RELATIONSHIPS BETWEEN DEGREE OF HEALTH RISK BEHAVIOR INVOLVEMENT AND VIOLENT ADOLESCENT BEHAVIOR

1
RELATIONSHIPS BETWEEN DEGREE OF HEALTH RISK
BEHAVIOR INVOLVEMENT AND VIOLENT ADOLESCENT
BEHAVIOR
Elizabeth M.G. Larkin, MSCenter for Adolescent
Health, Case Western Reserve University
emg9_at_cwru.edu
xxx_at_cwru.edu
ABSTRACT Purpose To examine the relationship
between the extent of adolescent involvement in
health risk behaviors and the frequency of
violent behavior. Methods 2,244 Midwestern
suburban high school students completed a
questionnaire assessing both health risk and
violent behavior involvement. Twenty-nine
questions from the Youth Risk Behavior Survey, an
87-item national survey, were used to assess
extent of health risk behavior involvement
including the use of alcohol, tobacco, and
illegal drugs sexual activity and
injury-causing behaviors. Additionally, students
were asked how often they had participated in
various violent behaviors, resulting in a
20-point scale demonstrating good reliability
(alpha .80). For each health risk behavior,
represented in a 5-point Likert scale,
preliminary Pearson correlations and one-way
ANOVAs were used to determine the association
between frequency of violent behavior and level
of health risk involvement. Dunnetts C post-hoc
tests were also performed. Results All health
risk behaviors were positively related to
frequency of violent behavior with Pearsons
correlations ranged from .111 to .370 and ANOVA F
statistics ranged from 19.2 to 102.5. Post-hoc
tests confirmed the positive relationship.
Conclusions These findings suggest that
increased engagement in risky behaviors, and/or
continual disregard for personal health and
safety are related to increases in adolescent
violent behavior. Therefore, targeting
adolescents with risk-taking proclivities for
entry into violence prevention programs may be an
effective method of maximizing the effectiveness
of interventions. These findings should be
incorporated into programs aimed at reducing both
health risk behaviors and violent behaviors in
adolescents.

• BIVARIATE ANALYSIS
• Males were found to score significantly higher on
  the violence scale than females (plt.001).
• Younger students in 9th and 10th grades were
  significantly more violent than those in the
  higher grade levels (plt.001).
• Students identifying themselves as American
  Indian/Alaskan Native or African American/Black
  had a significantly higher average violence score
  than those who did not identify with these groups
  (p.016 and plt.001). Those students identifying
  themselves as white demonstrated significantly
  lower violence scores than those students who do
  not describe themselves as white (plt.001).
• All 29 health risk behaviors were significantly
  associated with violence and demonstrated
  significant correlative relationships with
  frequency of violent behavior (ANOVA F plt .001,
  Pearsons R plt .003,) (See Figure 3).
• CONCLUSIONS
• Negative health behaviors of all types are
  correlated with adolescent violent behavior.
• Although no causal relationship can be deduced
  from this cross-sectional data, the consistency
  of these patterns suggest that an underlying
  willingness to engage in risky behaviors, and/or
  an underlying disregard for personal health and
  safety may be found to be a significant
  contributor to adolescent violence.
• Targeting adolescents with risk-taking
  proclivities for entry into violence prevention
  programs may be an effective method of maximizing
  the effectiveness of interventions.
• REFERENCES
• Bailey, S.L. Flewelling, R.L. Rosenbaum, D.P.
  (1997). Characteristics of students who bring
  weapons to school. J. Adolesc. Health, 20 (4),
  261-70.
• Clubb, P.A. Browne, D.C. Humphrey et al.
  (2001). Violent Behaviors in early adolescent
  minority youth Results from a Middle School
  Youth Risk Behavior Survey. Mat. Child Health
  J., 5(4), 225-235.
• Dukarm, C.P. Byrd, R.S. et al. (1996). Illicit
  substance use, gender, and the risk of violent
  behavior among adolescents. Arch. Pediatr.
  Adolesc. Med., 150(8), 797-801.
• DuRant, R.H. Krowchuk, D.P. et al. (1999).
  Weapon carrying on school property among middle
  school students. Arch. Pediatr. Adolesc. Med.,
  153 (1), 21-26.
• Grumbaum, J.A. Basen-Engquist, K. Pandey, D.
  (1998). Association between violent behaviors and
  substance use among Mexican-American and
  non-Hispanic white high school students. J.
  Adolesc. Health, 23 (3), 153-159.

• RESULTS
• DESCRIPTIVE STATISTICS
• The sample taken from the seven inner-ring high
  schools yielded a fairly diverse group. The
  majority of students ranged from 14 to 18 years
  old with over 45 of students identifying with a
  minority ethnicity (see Figure 1).
• UNIVARIATE ANALYSIS
• In the past year, 58 had told others they would
  hurt them, 51 had hit someone before they were
  hit, 68 had hit someone after they were hit, 26
  had beaten someone up, and 5 had attacked
  someone with a knife (See Figure 2).
• The scale appears reliable with a Cronbachs
  alpha of 0.80.
• The violence scale, ranging from 5 to 20,
  resulted in an average score of 7.8.

Figure 3 Self-reported Demographics

[phone]
[phone]
Alcohol Use
Injuries
Violence Scale
Violence Scale
Adolescent Health
Figure 1 Self-reported Demographics

• INTRODUCTION
• In the last ten years, much research has focused
  on the prevalence and possible risk factors of
  adolescent violence. 1-8
• A great deal of this work has centered around the
  Youth Risk Behavior Survey (Y.R.B.S.), a school
  based self-report survey developed by the Centers
  for Disease Control and Prevention (CDC). 9
• While the Y.R.B.S. has proven an effective
  instrument in monitoring most health behaviors,
  it may not always the best tool for conducting
  analysis of violent behaviors. Because of the
  more extreme violent behavior questions asked,
  previous studies have only been able to look at
  the data dichotomously, reporting only odds
  ratios for extreme behaviors1-8.
• A finer instrument is needed to examine violence
  in its earlier stages and less extreme forms,
  allowing a wider distribution of responses.
• This study proposes that a more sensitive scale
  may provide insight into the relationship between
  the degree of adolescent violent behavior and
  extent of risk behavior involvement.
• METHODOLOGY
• DESIGN - In 2002, a self-report questionnaire was
  administered to a random sample of 2,244 suburban
  high school students during regular school hours,
  in order to determine the relationship between
  health risk behavior involvement and adolescent
  violent behavior.
• SAMPLE/DATA COLLECTION - A clustered sampling
  method was used, with students randomly selected
  by classroom from each school. Student
  participation was both anonymous and voluntary
  and passive permission slips were used. A total
  of 2,244 complete surveys were received.
• INSTRUMENTATION -The basic 87-item CDC-developed
  questionnaire was administered with an additional
  5-question violence scale10. The questions
  focused on aggressive behavior, as opposed to the
  more extreme fighting and weapons questions
  already included in the national survey. A
  higher violence score indicated more violence
  involvement.
• ANALYSIS - ANOVA statistics, Dunnets C post-hoc
  tests, and Pearsons correlations were
  calculated, and significant results were graphed
  to illustrate the dose-response pattern of the
  relationships.

Cigarette Use
Smokeless Tobacco Use
Violence Scale
Violence Scale
Gender
Last Name, First Name
Larkin, Elizabeth M. G.
Marijuana Use
Hard Drug Use
Grade Level
Race
Figure 2 Distribution of Violence Question
Responses
Violence Scale
Violence Scale
Sexual Activity
Violence Scale