speak to the use of the term “cumulative” which might then reflect the building of promotive assets over time or the compounding of risks over time. This study, however, provides an important first step in highlighting the value in a cumulative cross-domain approach for risk and promotive factors for early adolescents and, in particular, the relevance of resiliency theory with regard to injury.
A further limitation to the study is the lack of specificity in our understanding of SES, and we used constructs of violence and transportation that represent a collection of different experiences in themselves, for example, our transportation behavior variable reflected behaviors as a passenger and unlicensed driver. While we cover multiple components to transportation, future research might examine implications for specific health compromising behaviors alongside specific risk and promotive factors. Furthermore, the low consent rates, while reflective of similar studies in the same jurisdiction (e.g., Martin et al., 2015), necessitate consideration of generalizability.
Conclusion
The research does, however, add to the work on cumulative risk and promotive factors but with an early adolescent sample. It suggests that the field focus on the way in which individual and social factors may aggregate and combine to affect adolescents’ lives. In addition, we importantly test models of resiliency theory for students in low SES areas and apply the theory with regard to transportation injury-risk behaviors. One in every five deaths among adolescents in high income countries are due to road traffic injuries, and thus, better understanding how risk and promotive factors impact on such injuries in this area, as well as violence, will provide relevant safety benefits for adolescents.
Footnotes
Acknowledgements
The authors thank the wider research team and they also thank the staff, students, and parents of the participating schools.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors acknowledge the financial support of the Australian Research Council (ARC-DP, DP110105043) who provided funding for the research.