Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Apr 1.
Published in final edited form as: J Sch Health. 2009 Apr;79(4):160–168. doi: 10.1111/j.1746-1561.2009.00385.x

Social environments and physical aggression among 21,107 students in the United States and Canada

William Pickett 1,2, Bruce Simons-Morton 3, Suzanne Dostaler 1,2, Ronald J Iannotti 3
PMCID: PMC2658746  NIHMSID: NIHMS53901  PMID: 19292848

Abstract

Background

Physical aggression is an important issue in North American populations. The importance of student social environments in the occurrence of physical aggression requires focused study. In this study, reports of physical aggression were examined in relation to social environment factors among national samples of students from Canada and the United States.

Methods

Students in grades 6–10 from the US (n=14,049) and Canada (n=7,058) who had participated in the Health Behaviour in School-aged Children Survey were studied. Rates of student physical aggression were compared between the two countries. School, family, socioeconomic, and peer-related factors were considered as potential risk factors. A simple social environment risk score was developed using the US data and was subsequently tested in the Canadian sample.

Results

Risks for physical aggression were consistently higher among US vs. Canadian students, but the magnitude of these differences was modest. The relative odds of physical aggression increased with reported environmental risk. To illustrate, US boys in grades 6 to 8 reporting the highest social-environment risk score (5+) experienced a relative odds of physical aggression 4.02 (95% CI 2.7–5.9) times higher than those reporting the lowest score (adjusted OR for risk scores 0 through 5+: 1.00, 1.19, 2.10, 2.01, 3.71, 4.02; ptrend<0.001).

Conclusions

Unexpectedly, rates of physical aggression and associations between social environments and student aggression were remarkably similar in Canada and the United States. Family, peer, and school social environments serve as risk or protective factors, with significant cumulative impact on physical aggression in both countries. Given the observed high rates and the many negative effects of aggression on long-term health, school policies aimed at the reduction of such behavior remain a clear priority.

Keywords: adolescents, aggression, fighting, social environments, violence

INTRODUCTION

Once seen as the exclusive purview of the criminal and juvenile justice systems, aggression is now recognized as a public health problem.1 There is no single problem that is more prevalent and perhaps more vexing to adolescents, parents, and school officials.2,3 Among adolescents, aggression typically manifests itself in the form of physical fighting and physical bullying.4–6 involvement in aggression is associated in turn with injury and negative social consequences.1,7

Physical aggression has been studied in past epidemiological analyses.8 In one cross-national survey, rates of engaging in at least one fight in the previous 12 months among 11–15 year olds ranged from 37% in Finland to over 60% in Scotland, Israel, Hungary and other East European countries; rates in Canada and the US were 48%.9 Rates of aggression are generally higher among males and minority populations.1,10,11 In regional and national surveys of adolescents in middle and high schools, 20–36% of US adolescents reported being in a physical fight in the past 12 months, with higher national rates observed among males (43%), Black (43%) and Hispanic (41%) populations, and older grades (e.g., 44% vs. 29% in grades 12 and grade 9, respectively).1,5,6,12 Most physical fights do not require medical attention,12,13 however, physical fighting is associated with an increased likelihood of future injury,9 mental health problems,7 substance use,1,14 school adjustment problems,1,15 and violent crime.16

Within the behavioral sciences, persistent aggressive behavior among adolescents has been considered from the perspectives of temperament, development, and environment.17,18 Clearly, some youth are both aggressive and poorly socialized.1 Although these youth represent a small proportion of the population, their “early onset” behaviors account for a disproportionate amount of aggression and violence.1,4 Most aggressive behavior is episodic and transient and associated with the interaction between adolescent development and environmental factors.1 Self-control develops gradually during adolescence19 and, in the absence of the socializing influence of attentive adults, minor interpersonal conflicts can escalate into physical violence. Variability in size and maturity of adolescents contributes to aggressive behavior29 along with established risk factors: socio-economic status; antisocial attitudes and behavior; harsh, abusive, or lax parenting; poor adjustment to school; and affiliation with anti-social peers.1,16,21 At a community level, exposure to violence, crime, and drugs in neighborhoods and schools are potential determinants.22 Conversely, school and community connectedness, authoritative parenting, and positive relationships with pro-social peers are considered to be protective factors.1,16,22,23

The United States and Canada are two countries with many shared values. However, there are important differences between these countries in terms of the prevalence of juvenile and adult violent crime. Violent crime and juvenile incarceration rates are both markedly higher in the United States.24 One important step in understanding these observed differences is to examine potential causes of physical aggression which may be precursors to more serious forms of violence. Factors characterizing social environments are of particular interest because they are broad indicators, reflect policy, and may be amenable to change.

Adolescent health and risk behaviors are affected by social influences in the home, peer-group, and school, as well as family socio-economic status,23 and risks associated with these social-environmental contexts are potentially cumulative.25 Risk factors for aggressive behavior may represent a single pattern or it is possible that these risk factors are independent and have a cumulative effect.26,27 A coherent pattern or syndrome would suggest that interventions need to simultaneously address a complex array of risk factors.

Study Aim

Physical aggression and social-environmental factors potentially associated with its occurrence were studied among samples of students from the United States and Canada. School, family, socioeconomic, and peer-related factors were considered as potential risk factors. A simple social environment risk score was developed using the US data and was subsequently tested in the Canadian sample. Cross-country differences and similarities were identified to evaluate the consistency of observed associations, with the ultimate goals of evaluating the protective effects of social relationships and informing prevention efforts for physical aggression at the school and societal levels in both countries.

METHODS

Health Behaviour in School-Aged Children

Analyses were conducted with data from the Health Behaviour in School-Aged Children Survey (HBSC), a cross-national survey of adolescent health behavior conducted every four years under the auspices of the World Health Organization.28 Canada has participated in this international survey since 1989 while the US became involved in 1998. The current analyses include national HBSC surveys in the US and Canada conducted in 2001/02.

Study Population and Procedures

Anonymous school-based surveys were conducted during the 2001/02 academic year according to the common HBSC research protocol.28 National estimates were developed for 11 to 15 year old children in both countries. In Canada, classes within schools were selected using weighted probability techniques to ensure that students from different geographic areas, religious affiliations and languages of instruction were equally likely to be sampled. Analogous sampling procedures were followed in the United States, although children from designated minority groups (African American, Latin American) were over-sampled to provide more accurate population estimates of these groups. Statistical criteria specify that the national samples were sufficient to provide confidence intervals of ±3% for representative estimates with sample design effects no more than 1.4 times greater than would be obtained from a simple random sample.

Primary Measures

Variables used in this study were obtained from the self-report questionnaire consisting of 122 core questions as well as optional school setting items that were available in both countries. Individual items and factor analytically derived scales were used to describe and model relationships between physical aggression and perceived social environments.

Physical Aggression

Participants were asked to report two measures of physical aggression: 1) the frequency of engagement in physical fighting in the previous 12 months;12 and 2) the frequency of engagement in physical bullying of another student(s) in the previous couple of months.29 The physical fighting measure was originally developed for the Youth Risk Behavior Survey which reported a high test-retest reliability for this item (Kappa: 68.2%).13 The physical bullying measures were originally conceived by Olweus, who also reported high internal consistency reliabilities at the classroom level (0.80 to 0.90) during developmental analyses. 29 A composite measure of physical aggression, defined as engagement in at least two episodes of physical fighting in the previous 12 months and/or reporting the perpetration of at least 2 to 3 episodes per month of physical bullying, comprised the primary outcome.

Perceived Social Environment

Both factor analytically derived scales and individual items were used as measures of perceived social environment. These were examined as individual risk factors and later combined into a summary environmental risk score.

Measures involving scales

HBSC contains a series of items describing perceptions about home, peer groups and school.28,30,31 These were rated by students using five-point Likert-type scales, with response options of (1) strongly agree through (5) strongly disagree. Specific questions assessed student opinions surrounding parental support, perceptions of safety at school, peer group cohesion at school, fairness of school rules, and school atmosphere. To derive scales from these items, maximum likelihood factor analyses were initially run with oblique rotations using SPSS (version 12.0 for Windows). Model fit was determined using the Scree plot, the interpretability of the pattern matrix, and the value of the root mean square error of approximation (RMSEA). Following the exploratory factor analyses, confirmatory factor analyses with the maximum likelihood method of estimation were run in LISREL (version 8.0). Two scales were constructed: 1) parental support; and 2) peer support at school. Participant scores for these scales were categorized into three levels (high, medium, low levels of support) based upon the mean values of included items. Test-retest reliability of the items that contributed to these scales is reported to be high.28,31 Post-hoc reliability analyses based on the current data yielded a Cronbach’s alpha >0.50.

Measures involving individual items

School environment was assessed by the item how do you feel about school at present?” Test-retest reliability of this item is high.28,31 Available responses were combined into three categories: 1) high- I like it a lot or a bit; 2) medium - I don’t like it very much; 3) low- I don’t like it at all. Socio-economic status was assessed using responses to an existing perceived family wealth item: how well off do you think your family is?28 Past analyses of this item suggest that it is well understood by school children and also correlates strongly with indicators of health and wellbeing, although formal reliability statistics are not available.32 Responses were combined into 3 categories: 1) high- very or quite well off; 2) medium – average; 3) low – not very or not at all well off.

Social Environment Risk Score

Based upon a priori theory,33 we combined the four measures of social environment (parental support, peer support, school environment, socio-economic status) into an additive social environment risk score. For each measure contributing to the summary score, values were assigned to each category as follows: 0 for high, 1 for medium, and 2 for low. With four contributing variables, the range of possible values for the summary score was 0 through 8, where 0 represented the highest possible level of support and 8 the lowest possible level.

Analyses

The two national samples were described demographically. Descriptive analyses were used to estimate the prevalence of student physical aggression stratified by country, grade level (6 to 8; 9 to 10), sex, and the four individual measures of perceived social environment (socio-economic status, parental support, peer support, school environment). Logistic models were then used to estimate the relative odds of physical aggression for the four individual measures of perceived social environment. All models were developed within strata defined by country, grade level and sex. We then applied the environmental risk score to examine the cumulative additive effects of the four social contexts on the relative odds of physical aggression. Ethnicity was forced into each model as a potential confounder based upon a priori assumptions. To account for the sample design and clustering at the classroom level, the regression estimates reflect error inflation by a design effect of 1.4. All analyses were performed in SPSS (version 12.0, Chicago, IL, USA).

Approval to conduct the survey and related research was obtained from the ethics review boards associated with Queen’s University and the National Institute of Child Health and Human Development. The international HBSC protocol was approved by the World Health Organization Europe.

RESULTS

Samples

Participants in the 2001/02 HBSC survey included students from 169 schools (Canada) and 340 schools (United States). In Canada, 74.2% of the students selected for the study completed the questionnaire, and their demographic profile was representative of Canadians in the same age range.34 In the United States, 82.0% of those selected completed the questionnaire; while the primary sample is representative of a cross-section of US school children there is an over-sampling of minority students. Complete responses to items about aggression were obtained from 14049 (94.8%) students in the United States and 7058 (97.5%) students in Canada; the demographic profiles of these respondents are provided in Table 1. Because of missing responses on select items, 6613 (91.4%) Canadian and 12955 (87.4%) American records contained complete data on key exposures, outcomes and covariates and were included in the final regression models.

Table 1.

Demographic profile of national samples of adolescents in the 2002 HBSC Surveys

UNITED STATES CANADA
n=14049 n=7058
BOYS n=6645 GIRLS n=7404 BOYS n=3247 GIRLS n=3811
Grades 6–8 - n (%) 4244 (63.9) 4711 (63.6) 2233 (68.8) 2477 (65.0)
Grades 9–10 - n (%) 2401 (36.1) 2693 (36.4) 1014 (31.2) 1334 (35.0)
Mean age in years (SD) 13.8 (1.5) 13.7 (1.5) 13.7 (1.5) 13.7 (1.5)
Ethnicity - n (%)
 White 3448 (51.9) 3730 (50.4) 2769 (85.3) 3295 (86.5)
 African American/Black 1208 (18.2) 1569 (21.2) 57 (1.8) 51 (1.3)
 Hispanic/Latin American 1273 (19.2) 1429 (19.3) 12 (0.4) 24 (0.6)
 Asian 297 (4.5) 301 (4.1) 199 (6.1) 245 (6.4)
 Native/Aboriginal 272 (4.1) 262 (3.5) 81 (2.1) 71 (1.9)
 Other/Unknown 147 (2.2) 113 (1.5) 106 (3.2) 116 (3.0)
Open in a new tab

Physical Aggression and Social Environment

Prevalence of engagement in physical aggression varied by grade level and sex in both countries, with highest rates reported in boys vs. girls, and young vs. older children (Table 2). Higher rates of physical aggression were associated with lower perceived environmental support in a dose-related fashion. This relationship was found for each of the four social environment measures, socioeconomic status, parental support, peer support, and school environment, and was observed for every demographic group except for a curvilinear relationship between aggression and parental support in Canadian boys in grades 9 to 10. The strength and consistency of the observed relationship was further illustrated in logistic regression analyses (Table 3). The relative odds of aggression rose significantly in association with lower perceived levels of socioeconomic status, parental support, peer support, and school environment; with the single exception noted above, the pattern of risk gradients remained consistent across age and gender. The observed risk gradients were strong in all demographic groups, but especially pronounced among Canadian girls in grades 6–8.

Table 2.

Prevalence of reported aggressive behavior among adolescents in the United States and Canada by grade level, sex, and perceived social environment

Measure of Social Environment Proportion reporting aggressive behavior in previous 12 months
U.S.A.
Canada
U.S.A.
Canada
Boys grades 6–8 Boys grades 9–10
Total n,% 4244* 32.8 2233* 29.9 2401* 29.8 1014* 24.5
SES High 2453 31.0 1292 29.1 1304 27.8 535 21.9
Medium 1166 35.8 603 34.3 803 31.6 375 25.9
Low 335 43.0 212 34.4 190 39.5 89 32.6
Parental support High 2962 28.9 1736 29.4 1546 26.8 693 24.5
Medium 814 38.5 385 34.8 487 30.4 238 19.7
Low 373 49.9 88 46.6 320 42.2 72 36.1
Peer support High 2207 27.8 1315 27.8 1078 25.2 544 21.0
Medium 1290 34.1 679 31.8 871 29.6 372 25.8
Low 584 46.9 207 48.3 381 42.5 87 37.9
School environment High 2823 27.3 1428 26.4 1441 24.5 605 19.5
Medium 856 40.8 507 35.3 553 30.0 261 28.7
Low 519 49.1 272 47.4 381 49.1 141 36.9
Girls grades 6–8 Girls grades 9–10
Total n, % 4711* 17.0 2477* 12.8 2693* 14.6 1334* 10.9
SES High 2632 14.0 1310 10.5 1268 12.5 685 9.6
Medium 1399 19.4 788 16.0 1119 15.5 546 11.4
Low 442 29.4 207 19.8 237 22.4 87 18.4
Parental support High 3482 14.5 1943 10.4 1815 12.8 922 8.9
Medium 821 19.1 405 18.3 560 15.2 295 12.9
Low 345 34.8 106 35.8 299 22.7 114 21.9
Peer support High 2730 14.0 1580 10.5 1362 11.5 792 9.1
Medium 1401 17.1 696 15.2 1008 15.8 432 12.0
Low 467 31.5 183 23.5 298 23.5 106 17.9
School environment High 3563 13.8 1867 10.1 1841 11.7 920 8.5
Medium 783 23.5 439 18.2 598 17.9 283 11.3
Low 328 36.3 146 29.5 234 28.6 123 28.5
Open in a new tab
*

total response for aggressive behavior, small variation in number of missing for different social environments

Table 3.

Relative odds of aggressive behavior among adolescents in the United States and Canada by perceived social environment

Measure of social environment Odds of reporting aggressive behavior in previous 12 months
U.S.A.
Canada
U.S.A.
Canada
OR 95% CI OR 95% CI OR 95% CI OR 95% CI
Boys grades 6–8 Boys grades 9–10
SES High 1.00 1.00 1.00 1.00
Medium 1.24 (1.01–1.52) 1.27 (0.95–1.71) 1.20 (0.92–1.57) 1.25 (0.81–1.92)
Low 1.68 (1.21–2.32) 1.28 (0.86–1.98) 1.70 (1.09–2.64) 1.73 (0.90–3.42)
Parental support High 1.00 1.00 1.00 1.00
Medium 1.54 (1.23–1.93) 1.28 (0.96–1.78) 1.19 (0.87–1.63) 0.76 (0.46–1.26)*
Low 2.45 (1.80–3.32) 2.10 (1.15–3.84) 2.00 (1.41–2.83) 1.74 (0.85–3.55)*
Peer support High 1.00 1.00 1.00 1.00
Medium 1.34 (1.09–1.65) 1.21 (0.92–1.61) 1.25 (0.94–1.65) 1.31 (0.85–2.02)
Low 2.29 (1.76–2.98) 2.43 (1.60–3.69) 2.19 (1.55–3.08) 2.31 (1.18–4.51)
School environment High 1.00 1.00 1.00 1.00
Medium 1.84 (1.47–2.29) 1.52 (1.12–2.06) 1.32 (0.97–1.79) 1.66 (1.04–2.66)
Low 2.58 (1.97–3.36) 2.52 (1.73–3.65) 2.97 (2.14–4.12) 2.41 (1.38–4.20)
Girls grades 6–8 Girls grades 9–10
SES High 1.00 1.00 1.00 1.00
Medium 1.49 (1.17–1.89) 1.63 (1.13–2.35) 1.29 (0.94–1.79) 1.20 (0.72–2.01)
Low 2.56 (1.85–3.55) 2.12 (1.23–3.62) 2.02 (1.24–3.29) 2.11 (0.93–4.88)
Parental support High 1.00 1.00 1.00 1.00
Medium 1.39 (1.06–1.84) 1.92 (1.28–2.88) 1.22 (0.83–1.80) 1.51 (0.86–2.69)
Low 3.14 (2.24–4.40) 4.79 (2.65–8.66) 2.00 (1.31–3.06) 2.88 (1.43–5.78)
Peer support High 1.00 1.00 1.00 1.00
Medium 1.27 (0.99–1.63) 1.53 (1.06–2.20) 1.45 (1.04–2.02) 1.37 (0.81–2.32)
Low 2.82 (2.07–3.86) 2.62 (1.54–4.44) 2.37 (1.53–3.69) 2.18 (1.01–4.73)
School environment High 1.00 1.00 1.00 1.00
Medium 1.92 (1.47–2.51) 1.98 (1.33–2.95) 1.65 (1.16–2.35) 1.38 (0.76–2.53)
Low 3.56 (2.53–5.01) 3.71 (2.16–6.37) 3.03 (1.95–4.73) 4.29 (2.27–8.12)
Open in a new tab
*

p for trend =0.44; all others p<0.05

There was also strong evidence for an increase in the relative odds of physical aggression observed in association with the additive environmental risk score. As the quality of the social environment declined (as indicated by a higher score) the relative odds of aggression generally increased. Again, these gradients were found within each of the four demographic groups under study in both countries (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Relative odds of reporting physical aggression among adolescents in the United States and Canada by grade level, sex, and environmental risk score. Odds ratios adjusted for race/ethnicity. Error bars represent 95% confidence intervals.

DISCUSSION

Consistent with some studies,1,23 evidence on adolescent neurological development,35 and theories predicting improved coping with social conflict as cognitive and interpersonal skills develop,36,37 physical aggression in the United States and Canada decreased with age from grades 6 through 10 and was higher in boys than girls. The literature, however, is inconsistent12,23 and other some theories of problem behavior predict increases in problem behavior during this period.38

A major finding of this cross-national study was that student populations in the United States and Canada had remarkably similar rates of physical aggression. Our inability to distinguish between minor from more major incidents of aggression likely contributed to the lack of observed cross-national differences. However, the consistency of this finding is striking in that no differences were observed across demographic strata. The reported North American rates also appear high when compared with rates for the average European adolescent.39 Negative social environments were associated with higher relative odds of physical aggression in both countries. In addition, each of the four dimensions of the lack of social environment (lower socio-economic status, lower support from parents, lower support from peers, and lower satisfaction with school environments) appeared to be independently associated with higher odds of physical aggression. That is, all four measures of lack of environmental support contributed independently to an index of risk of physical aggression; increases in the relative odds of physical aggression were observed in association with the environmental risk score. The observed risk gradients were strong in all groups examined, and point to the presence of health disparities in both males and females. The observed differences in experiences with physical aggression by the environmental risk score were especially notable among Canadian females.

A major strength of the HBSC is that it permits cross-country comparisons of student health and behavior based upon standard measures and methods. Both national surveys were large, response rates were high, and the vast majority of students who engaged in the survey completed the items of interest. Measures used to define key variables used in this analysis have been tested and used extensively in United States and international analyses, which minimizes the potential for misclassification. These facts speak to the internal validity of the observed analytic findings.

Limitations of the HBSC and these particular analyses warrant recognition. First, although the approach to modeling used here may imply a temporal sequence, similar to other adolescent health surveys,12 HBSC is cross-sectional and any observed associations require confirmation in longitudinal analyses. Standard criteria used in the assessment of causation include consideration of the strength of associations, statistical significance, biological or social plausibility, consistency, dose-response, as well as temporality.40 Our analyses support all but the latter criterion. Second, findings from this study can only be generalized to populations of students who are able to participate in a school-based health survey. This may exclude portions of the student population that experience differential risks for physical aggression, including adolescents who are truant, cognitively impaired, or incarcerated. Any selection bias due to the non-participation of these groups is likely to attenuate observed estimates of effect. Third, the etiology of physical aggression is complex, and these analyses consider four of many possible determinants. There is an ongoing need for more complex models and methods of study that tease out the relative impacts of a complex array of contributing factors. Our environmental risk score represents a possible over-simplification of this social model in that each environmental risk factor is weighted equally. Finally, surveys such as the HBSC have the inherent limitation of being reliant on self-reports from children who actually attend school. Such assessments may not provide an optimal basis for explaining similarities and differences in violence between the full population of adolescents from different countries and/or cultures.

Physical aggression among students is clearly a negative health behavior that should be discouraged. Yet, our results suggest that physical aggression is relatively normative in North American adolescent populations, with substantial minorities of boys and girls in both the United States and Canada reporting aggressive behavior. Speculatively, physical aggression during adolescence could influence future development. Occasional fighting may encourage most youth to figure out how to avoid violent behavior through the consequences of their actions,38 particularly for those with reasonably involved parents, teachers and other adult role models who discourage this behavior and help the adolescent to resolve the issue and avoid future confrontations.25 However, persistent aggressive behavior is associated with a trajectory of increased aggression and problem behavior, particularly in the absence of a positive and supportive social environment.25,38,41

The finding that social factors were consistently associated with fighting in multiple age groups and sexes and in both countries is consistent with Problem Behavior Theory which posits that antisocial and problem behavior is the product of multiple social influences in homes, schools and neighborhoods that may interact with personal characteristics to create negative health outcomes.33 From an adolescent development perspective, anti-social behaviors, including aggression, rapidly becomes normative during adolescence because of what Moffitt describes as the maturity gap, where adolescents are biologically mature, but socially immature, lacking meaningful adult social roles and responsibilities.38 For most, however, continued aggression would not be reinforcing, particularly among those with pro-social contacts and who gain access to meaningful, more adult-like status and responsibilities. The positive dimension of these social influences may serve as protective factors that discourage physical aggression or increase personal resilience against negative environmental influences.25 Consideration of the additive effects of different social factors on risks for physical aggression via the environmental risk score was also inspired by these theoretical perspectives and previous studies of adolescent behavior.26 In past analyses, we have applied a similar method to study the health effects of multiple risk behavior.26,42 This approach to exposure assessment is simple, has a solid theoretical basis, and appears to be a useful tool to foster understanding of clustered risk factors and their consequences. With refinement, the risk score may provide a screening mechanism for the early identification of youth at high risk for violence and other negative health outcomes.

One aim of the current analysis was to develop evidence to think about differences in serious violence (e.g. violent crime) observed in Canada and the United States.24 In this respect, our analysis is most unhelpful. It is curious that the prevalence of physical aggression and its association with socio-environmental factors are so similar among adolescents in the two countries, and yet violent crime and incarceration of juveniles and adults are so much greater in the United States.24 One explanation may be the possible insensitivity of our measures and study design. Physical aggression is one component of a syndrome of problem behaviors,26,33 and measures of aggression considered in isolation may not be the best indicator of a violent lifestyle. Second, we were unable to distinguish minor physical fights from more intense episodes of violence, and transient episodes from life-long patterns of anti-social behavior, which might partially mask national differences. Third, events involving physical aggression or the aftermath of these events may contribute to negative outcomes associated with social sanctions, especially in school environments where policies may be fairly standardized. This situation too may distort the findings of this cross-sectional analysis. Fourth, there are numerous contextual factors, e.g., characteristics of the youth’s neighborhood, school, and family, that are not considered in these analyses yet have been shown to have significant influence on violence and aggression.16 Fifth, these additional contextual factors and those influences included in the model may have a different mediating and moderating effects across different economic, racial/ethnic, and rural/urban strata. Finally, it may well be that in the United States compared to Canada approaches to the control of physical aggression are more lax, unconstructive, or punitive, social norms regarding violence vary, or the relatively greater access in the United States to weapons of violence account for the observed differences in criminal violence.43

This research provides additional evidence of the importance of the social environments as factors in adolescent aggressive behavior. The similarity of influences across countries suggest that these social environmental factors and the mechanisms by which they influence aggression are robust. It does not appear that children in the US have a propensity to engage in more physical aggression than their neighbors to the north. In both countries, family, peer, and school social environments clearly serve as risk or protective factors, with significant cumulative impact on physical aggression in school and other environments.

Finally, the study findings have implications for school personnel. First, these findings provide new data about the magnitude and distribution of the physical aggression problem in Canada and the United States. Violence in schools has become a universal concern for teachers and administrators, and this large cross-national profile can contribute to the formulation and justification of contemporary educational policy in both countries. Second, the environmental risk score developed for our analysis has potential as a screening tool and when combined with measures of prevalence could provide schools with information about risk at the school level. School violence prevention programs might benefit from the use of an assessment of environmental risk. Finally, given what is understood about the long-term effects of this behavior on rates of adult aggression, and the more proximal negative effects on relationships with peers and attitudes towards school, school personnel are advised to implement policies to reduce aggressive behavior at school as a priority.

Acknowledgments

This study was supported financially, in part, by the Canadian Institutes of Health Research (operating grant 2004MOP-CHI-128223-C), the Public Health Agency of Canada (Contract: HT089-05205/001/SS), the US Health Resources and Services Administration, and the US National Institute of Child Health and Human Development. Health Behaviour in School-aged Children is a World Health Organization/European Region collaborative study. The international coordinator of the 2001/2002 study is Candace Currie, University of Edinburgh, Scotland; and the data bank manager is Oddrun Samdal, University of Bergen, Norway. We thank Matthew King (national coordinator), Dr. Will Boyce (principal investigator), and Danielle Shelley from the Canadian HBSC team, and Dr. Mary Overpeck (principal investigator) and Kimberly Chambers from the United States HBSC team.

References

  • 1.U.S. DHHS. Youth Violence: A Report of the Surgeon General. Rockville, MD: US. Department of Health and Human Services; Washington DC: US GPO; 2001. [Google Scholar]
  • 2.Hamburg MA. Youth violence is a public health concern. In: Delbert E, Hamburg B, Kirk WR, editors. Violence in American Schools: A new perspective. New York, NY: Cambridge University Press; 1998. pp. 31–54. [Google Scholar]
  • 3.Kingery PM, Coggeshall MB, Alford AA. Violence at school: recent evidence from four national surveys. Psychology in the Schools. 1998;35:247–258. [Google Scholar]
  • 4.Loeber R, Farrington DP, Stouthamer-Loeber M, Van Kammen WB. Antisocial Behavior and Mental Health Problems: Explanatory Factors in Childhood and Adolescence. Mahwah, NJ: Lawrence Elbaum Associates; 1998. pp. 185–200. [Google Scholar]
  • 5.Malek MK, Chang B, Davis TC. Fighting and weapon carrying among seventh-grade students in Massachusetts and Louisiana. J Adolesc Health. 1998;2:92–102. doi: 10.1016/s1054-139x(98)00024-x. [DOI] [PubMed] [Google Scholar]
  • 6.St George DM, Thomas SB. Perceived risk of fighting and actual fighting behavior among middle school students. J Sch Health. 1997;67:178–181. doi: 10.1111/j.1746-1561.1997.tb07166.x. [DOI] [PubMed] [Google Scholar]
  • 7.Loeber R, Green S, Lahey BB, Kalb L. Physical fighting in childhood as a risk factor for later mental health problems. Journal of the American Association of Child and Adolescent Psychiatry. 2000;39:421–428. doi: 10.1097/00004583-200004000-00010. [DOI] [PubMed] [Google Scholar]
  • 8.Brener ND, Simon TR, Krug EG, Lowry R. Recent trends in violence-related behaviors among high school students in the United States. JAMA. 1999;282:440–446. doi: 10.1001/jama.282.5.440. [DOI] [PubMed] [Google Scholar]
  • 9.Pickett W, Craig W, Harel Y, et al. Cross-national study of fighting and weapon carrying as determinants of adolescent injury. Pediatrics. 2006;116:e855–863. doi: 10.1542/peds.2005-0607. [DOI] [PubMed] [Google Scholar]
  • 10.Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Web-based injury statistics query and reporting system (WISQARS) [online] 2006 [cited 2006 Feb 8]. Available from: URL: www.cdc.gov/ncipc/wisqars.
  • 11.Clubb PA, Browne DC, Humphrey MS, et al. Violent behaviors in early adolescent minority youth: results from a “Middle School Youth Risk Behavior Survey”. Matern Child Health J. 2001;5:225–235. doi: 10.1023/a:1013076721400. [DOI] [PubMed] [Google Scholar]
  • 12.Centers for Disease Control and Prevention. Youth Risk Behavior Surveillance-United States. MMWR 2006. 2005;55(SS5) [Google Scholar]
  • 13.Brener ND, Collins JL, Kann L, Warren C, Williams BI. Reliability of the youth risk behavior survey questionnaire. Am J Epidemiol. 1995;141:575–80. doi: 10.1093/oxfordjournals.aje.a117473. [DOI] [PubMed] [Google Scholar]
  • 14.Epstein JA, Botvin GJ, Diaz T, Williams C, Griffin K. Aggression, victimization, and problem behavior among inner-city minority adolescents. Journal of Child and Adolescent Substance Abuse. 2000;9:51–66. [Google Scholar]
  • 15.Graham S, Bellmore AD, Mize J. Peer Victimization, aggression, and their co-occurrence in middle school: pathways to adjustment problems. J Abnorm Child Psychol. 2006;34:363–378. doi: 10.1007/s10802-006-9030-2. [DOI] [PubMed] [Google Scholar]
  • 16.Loeber R, Pardini D, Homish L, et al. The prediction of violence and homicide in young men. J Consult Clin Psychol. 2005;73:1074–1088. doi: 10.1037/0022-006X.73.6.1074. [DOI] [PubMed] [Google Scholar]
  • 17.Fraser MW. Aggressive behavior in childhood and early adolescents: an ecological-developmental perspective on youth violence. Soc Work. 1996;41:347–361. [PubMed] [Google Scholar]
  • 18.Gottfredson MR, Hirschi T. A general theory of adolescent problem behavior: problems and prospects. In: Ketterlinus RD, Lamb ME, editors. Adolescent problem behaviors: issues and research. Hillsdale NJ: Lawrence Erlbaum Associates, Inc; 1994. pp. 41–56. [Google Scholar]
  • 19.Barber BK. Adolescent socialization in context: the role of connection, regulation, and autonomy. Journal of Adolescent Research. 1997;12:5–11. [Google Scholar]
  • 20.Williams JM, Dunlop LC. Pubertal timing and self-reported delinquency among male adolescents. J Adolesc. 1999;22:157–171. doi: 10.1006/jado.1998.0208. [DOI] [PubMed] [Google Scholar]
  • 21.Simons-Morton BG, Hartos J, Haynie DL. Prospective analyses of peer and parent influences on minor aggression among early adolescents. Health Edu Behav. 2004;31:22–33. doi: 10.1177/1090198103258850. [DOI] [PubMed] [Google Scholar]
  • 22.Brookmeyer KA, Fanti KA, Henrich CC. Schools, parents, and youth: a multilevel, ecological analysis. J Clin Child Adolesc Psychol. 2006;35:504–514. doi: 10.1207/s15374424jccp3504_2. [DOI] [PubMed] [Google Scholar]
  • 23.Resnick MD, Bearman PS, Blum RW, et al. Protecting adolescents from harm. Findings from the National Longitudinal Study on Adolescent Health. JAMA. 1997;278:823–832. doi: 10.1001/jama.278.10.823. [DOI] [PubMed] [Google Scholar]
  • 24.Pritchard C, Butler A. A comparative study of children and adult homicide rates in the USA and the major western countries 1974–1999: Grounds for concern? Journal of Family Violence. 2003;18:341–350. [Google Scholar]
  • 25.Rutter M. Resilience: Some conceptual considerations. J Adolesc Health. 1993;14:626–631. doi: 10.1016/1054-139x(93)90196-v. [DOI] [PubMed] [Google Scholar]
  • 26.Pickett W, Schmid H, Boyce WF, et al. Multiple risk behaviour and injury: an international analysis of young people. Arch Pediatr Adolesc Med. 2002;156:786–793. doi: 10.1001/archpedi.156.8.786. [DOI] [PubMed] [Google Scholar]
  • 27.Stronski SM, Ireland M, Michaud P, Narring F, Resnick MD. Protective correlates of stages in adolescent substance use: A Swiss National Study. J Adolesc Health. 2000;26:420–427. doi: 10.1016/s1054-139x(99)00057-9. [DOI] [PubMed] [Google Scholar]
  • 28.Currie C, Samdal O, Boyce W, Smith B, editors. Health Behavior in School-Aged Children: a World Health Organization Cross-National Study (HBSC), Research Protocol for the 2001/2002 Survey. University of Edinburgh; Scotland: 2001. [Google Scholar]
  • 29.Olweus D. Bullying among schoolchildren: intervention and prevention. In: Peters RD, McMahon RJ, Quinsey VL, editors. Aggression and violence throughout the lifespan. Newbury Park: Sage Publications; 1992. pp. 127–139. [Google Scholar]
  • 30.Nutbeam D, Smith C, Moore L, Bauman A. Warning! Schools can damage your health: alienation from school and its impact on health behavior. J Paediatr Child Health. 1992;29(Suppl 1):S25–30. doi: 10.1111/j.1440-1754.1993.tb02256.x. [DOI] [PubMed] [Google Scholar]
  • 31.Torsheim T, Wold B, Samdal O. The teacher and classmate support scale - Factor structure, test-retest reliability and validity in samples of 13-and 15-year-old adolescents. School Psychology International. 2000;21:195–212. [Google Scholar]
  • 32.Mullan E, Currie C. Social inequalities and adolescent health: Cross-national patterns from the 1998 HBSC survey. Policy Series: Health Policy for Children and Adolescents. Copenhagen: World Health Organization; 2000. [Google Scholar]
  • 33.Jessor R. Risk behavior in adolescence: a psychosocial framework for understanding and action. J Adolesc Health. 1991;12:597–605. doi: 10.1016/1054-139x(91)90007-k. [DOI] [PubMed] [Google Scholar]
  • 34.Boyce W, King M. Introduction. In: Boyce W, editor. Young people in Canada: their health and well-being. Ottawa: Health Canada; 2004. pp. 1–7. [Google Scholar]
  • 35.Dahl RE. Adolescent brain development: a period of vulnerabilities and opportunities. Keynote address. Ann N Y Acad Sci. 2004 Jun;1021:1–22. doi: 10.1196/annals.1308.001. [DOI] [PubMed] [Google Scholar]
  • 36.Carpendale JI, Lewis C. Constructing an understanding of mind: the development of children’s social understanding within social interaction. Behav Brain Sci. 2004;27:79–96. doi: 10.1017/s0140525x04000032. [DOI] [PubMed] [Google Scholar]
  • 37.Dodge K, Pettit GS, McClaskey CL, Brown MM. Social competence in children. Monogr Soc Res Child Dev. 1986;51:1–85. [Google Scholar]
  • 38.Moffitt TF. Adolescence-limited and life-course-persistent antisocial behaviour: a developmental taxonomy. Psychol Rev. 1993;100:674–701. [PubMed] [Google Scholar]
  • 39.Currie C, Roberts C, Morgan A, et al., editors. WHO Policy Series: Health Policy for Children and Adolescents Issue 4. Copenhagen, Denmark: WHO Regional Office for Europe; 2004. Young people’s health in context: international report from the HBSC 2001/02 survey. [Google Scholar]
  • 40.Hill AB. The environment and disease: association or causation? Proc R Soc Med. 1965;58:296–300. doi: 10.1177/003591576505800503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Woodward LJ, Fergeson DM. Childhood and adolescent predictors of physical assault: a prospective longitudinal study. Criminology. 2000;38:233–262. [Google Scholar]
  • 42.Simpson K, Janssen I, Boyce WF, Pickett W. Risk taking and recurrent health symptoms in Canadian adolescents. Prev Med. 2006;43:46–51. doi: 10.1016/j.ypmed.2006.03.005. [DOI] [PubMed] [Google Scholar]
  • 43.Cukier W. Firearms regulations: Canada in the international context. Chronic Dis Can. 1998;19:25–34. [PubMed] [Google Scholar]

RESOURCES