Citation
Risk and protective factors of suicidal adolescents

Material Information

Title:
Risk and protective factors of suicidal adolescents
Creator:
Stappert, Kelsey ( author )
Language:
English
Physical Description:
1 electronic file (49 pages). : ;

Subjects

Subjects / Keywords:
Suicide -- Prevention ( lcsh )
Youth -- Suicidal behavior ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
Recognizing risk and protective factors for adolescent suicide attempts is an important factor for identification and suicide prevention strategies. The purpose of the study was to assess the relationship between adolescent risk factors and suicide. Nationally representative data from the 2013 Youth Risk Behavior Survey (YRBS) created by the Centers for Disease Control and Prevention (CDC) were used. The relationship of demographic characteristics, bullying, substance abuse, media exposure, and physical activity to suicide attempts was assessed. Results concluded distinguishable characteristic among demographics and suicide risk factors: gender, race/ethnicity, grade level, experience of bullying-school property and electronically, exercise-physical activity and sports participation, and media exposure-video game/internet use. Implications for practice and future research are discussed.
Thesis:
Thesis (Psy.D.)--University of Colorado Denver.
Bibliography:
Includes bibliographic references
System Details:
System requirements: Adobe Reader.
General Note:
Department of Psychology
Statement of Responsibility:
by Kelsey Stappert..

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
952104854 ( OCLC )
ocn952104854

Downloads

This item has the following downloads:


Full Text
RISK AND PROTECTIVE FACTORS OF SUICIDAL ADOLESCENTS
by
KELSEY STAPPERT
B.G.S., University of Kansas, 2010
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Psychology
School Psychology Program
2016


2016
KELSEY STAPPERT
ALL RIGHTS RESERVED


This thesis for the Doctor of Psychology degree by
Kelsey Stappert
has been approved for the
School Psychology Program
by
Franci Crepeau-Hobson, Chair
Bryn Harris
Colette Hohnbaum


Stappert, Kelsey (Psy.D., School Psychology)
Risk and Protective Factors of Suicidal Adolescents
Thesis directed by Associate Professor Franci Crepeau-Hobson
ABSTRACT
Recognizing risk and protective factors for adolescent suicide attempts is an important factor
for identification and suicide prevention strategies. The purpose of the study was to assess the
relationship between adolescent risk factors and suicide. Nationally representative data from the 2013
Youth Risk Behavior Survey (YRBS) created by the Centers for Disease Control and Prevention
(CDC) were used. The relationship of demographic characteristics, bullying, substance abuse, media
exposure, and physical activity to suicide attempts was assessed. Results concluded distinguishable
characteristic among demographics and suicide risk factors: gender, race/ethnicity, grade level,
experience of bullying-school property and electronically, exercise-physical activity and sports
participation, and media exposure-video game/internet use. Implications for practice and future
research are discussed.
The form and content of this abstract are approved. I recommend its publication.
Approved: Franci Crepeau-Hobson
IV


ACKNOWLEDGEMENTS
Life has been a spectacular journey thus far. I have been fortunate to have many supportive
mentors who have provided me with the expertise and experience that lead me to constantly challenge
my knowledge of the world. Without each of you, this wonderful journey may not have happened.
Thank you for your patience and guidance along the way.
v


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION...................................................................1
Adolescent Suicide Prevalence and Risk.....................................1
Suicide Risk and Protective Factors........................................1
Demographics....................................................... 1
Bullying........................................................2
Substance Abuse.....................................................2
Media Exposure......................................................3
Exercise and Sports Participation...................................4
Specific Aims..............................................................4
II. METHODS.......................................................................7
Sample.....................................................................7
Measures...................................................................8
Demographics........................................................8
Bullying............................................................8
Substance Abuse.....................................................9
Media Exposure......................................................9
Exercise and Sports Participation..................................10
Suicide........................................................... 11
Data Analysis............................................................ 12
Assumptions....................................................... 13
III. RESULTS.................................................................... 14
Demographics as Predictors of Suicide Attempt............................ 14
Risk and Protective Factors as Predictors of Suicide Attempt............. 15
Demographics with Risk and Protective Factors Predicting Suicide Attempt. 17
vi


Bullying......................................................... 17
Substance Abuse..................................................22
Media Exposure...................................................24
Exercise and Sports Participation................................27
IV. DISCUSSION.................................................................30
Interpretation..........................................................30
Limitations and Future Research Directions..............................33
Implications for Practice...............................................34
Competing Interests.....................................................35
REFERENCES.............................................................................36
vii


LIST OF TABLES
TABLE
la. Sample Characteristics of the Population.............................................. 11
lb. Sample Characteristics of Risk and Protective Factors................................. 12
2. Logistic Regression Results for Demographic Predictors of Gender Race and Grade Level: Suicide
Attempt................................................................................ 15
3. Logistic Regression Results for Predictors of Risk and Protective Factors: Suicide Attempt.17
4. Chi-Square Results for Predictors of Gender, Race, and Grade Level by Bullying Factors: Suicide
Attempt................................................................................ 19
5. Chi-Square Results for Predictors of Gender, Race, and Grade Level by Substance Use Factors:
Suicide Attempt.........................................................................22
6. Chi-Square Results for Predictors of Gender, Race, and Grade Level by Media Exposure Factors:
Suicide Attempt.........................................................................25
7. Chi-Square Results for Predictors of Gender, Race, and Grade Level by Exercise and Sports
Participation Factors: Suicide Attempt..................................................28
vm


CHAPTER I
INTRODUCTION
Adolescent Suicide Prevalence and Risk
According to the Centers for Disease Control (CDC), suicide is the third leading cause of
death for people aged 15 to 24 years (CDC, 2010a; Buda, Berman, Jobes & Silverman, 2007;
Murphy, Xu, & Kochanek, 2012; Heron, 2013). The American Association of Suicidology
indicates increasing rates of depression and death by suicide for across adolescent populations
(McIntosh, 2012). With adolescent suicide rates on the rise, there has also been subsequent
advancement in identifying and understanding factors associated with adolescent suicide (Cash &
Bridge, 2009).
The developmental stage of adolescence, a phase of life when children transition to
adulthood, is a crucial period of physical maturation and brain development (Crone & Dahl,
2012; Konrad, Firk, & Uhlhaas, 2013). With changes occurring in brain development during
adolescence, there is often an increase in risk-taking behaviors and sensation seeking (Dahl,
2004; Fine, & Sung, 2014). Some of these risk-taking behaviors can lead to exposure to potential
adverse life events and an increased vulnerability to mental illness (Dahl, 2004). By further
examining the risk-taking behaviors of suicidal adolescents, it may point to potential prevention
initiatives and strategies to combat further adolescent death.
Suicide Risk and Protective Factors
Since suicide is one of the leading causes of adolescent death, identifying suicide-related
behaviors and both risk and protective factors is of critical importance. Risk and protective
factors can affect the likelihood that a suicide attempt will occur. There are several suicide-related
factors that are associated with increased and decreased suicide attempts.
Demographics. In a study using a national adolescent sample in the United States, gender
was identified as a major demographic risk factor for suicide (Epstein & Spirito, 2010;Wolitzky-
Taylor, Ruggiero, McCart, Smith, Hansen, Resnick, & Kilpatrick, 2010; Chung & Joung, 2012).
1


In the United States, the rate of suicidal ideation and attempts is higher among females than for
their male peers. Additionally, according to research conducted by Pena and colleagues, suicide
attempts are disproportionately female, Black, and Hispanic (Pena, Matthieu, Zayas, Masyn, &
Caine, 2012). They also found that of subpopulations, Hispanic females had higher rates of
suicide attempts than other minority groups; whereas, White males had lower rates of suicide
attempts than other subpopulations. Additionally, research indicates that men die by suicide three
and a half time more often than women (CDC, 2014). Since demographic factors such as gender
and race/ethnicity are associated with increased suicide rates among adolescent youth, further
analysis may indicate additional risk factors.
Bullying. Data from the 2011 National Youth Risk Behavior Survey (YRBS) indicates
over 20% of high school youth are bullied on school property and 16% are victims of electronic
bullying or cyberbullying (CDC, 2012/ Victims of cyber bullying are more likely to report
depressive symptoms than victims of other types of bullying (Wang, Nansel, & Iannotti, 2011).
There is a large body of research that supports the relationship between bullying and adolescent
suicide (Klomek, Sourander, & Gould, 2010). This includes a wide range of bullying behaviors:
being belittled about religion, race, looks, or speech; being the subject of lies, rumors, or sexual
jokes or comments, and cyberbullying (Bonanno & Hymel, 2013; Hinduja & Patchin, 2010;
Klomek et al., 2010; Undheim, 2013). Given the high prevalence of bullying and its strong
association with adolescent suicide risk, it is necessary to further study this relationship.
Substance Abuse. Another risk factor linked with suicidal ideation and attempts among
adolescents include substance use in the United States (Dunn, Goodrow, Givens, & Austin,
2008). Adolescent substance use is highly prevalent, particularly among high school students
(Johnston, OMalley, Bachman, & Schulenberg, 2008). Research has identified that adolescents
experiencing suicidal ideation are at a greater risk for substance abuse (Department, 2001;
American Academy of Child and Adolescent Psychiatry, 2004).
2


Studies have found associations between adolescent illicit drug use and suicide, although
types of drugs are frequently not identified or compared (Miller, 2011). Alcohol use has been
associated with an increased risk for suicidal behaviors among adolescents (Roy, Lamparski,
DeJong, Moore, & Linnoila, 1990; Borowsky, Ireland, & Resnick, 2001; Bossarte, & Swahn,
2011). However, the association o marijuana use and suicide has been inconsistent (Degenhardt,
Hall, & Lynskey, 2003; Pedersen, 2008; Price, Hemmingsson, Lewis, Zammit, & Allebeck, 2009;
Rasic, Weerasinghe, Asbridge, & Langille, 2013; Chabrol, Melioli, & Goutaudier, 2014;
Rylander, Valdez, & Nussbaum, 2014). A study by Rasic et al. (2013) found and association
between heavy marijuana use and depression but identified no association with suicidal ideation
or attempts. Overall, the use of alcohol and marijuana use have all been associated with
suicidality (Kim, 2010; Roberts 2010); however, extensive research into specific use and suicide
attempts have not been consistent.
Media Exposure. Currently, video games and internet use are common activities among
children and adolescents (Lenhart & Madden, 2007; Marshall, Gorely, & Biddle, 2006).
According to a study from Rideout and colleagues, approximately 83% of United States youth
between 8 and 18 years have video game consoles at home (Rideout, Roberts, & Foehr, 2005).
According to this study, 86% have a home computer, 35% have a computer in their bedroom, and
the average time of Internet use for recreation was estimated for over an hour. Additionally,
compared to adolescents who report no video game or Internet use, adolescents reporting 5 or
more hours of daily video game and Internet use were more likely to have experienced suicidal
ideation or have made a suicidal plan. (Messias, Castro, Saini, Usman, & Peeples, 2011;
Gauthier, Zuromski, Gitter, Witte, Cero, Gordon, & Joiner, 2014). Such results are consistent
with other studies that found a relationship between depression, video gaming, and suicide (Kim,
Namkoong, Ku, & Kim, 2008; Mathers, Canterford, Olds, Hesketh, Ridley, & Wake, 2009;
Rehbein, Kleimann, & Mossle, 2010; Weaver, Mays, Sargent Weaver, Kannenberg, Hopkins,
Eroglu, et al., 2009) including Internet overuse (Lam & Peng, 2010). Although research has
3


found associations between video game and internet use with suicide, there has been limited
research that examines the relationship between number of hours spent watching television and
suicide.
Exercise and Sports Participation. Research suggests that exercise and sports
participation might be protective factors in suicide attempts and suicide. Youth physical activity
guidelines, according to the U.S. Department of Health and Human Services, recommend that
children and adolescents aged 6-17 years should have 60 minutes or more of physical activity
each day (2008). A number of studies with adolescents who had regular participation in athletic
activity displayed lower reported instances of hopelessness, depression, suicidal ideation, and
suicide (Babiss, & Gangwisch, 2009; Brown, Galuska, Zhang, Eaton, Fulton, Lowry, & Maynard,
2007; Miller, & Hoffman, 2009; Taliaferro, Rienzo, Miller, Pigg, & Dodd, 2009). Other studies
also provided evidence of physical inactivity was a risk factor for suicide. For example, research
indicates that individuals who attempted suicide were less physically active than those who did
not attempt (Simon, Powell, & Swann, 2004; Chioqueta, & Stiles, 2007). Although many case
studies have examined physical activity as a protective factor of suicide in adolescents, further
examination may clarify types of physical activity and its contribution to protective.
By contributing to research that further examines potential risk and protective factors in
adolescent populations may assist in understanding and prevention efforts to combat adolescent
suicide.
Specific Aims
Of adolescents who experience the distress of suicidal ideation, only a small proportion
of suicidal adolescents is recognized by adults, including parents and school staff who may be the
best prospects to help (Brown, Wyman, Brinales, & Gibbons, 2007). Additionally, the majority of
suicidal adolescents do not seek adult support (Kerr, Owen, Pears, & Capaldi, 2008; Wyman,
Brown, Inman, Cross, Schmeelk-Cone, Guo, & Pena, 2008; Klaus, Mobilio, & King, 2009).
Parent and school staff s inability to identify and detect adolescents at risk of suicide is a major
4


barrier to reducing suicide deaths. Although an increasing number of school-based suicide
prevention efforts are aimed at improving help-seeking behavior from adolescents (Aseltine,
James, Schilling, & Glanovsky, 2007; Freedenthal, 2010), there remains a gap in clear
identification of adolescent suicide risk and protective factors. The present study seeks to address
this gap by investigating the relationship between adolescent suicidal ideation and a variety of
risk and protective factors for on a nationally representative sample of high school students in the
United States. Specifically, this study examined demographics (i.e., gender, race/ethnicity, and
grade), bullying (i.e., on school property and electronically), substance abuse (i.e., alcohol and
marijuana), media exposure (i.e., hours per school day spent watching television and video
game/Internet usage), as well as exercise and physical activity (i.e., days per week of physical
activity and sports participation).
This study utilized data from Youth Risk Behavior Surveillance System (YRBS; Brener
et al., 2013). Data from a nationally representative sample such as the YRBS can be used to
determine the degree of the relation between demographics of risk and protective factors and
suicide attempts to inform the efforts of early identification and intervention of youth at-risk for
suicide. The purpose of the current study was to assess how well the following variables predicted
suicide attempts: gender, grade, race/ethnicity, bullied- on school property and electronically,
substance use- alcohol and marijuana, media exposure- television and video games/internet, and
exercise and sports participation. The following research questions were investigated:
RQ1: Do the demographic characteristics of gender, grade, or race/ethnicity predict
whether a student attempted suicide?
RQ2: Does the experience of being bullied- on school property and electronically,
substance use- alcohol and marijuana, media exposure- television and video
games/internet, and exercise and sports participation that best predicts whether a student
attempted suicide?
5


RQ3: Is there a combination gender, grade, race/ethnicity, gender, experience of being
bullied- on school property and electronically, substance use- alcohol and marijuana,
media exposure- television and video games/intemet, and exercise and sports
participation that best predicts whether a student attempted suicide?
Knowing the answers to these questions is important for identifying students at-risk for
suicide attempts in order to intervene and save adolescent lives.
6


CHAPTER II
METHODS
Sample
The Youth Risk Behavior Survey (YRBS) is a scale created by the Centers for Disease
Control and Prevention (CDC) to assess the prevalence of adolescent behaviors that contribute to
the leading causes of death, disability, and social problems in the United States. The survey
contains about 98 self-report items designed to measure the frequency and the severity of
behaviors within six categories: (1) violent and self-injurious behavior, (2) tobacco use, (3)
alcohol and other drug use, (4) sexual behavior, (5) dietary behaviors, and (6) physical inactivity
(CDC, 2013).
The YRBS utilizes a three-stage cluster sample design and produces a representative
sample of 9th through 12th grade United States high school students aged 14-18 years. YRBS
surveys were distributed to all regular members of public, Catholic, and private schools between
grades 9 and 12 across 50 states. A weighting factor was applied to each student record to adjust
for a nonresponse and the oversampling of students in the sample. According to the YRBS
Combined Sets Users Guide, overall weights were scaled so the weighted count of students was
equal to the total sample size, and the weighted proportions of students in each grade matched
population projections for each survey year.
Data from the most recent (2013) Youth Risk Behavior Surveillance System was utilized
for this study. The final sample for the 2013 YRBSS was collected from 148 of the 193 high
schools sampled across the country (77% school response rate). Of the 15,480 students sampled,
13,633 submitted questionnaires. Surveys that contained incomplete and missing data were
omitted during the YRBS data scrub, which left 13,583 usable questionnaires. Several studies
have examined and determined the psychometric properties of the YRBS, indicating that the
YRBS has sufficient levels of test-retest reliability (e.g., Brener et al. 1995, 2002, 2003).
7


Measures
Due to the infrequency of multiple occurrences of the independent variables under
consideration and the strong positively skewed distributions (i.e., majority of the scores in the
lower range of the scales), predictor variables that were ordinal in nature (i.e., 0 times/days, 1
time/day, 2 or 3 times/days, 4 or 5 times/days, 6 or more times/days) were dichotomized using
dummy variables. This was also done to meet the condition of logistic regression that all
predictors be either scale or dichotomous in nature. Other researchers have used similar
procedures (e.g., Nickerson & Slater, 2009; Shilubane et al., 2012).
Demographics. Gender, grade level, year and race/ethnicity were provided by the data. In
the 2013 survey, gender included the categories of Female or Male. Grade subscale was
measured by providing response answers of 9th, 10th, 11th, 12th and ungraded or other.
For the purpose of this study the ungraded or other grade level category contained only 23
responses, and was omitted from the analysis. The race/ethnicity was broken down into two
questions: Are you Hispanic or Latino? and What is your race? Although there was an
extensive race/ethnicity category of eight levels, the race/ethnicity category was modified for
utilization for the present study. The race/ethnicity category was broken down from the previous
eight categories and abbreviated into four: White, Black or African American,
Hispanic/Latino, and All other race/ethnicities. Of the previous eight race/ethnicity
categories, White and Black or African American remained the same, whereas
Hispanic/Latino was combined with Multiple Races-Hispanic and All other race/ethnicities
included American Indian/Alaska Native, Asian, Native Hawaiian/Other Pacific Islander,
and Multiple Race- Non-Hispanic.
Bullying. Bullying was measured based on 2 items from the Youth Risk Behavior Survey
(CDC, 2013) in order to measure how frequently adolescents were victims of bullying at school
and electronically. Bullying on school property was examined by the question, During the past
12 months, have you ever been bullied on school property? This question provided responses
8


from Yes or No. The electronic bullying measurement consisted of 1 item intended to
measure how frequently adolescents were bullied, by peers, through electronic communication.
This item was also binary to responses of Yes or No to the question, During the past 12
months, have you ever been electronically bullied? (Include being bullied through e-mail, chat
rooms, instant messaging, Websites, or texting).
Substance Abuse. Alcohol and Marijuana abuse were measured based on 2 items from the
Youth Risk Behavior Survey (CDC, 2013). Alcohol abuse was examined by the question,
During the past 30 days, on how many days did you have 5 or more drinks of alcohol in a row,
that is, within a couple of hours? This question provided responses from 0 days, 1 day, 2
days, 3 to 5 days, 6 to 9 days, 10 to 19 days, and 20 or more days. For this study, the
alcohol abuse variable examined all responses of consuming 5 or more drinks of alcohol in the
last 30 days, for 1 or more days as compared to 0 days. Marijuana use was analyzed similarly
to alcohol abuse. Marijuana abuse was examined by the question, During the past 30 days, how
many times did you use marijuana? Question responses included, 0 times, 1 or 2 times, 3 to
9 times, 10 to 19 times, 20 to 39 times, and 40 or more times. Additionally, for this study,
marijuana use was analyzed by separating responses into 0 times and 1 or more times of
marijuana use in the last 30 days. This variable was analyzed to compare those who are using
marijuana 1 or more time to the entire sample of questionnaires. This variable was modified to
display marijuana use, the dummy variable of 1 was used for 0 days and 2 was used for 1
or more days.
Media Exposure. In order to assess media exposure, variables that measured hours of
watching television and video game/intemet use were examined. This YRBS item asked, On an
average school day, how many hours do you watch TV?. Item responses included, I do not
watch TV on an average school day, Less than 1 hour per day, 1 hour per day, 2 hours per
day, 3 hours per day, 4 hours per day, and 5 or more hours per day. For the purpose of the
study, only responses of 3 or more hours of television on an average school day were examined.
9


Additionally, average hours per school day and video game/internet use were examined. The
YRBS item question inquired, On an average school day, how many hours do you play video or
computer games or use a computer for something that is not schoolwork? (Count time spent on
things such as Xbox, PlayStation, an iPod, an iPad or other tablet, a smartphone, YouTube,
Facebook or other social networking tools, and the Internet.) Respondents were given the
following responses to choose from, I do not play video or computer games or use a computer
for something that is not school work, Less than 1 hour per day, 1 hour per day, 2 hours per
day, 3 hours per day, 4 hours per day, and 5 or more hours per day. In order to examine
excess use of video game/internet, only responses of 3 or more hours on an average school day
were examined, specifically the number of hours per average school day was coded as 1 for
less than 3 hours and more than 3 hours was coded as 2.
Exercise and Sports Participation. Examining adolescent inactivity or frequency in
exercise and sports participation, two survey items were analyzed. Regular physical activity was
measured by responses of 0 days, 1 day, 2 days, 3 days, 4 days, 5 days, 6 days,
and 7 days to the survey item, During the past 7 days, on how many days were you physically
active for a total of at least 60 minutes per day? (Add up all the time you spent in any kind of
physical activity that increased your heart rate and made you breathe hard some of the time.).
The physical activity item was analyzed based on self-response to questions of 5 days or more.
This variable was modified with the dummy code 1 for less than 5 days and 2 was used for
more than 5 days of physical activity.
Sports participation was measured by self-report to the YRBS item, During the past 12
months, on how many sports teams did you play? (Count any teams run by your school or
community groups.) Responders were provided responses of 0 teams, 1 team, 2 teams,
and 3 or more teams. Sports participation is examined by responses of one or more sports teams
in the past 12 months. Similarly, for number of sports teams, 1 was used for being on 0 teams
and 2 was used for 1 or more teams.
10


Suicide. The suicidal behavior subscale contained four items from the YRBS, although
items were only measured based on self-reported suicide attempts. This item asked adolescents to
respond to items of 0 times, 1 time, 2 or 3 times, 4 or 5 times, 6 or more times
measuring suicide attempts (e.g., During the past 12 months, how many times did you actually
attempt suicide?). From this question, a binary categorical variable was created: attempted
suicide as yes which indicated a response of one or more times and no which indicated a
response of 0 times. Table la and Table lb present the n for all variables included in the
analysis.
Table la Sample Characteristics of the Population (N =13,583)
Characteristics n (%)
Gender
Female 6,621 48.8
Male 6,950 51.2
Grade Level
9th 3,588 26.6
10th 3,152 23.4
11th 3,184 23.6
12th 3,557 26.4
Ethnicity
White 5,449 41.1
Black or African-American 2,993 22.6
Hispanic/Latino 3,395 25.6
All Other 1,428 10.8
Suicide Attempt in past 12 months
No 10,967 91.5
Yes 1,015 8.5
11


Table lb Sample Characteristics of Risk and Yotective Factors (N =13,583)
Variable n (%)
Bullied
Bullied on School Property
Yes 2,508 18.6
No 11,007 81.4
Bullied Electronically
Yes 1,878 13.9
No 11,623 86.1
Substance Use
Alcohol Abuse (5+ drinks in a row in past 30 days)
0 days 10,440 79.9
1+ day 2,620 20.1
Marijuana Abuse (Use in last 30 days)
0 times 9,947 74.7
1+ times 3,367 25.3
Media Exposure
TV (Average hours/school day)
Less than 3 8,463 63.9
More than 3 4,782 36.1
Video Game/Intemet Use (Average hours/school day)
Less than 3 7,586 57.2
More than 3 5,669 42.8
Exercise and Sports Participation
Physical Activity (60 min/day)
0-4 days 7,120 53.5
5+ days 6,190 46.5
Sports Team Participation (Previous 12 months)
0 teams 6,151 46.6
1+ teams 7,044 53.4
Data Analysis
Analyses were conducted using SPSS Version 23. YRBS data provides weights based on
gender, race/ethnicity, and grade level. These weights were applied to provide representative data
to adolescents in the United States. Descriptive statistics were used to describe the youth
demographics and their risk and protective factors.
Logistic regression analyses were used to describe the relationship(s) among risk and
protective factors. The logistic regressions were performed to assess how well the following
variables predicted suicide attempts: gender, grade, race/ethnicity, bullied on school property and
12


electronically, substance use- alcohol and marijuana, media exposure- television and video
games/intemet, and exercise and sports participation.
Prevalence of student suicide attempts was calculated based on the three demographic
variables: gender, grade, and race/ethnicity. SPSS Crosstabs function was applied to each risk
characteristic with suicide attempt with chi-square analyses were utilized. Pearson chi-square was
conducted to determine strength of association between risk and protective characteristics and
suicide attempts within demographics. Additionally, pairwise comparisons were made for each
variable and Cramers V and Phi were used to examine effect size and strength of associations.
Column proportions were analyzed for significant differences at the .05 level to determine
differences within risk and protective factors by demographics. These variables were included in
the analysis based on what is known about risk factors for suicidal behavior and to determine
specific risk and protective factors for different groups.
Assumptions. Assumptions of logistic regression were checked and met, including
observations being independent and independent variables being linearly related to the logit of the
dependent variable. Odds ratios (OR) were checked to assess the ratio of the probability of the
occurrence of the outcome variable based on the predictor variable. Pearson chi-square is used
when your research question includes two or more nominal or ordinal variables. The assumptions
of (1) having a large sample size and (2) having at least 5 for expected frequencies in each cell
were checked and met.
13


CHAPTER III
RESULTS
Demographics as Predictors of Suicide Attempt
A logistic regression was conducted to answer the first research question: Do the
demographic characteristics of gender, grade, or race/ethnicity predict whether a student
attempted suicide? When all predictor variables were considered together, they significantly
distinguished between demographic characteristics of students who attempted suicide in the
previous 12 months, %2 = 78.76, df= 1, n = 11,689,/? < .001. This model predicted 4% of the
variance in the outcome variable, Nagelkerke R2 = .037. Table 2 presents the demographic
characteristic predictors for students who attempted suicide, Wald OR, and 95% confidence
intervals (Cl) for the ORs. Gender, race/ethnicity, and grade level were statistically significant
predictors for suicide attempt, specific predictors included females ((3 = -.77, Wald %2 (1) =
120.95 ,/? <001, OR = 2.17), White ((3 = .36, Wald %2 (1) = 10.38,/? <001, OR = .61),
Black/African American ((3 = .11, Wald %2 (1) = .88,/? <.001, OR= .71), Hispanic/Latino ((3 =
-.13, Wald x2 (1) = 1.37,/? =.10, OR= .84), 9th grade ((3 = -.50, Wald %2 (1) = 28.60,/? <001,
OR = .61), and 10th grade ((3 = -.35, Wald %2 (1) = 12.26,/? <.001, OR = .71). Although 12th grade
and all other race/ethnicity, were not found significant, these factors were omitted due to
redundancy. This finding is likely related to the differences in the sample sizes for the all other
race/ethnic group and 12th grade, leading to likely insufficient statistical power to detect
significance for these two groups. Results of these analyses are presented in Table 2.
14


Table 2 Logistic Regression Results for Demographic Predictors of Gender, Race, and Grade Level: Suicide Attempt
Variable B Wald Sig. OR 95% Cl for OR
Gender .77 120.27 <001 2.17 1.89 2.49
Female -.77 120.95 <001 .46 .40 .53
Male 0b
Race/Ethnicity -.17 29.08 <001 .85 .80 .90
White .36 10.38 <001 .61 .50 .73
Black/Affican American .11 .88 <001 .71 .58 .86
Hispanic/Latino -.13 1.37 .10 .84 .69 1.03
All other 0b
Grade Level .17 32.57 <001 1.18 1.12 1.26
9th -.50 28.60 <001 .61 .50 .73
10th -.35 12.26 <001 .71 .58 .89
11th -.17 2.73 .10 .84 .69 1.03
12th 0b
Note. aThe reference category is: Yes. bThis parameter is set to zero because it is redundant.
Risk and Protective Factors as Predictors of Suicide Attempt
A logistic regression was conducted to answer the second research question for students
who attempted suicide in the past 12 months: Does the experience of being bullied on school
property and electronically, substance use- alcohol and marijuana, media exposure- television and
video games/intemet, and exercise and sports participation that best predicts whether a student
attempted suicide? Each risk and protective factor was analyzed by variable group, therefore by
running four logistic regressions by variable group. Table 3 presents the risk and protective
predictors for students who attempted suicide, Wald %, OR, and 95% confidence intervals (Cl)
for the ORs.
When predictor variables of bullying on school property and electronically were
considered together, they significantly predicted whether or a student attempted suicide, %2 =
561.19, df = 2, N = 11,945, p < .001. This model predicted 10% of the variance in die outcome
variable, Nagelkerke R2 = 10 (Table 3). Table 3 presents odds ratios, which suggests being
15


bullied on school property ((3 = -.89, Wald % (1) = 123.29,p <.001, OR = .41), and electronically
((3 = -1.14, Wald % (1) = 191.86,p <.001, OR = .32) were predictors of suicide attempt.
When substance use variables of alcohol and marijuana were considered together, they
significantly predicted whether a student attempted suicide, %2 = 250.13, df = 2, N = 11,500, p <
.001. This model predicted 5% of the variance in the outcome variable, Nagelkerke R2 = .05
(Table 3). Odds ratios, which suggests abusing alcohol 0 times in the past 30 days (P = .55,
Wald % (1) = 43.82,p <.001, OR = 1.73), and using marijuana 0 times (P = .82, Wald %2 (1) =
105.68, p <.001, OR = 2.26) were predictors of suicide attempt.
When media exposure variables of television and video games/intemet use were
considered together, they significantly predicted whether or not a student attempted suicide, %2 =
50.18, df = 2, N = 11,708, p < .001. This model predicted 1% of the variance in the outcome
variable, Nagelkerke R2 = .01 (Table 3). Odds ratios, which suggests watching TV less than
three hours on an average school day (P = .55, Wald %2 (1) = 43.82,p <.001, OR = 1.73), and
spending on average less than three hours playing video games or internet on a school day (P =
.19, Wald x (1) = 7.48, p =.006, OR = 1.21) were predictive of a student suicide attempt.
Additionally, when exercise and sports participation were considered together, they
significantly predicted whether a student attempted suicide, %2 = 36.66 df = 2, N = 11,654, p <
.001. This model predicted 1% of the variance in the outcome variable, Nagelkerke R2 = .007
(Table 3). Odds ratios, which suggests being physically active for 60 minutes for an average of
0-4 days per week (P = -.34, Wald %2 (1) = 22.16,p <.001, OR = .71) were also predictive of a
student suicide attempt, although sports participation was not. Some of the significant findings
were similar to the overall populations report of less substance abuse, media exposure, and video
game/intemet use potentially creating a higher statistical power to reported areas (See Table 3).
16


Table 3 Logistic Regression Results for Predictors of Risk and Protective Factors: Suicide Attempt
Variable B Wald Sig. OR 95% Cl for OR
Bullying
On School Property .89* 123.29 <001 2.42 2.07 2.83
Yes -.89 123.29 <001 .41 .35 .48
No 0b
Electronically 1.14* 191.86 <001 3.12 2.66 3.66
Yes -1.14 191.86 <001 .32 .27 .38
No 0b
Substance Use
Alcohol -.55 43.82 <001 .58 .49 .68
0 days .55 43.82 <001 1.73 1.47 2.03
1+ day 0b
Marijuana -.82 105.68 <001 .44 .38 .52
0 times .82 105.68 <001 2.26 1.94 2.64
1+ time 0b
Media Exposure
TV Use (Average hours/school day) -.19 7.48 .006 .83 .72 .95
Less than 3 .41 35.50 <001 1.50 1.31 1.71
More than 3 0b
Video Game/Intemet Use (Average hours/school day) -.41 35.50 <001 .67 .58 .76
Less than 3 .19 7.48 .006 1.21 1.06 1.38
More than 3 0b
Exercise and Sports Participation
Physical Activity (60 min/day) .34 22.16 <001 1.40 1.22 1.61
0-4 days -.34 22.16 <001 .71 .62 .82
5+ days 0b
Sports Team Participation (Previous 12 months) .15 4.56 .033 1.16 1.01 1.33
0 teams -.15 4.56 .033 .86 .75 .99
1+ teams 0b
Note. aThe reference category is: Yes. bThis parameter is set to zero because it is redundant. *Did not meet assumptions of logistical regression when run as factors
Demographics with Risk and Protective Factors Predicting Suicide Attempt
Pearson chi-square was conducted to answer the third research question: Is there a
combination gender, grade, race/ethnicity, gender, experience of being bullied on school property
and electronically, substance use- alcohol and marijuana, media exposure (television and video
17


games/intemet), and exercise and sports participation that best predicts whether a student
attempted suicide? Each risk and protective factor was analyzed by variable group with
demographics; therefore, four chi-squares were conducted with z-test of column portion
significant differences were applied to examine data. Assumptions were checked and met. Tables
4-7 presents the risk and protective predictors combined with student demographics who
attempted suicide, with Cramers V, Phi, and column z-test.
Bullying. To investigate whether student demographics and bullying on school property
were related to student suicide attempt, a chi-square was used (Table 4). Assumptions were
checked and met. Pearson chi-square results indicate that suicide attempt, gender, and being
bullied on school property are significantly related to one another (%2 = 464.55, df= 1 ,N =
11,953), p < 001). Cramers V. which indicates the strength of the association between the two
variables, was .20 and, thus, the effect size is considered moderate. The Pearson chi-square
results indicate that suicide attempt, race/ethnicity, and being bullied on school property were
also significantly related to one another (y = 565.39, df= 1 ,N= 11.702).p < 001). Cramers V.
which indicates the strength of the association between the two variables, was .22 and, thus, the
effect size is considered moderate. Lastly the Pearson chi-square results also indicated that
suicide attempt, grade level, and school property bullying were also significantly related to one
another (y1 = 576.649, df= 1. N = 11,869),p < 001). Cramers V. which indicates the strength of
the association between the two variables, was .22 and, thus, the effect size is considered
moderate.
18


Table 4 Chi-Square Results for Predictors of Gender, Race, and Grade Level by Bullying Factors: Suicide Attempt
Variable Bullied on School Property Total Bullied Electronically Total
Yes No Yes No
Gender
Female 322a 346b 668 295a 373b 668
Male 128a 211b 339 107a 232b 339
Total 450a 557b 1007 402a 605b 1007
Race/Ethnicity
White 194a 150b 344 187a 158b 345
Black/African American 65a 136b 201 41a 159b 200
Hispanic/Latmo 124a 191b 315 117a 198b 315
All other 59a 64b 123 45a 78b 123
Total 442a 541b 983 390a 593b 983
Grade Level
9th 172a 152b 324 128a 196b 324
10th 118a 131b 249 103a 147b 250
11th 83a 130b 213 76a 135b 211
12th 72a 138b 210 88a 123b 211
Total 445a 551b 996 395a 601b 996
Note. Each subscript letter denotes a subset of Bully categories whose column proportions do not differ significantly from each other at the .05 level. The column proportions test table assigns a subscript letter to the categories of the column variable. For each pair of columns, the column proportions are compared using a z test. If a pair of values is significantly different, the values have different subscript letters assigned to them.
To further investigate the relationship, a z-test as applied to column proportions of
students who attempted suicide and were bullied on school property. Potential differences based
on gender, race/ethnicity, and grade level of students who attempted suicide and were bullied on
school property were also examined. Of all students who reported attempting suicide in the
previous year, there were significantly more who were not bullied on school property (=557,
p=.05) than students who were bullied on school property (n 450, p=. 05; Table 4). Of female
students who had attempted suicide in the previous year, there were significantly more who were
not bullied on school property (n 346, p=.05) than female students who were bullied on school
property (n=346, p=. 05). Additionally, of male students who had attempted suicide in the
19


previous year, there was significantly more who were not bullied on school property (n= 211,
p=.05) than males students who were bullied on school property (/? 128, p=.05).
When considering race/ethnicity, being bullied on school property, and suicide attempt,
z-tests indicated that of White students who attempted suicide, there were significantly more who
were bullied on school property ( = 194, p=.05) than those who were not (/? 150, p=.05).
Although the race/ethnicity groups of Black/African American, Hispanic/Latino, and all other
races did indicate a significant difference between column proportions of those who were bullied
on school property and those who were not, these race/ethnicity groups of students reflected
significantly more students were not bullied on school property. When considering grade level,
being bullied on school property, and suicide attempt, z-tests indicated that among 9th grade
students who attempted suicide, there were significantly more who were bullied on school
property (n= 172, p=.05) than those who were not (/? 152, p=.05). Whereas of 10th, 11th, and 12th
grade students who reported a previous suicide attempt, they reported significantly less bullying
on school property (Table 4).
To investigate whether student demographics and electronic bullying were related to
student suicide attempt, a chi-square was used (Table 4). Assumptions were checked and met.
Pearson chi-square results indicate that suicide attempt, gender, and electronic bullying are
significantly related to one another (%2 = 592.68, df= 1 ,N= 11,947),< 001). Cramers V,
which indicates the strength of the association between the two variables, was .22 and, thus, the
effect size is considered moderate. Suicide attempt, race/ethnicity, and electronic bullying were
also significantly related to one another (% = 565.39, df= 1 ,N= 11.702).p < 001). Cramers V.
which indicates the strength of the association between the two variables, was .22 and, thus, the
effect size is considered moderate. Lastly Pearson chi-square results indicated that suicide
attempt, grade level, and electronic bullying were also significantly related to one another (%2 =
576.649, df= 1 ,N= 11,869), < 001). Cramers V, which indicates the strength of the
association between the two variables, was .22 and, thus, the effect size is considered moderate.
20


The relationship between student demographics and electronic bullying with suicide
attempt was further investigated by applying a z-test to column proportions of students who
attempted suicide. This was done in order to examine if there was a significant difference
between demographic characteristics of students who attempted suicide and were bullied
electronically and those who were not. Of all students who reported attempting suicide in the
previous year (results from gender and electronic bullying chi-square), there were significantly
more who were not bullied electronically (n=605, p=.05) than students who were bullied
electronically (n 402, p=. 05). Of female students who had attempted suicide in the previous year
there was significantly more who were not bullied electronically (n=295, p=.05) than females
students who were bullied electronically (n=373, p=.05). Additionally of male students who had
attempted suicide in the previous year, there were also significantly more who were not bullied
electronically (n=232, p=.05) than males students who were bullied electronically (/? 107,
P= 05).
Although there were significantly fewer reports of being bullied electronically among
students who had attempted suicide in the previous year, z-tests indicated some differences by
race/ethnicity. Similarly to the on-school property bullying results from the z-test, White students
who had previously attempted suicide had significantly higher proportions of students who had
been bullied electronically ( = 187, p=.05) than those who were not ( = 158, p=.05). Compared to
other race/ethnicity categories, only the White students reported higher proportions of electronic
bullying.
Additionally, students who had previously attempted suicide and were bullied
electronically and were analyzed by grade level. Z-test results from column comparisons
indicated that students from all grade levels reported significantly less proportions of electronic
bullying (p=. 05). Overall each grade level was representative of the overall sample of students
who had attempted suicide, where there were significantly higher reports of not being bullied
electronically (n = 601) than those who were ( =395,/?=.05).
21


Substance Use. To investigate whether student demographics and substance use were
related to student suicide attempt, a chi-square statistic was used to compare alcohol use and
marijuana use with student demographic characteristics (Table 5). Assumptions for alcohol use
chi-squares were checked and met. Pearson chi-square results indicate that suicide attempt,
gender, and alcohol use are significantly related to one another (%2 =170.98, df= 1 ,N =
11,599,p <. 05). The Pearson chi-square results for suicide attempt, race/ethnicity, and alcohol
use are significantly related to one another (%2 =168.86, df= 1 ,N= 11,365,p < 05). Lastly, chi-
square results for suicide attempt, grade level, and alcohol use are significantly related to one
another (y2 =165.89, df= 1 ,N= 11,527,p < 05). Cramers V. which indicates the strength of the
association between the two variables was 12 for each chi-square and, thus the effect size is
considered very weak.
Table 5 Chi-Square Results for Predictors of Gender, Race, and Grade Level by Substance Use Factors: Suicide Attempt
Variable Alcohol Total Marijuana Total
0 day 1+ day 0 times 1+ times
Gender
Female 409a 225b 634 376a 277b 653
Male 187a 121b 308 147a 163b 310
Total 596a 346b 942 523a 440b 963
Race/Ethnicity
White 201a 126b 327 179a 158b 337
Black/Affican American 138a 49b 187 iooa 93b 193
Hispanic/Latino 158a 130b 288 157a 143b 300
All other 83a 35b 118 73a OO CO 111
Total 580a 340b 920 509a 432b 941
Grade Level
9th 21 la 96b 307 185a 127b 312
10th 152a 87b 239 132a 109b 241
11th 127a 72b 199 106a 94b 200
12th 102a 86b 188 95a 105b 200
Total 592a 34 lb 933 518a 435b 953
Note. Each subscript letter denotes a subset o proportions do not differ significantly from e table assigns a subscript letter to the categoric the column proportions are compared using a values have different subscript letters assigne ' Substance Use categories w ich other at the .05 level. The ;s of the column variable. For z test. If a pair of values is si d to them. lose column column proportions test each pair of columns, gnificantly different, the
22


The association between student demographics and marijuana use were also investigated
through the Pearson chi-square statistic. Assumptions were checked and met. Pearson chi-square
results indicate that suicide attempt, gender, and marijuana use are significantly related to one
another (%2 =246.37, df= 1 ,N= 11,790,/? < 05), and the effect size is considered weak
(Cramers V= .15). Chi-square results for suicide attempt, race/ethnicity, and marijuana use were
also significantly related to one another (%2 =243.46, df= 1 ,N= 11,548,p < 05), although also
considered a weak effect size (Cramers V= .15). Lastly, chi-square results for suicide attempt,
grade level, and marijuana use are significantly related to one another (%2 =243.77, df= 1 ,N =
11,714 ,p <. 05), and the effect size is considered very weak (Cramers F=.14).
Although each association was considered weak or very weak, further investigations of
gender, race/ethnicity, and grade level of alcohol and marijuana use were analyzed for significant
differences between column proportion z-tests. Results from the z-tests indicated that both female
and male students who had attempted suicide had significantly less alcohol use of 0 days
drinking 5 or more drinks in the past month (female n 409. male n= 187) than those who
reported they had 1 or more days (female n=225, male n 121 ,P<- 05). Although gender was
not a factor indicator of higher reports of prior suicide attempt and alcohol use, gender was a
factor for significantly higher reports of marijuana use and previous suicide attempt. Column
proportions for female students who attempted suicide and used marijuana in the prior month
indicated significantly less marijuana use (n=376 used 0 times, n= 277 used 1+ time), while
their male counterparts indicated significantly higher marijuana use in the past month (w=147
used 0 times, n= 163 used 1+ time,/? < 05).
When comparing race/ethnicity characteristics with alcohol use and previous suicide
attempt, as presented in Table 5, indicated significant reports of 0 days of consuming 5 or more
drinking in the past month. This was also the case across grade levels, where all students who
reported attempting suicide had significantly higher reports of 0 days of using alcohol in excess
during the previous month. Additionally, marijuana use was significantly lower across all
23


race/ethnicities of students who had attempted suicide in the previous year (Table 5). However,
when considering grade level marijuana use in the previous month, 12th grade students who had
attempted suicide in the previous year also reported significantly higher marijuana use in the past
month (n= 105) than those who reported they had not (n= 95,p < 05); whereas Table 5 presents
that all other grade levels had significantly higher reports for using marijuana 0 times over the
past month.
Media Exposure. Pearson chi-square statistic was used to compare student demographics
and media exposure as related to student suicide attempt (Table 6). Assumptions for television
use chi-squares were checked and met. Pearson chi-square results indicate that suicide attempt,
gender, and television use are significantly related to one another (%2 =14.18, df= 1 ,N =
11,729,p <. 05), although Cramers V was .04 indicating the effect size is considered very weak.
The Pearson chi-square results for suicide attempt, race/ethnicity, and television use are
significantly related to one another (%2 =12.81, df 1. N = 11,490,p < 05), although Cramers V
was .03 displaying another very weak effect size. Lastly, chi-square results for suicide attempt,
grade level, and television use are significantly related to one another (%2 =12.432, df 1 ,N =
11,652, p < 05). Cramers V. which indicates the strength of the association between the two
variables, was also .03 and, thus, the effect size is considered very weak.
24


Table 6 Chi-Square Results for Predictors of Gender, Race, and Grade Level by Media Exposure Factors: Suicide Attempt
Variable TV Total Video Game/Internet Total
Less than 3 More than 3 Less than 3 More than 3
Gender
Female 407a 254b 661 313a 349b 662
Male 180a 143b 323 152a 173b 325
Total 587a 397b 984 465a 522b 987
Race/Ethnicity
White 245a 93a 338 174a 165b 339
Black/African American 81a 114a 195 93a 102a 195
Hispanic/Latino 17 la 135b 306 136a 17 lb 307
All other 77a 43a 120 48a 72b 120
Total 574a 385b 959 45 la 510b 961
Grade Level
9th 194a 123a 317 143a 176b 319
10th 142a 99b 241 117a 124b 241
11th 133a 74a 207 103a 104b 207
12th 114a 93b 207 98a Os o 207
Total 583a 389b 972 46 la 513b 974
Note. Each subscript letter denotes a subset of IV proportions do not differ significantly from each test table assigns a subscript letter to the categor columns, the column proportions are compared i different, the values have different subscript lcttc edia Exposure categories whose column other at the .05 level. The column proportions es of the column variable. For each pair of ising a z-test. If a pair of values is significantly us assigned to them.
The relationship between student demographics and video game/intemet use was also
investigated through the Pearson chi-square statistic. Assumptions were checked and met.
Pearson chi-square results indicate that suicide attempt, gender, and video game/intemet use are
significantly related to one another (%2 =43.39, df= 1 ,N= 11,731,p < 05), and the effect size is
considered very weak (Cramers V= .06). Chi-square results for suicide attempt, race/ethnicity,
and video game/intemet use were also significantly related to one another (%2 =43.97, df= 1 ,N =
11,492,p < 05), although also considered a very weak effect size (Cramers V= .06). Lastly, chi-
square results for suicide attempt, grade level, and video game/intemet use are significantly
25


related to one another (%2 =41.00, df= 1 ,N= 11,654,p<. 05), and the effect size is considered
very weak (Cramers V=.06).
Although media exposure effect sizes were considered weak to very weak, further
investigations of gender, race/ethnicity, and grade level of television and video game/intemet use
were analyzed for significant differences between column proportion z-tests. Results from the z-
tests indicated both female and male students who had attempted suicide reported television use
of less than 3 hours on an average school day (female n=407, male n= 180) which was
significantly less than those who reported they watched more than 3 hours of television on an
average school day (female =254, male n= 143,p < 05). Average television use on a school
night of students who had attempted suicide produced similar results for race/ethnicity and grade
levels. Overall, there were significantly higher reports of less than 3 hours on an average school
night across all grade levels and race/ethnicities, except for Black/African American and all other
races categories (Table 6). For Black/African American and all other races, results indicated that
for students who had attempted suicide in the prior year, there was no significant difference
between those who watched less than or more than three hours of television on an average school
night.
As for video game/internet use, z-tests indicated that overall students who had previously
attempted suicide reported significantly higher video game/intemet use (per gender and video
game/intemet use chi-square statistics, Table 6). Gender and video game/intemet use indicated
that both female and male students who had attempted suicide reported video game/intemet use
of more than 3 hours on an average school day (female n 349. male n 173) was significantly
higher than those who reported they watched less than 3 hours of television on an average
school day (female =313, male n= 152,/? < 05). Grade level comparisons indicated similar
findings, where there were significantly higher findings for students who attempted suicide in the
past year reported and more than 3 hours of video game/intemet use on an average school
night. When comparing video game/intemet use across race/ethnicities there was no significant
26


difference between video game/internet use for Black/African American students who had
previously attempted suicide. Presented on Table 6, White students who had previously attempted
suicide reported significantly less video game/internet use on an average school night (=174) of
less than 3 hours, which is different than the collective sample. Whereas Hispanic/Latino and
all other race/ethnicity categories indicated significantly higher video game/internet use of more
than 3 hours on an average school night (Hispanic/Latino w = 171, All other race/ethnicities
n = 72, p < 05).
Exercise and Sports Participation. Pearson chi-square statistic was used to compare
student demographics with exercise and sports participation as related to student suicide attempt
(Table 7). Assumptions for physical activity chi-squares were checked and met. Pearson chi-
square results indicate that suicide attempt, gender, and physical activity are significantly related
to one another (%2 =33.05, df= 1. N = 11,777,p < 05), although Cramers V w as .05 indicating
the effect size is considered very weak. The Pearson chi-square results for suicide attempt,
race/ethnicity, and physical activity are significantly related to one another (%2 =37.26, df= 1 ,N =
11,534,p <. 05), although Cramers V was .06 displaying another very weak effect size. Lastly,
chi-square results for suicide attempt, grade level, and physical activity are significantly related to
one another (%2 =34.45, df 1. N = 11,700,p < 05). Cramers V. which indicates the strength of
the association between the two variables, was also .05 and, thus, the effect size is considered
very weak.
The association between student demographics and sports participation was also
investigated through the Pearson chi-square statistic. Assumptions were checked and met.
Pearson chi-square results indicate that suicide attempt, gender, and sports participation use are
significantly related to one another (% =43.39, df= 1 ,N= 11,731,p < 05), and the effect size is
considered very weak (Cramers V= .04). Chi-square results for suicide attempt, race/ethnicity,
and sports participation were also significantly related to one another (%2 =14.79. df 1 ,N =
11,428, p <. 05), although also considered a very weak effect size (Cramers V= .04). Lastly, chi-
27


square results for suicide attempt, grade level, and sports participation are significantly related to
one another (%2 =14.59, df= 1 ,N= 11,592,p<. 05),and the effect size is considered very weak
(Cramers F=.06).
Table 7 Chi-Square Results for Predictors of Gender, Race, and Grade Level by Exercise and
Sports Participation Factors: Suicide Attempt
Variable Physical Activity Total Sports Team P articipation Total
0-4 days 5+ days 0 teams 1+ teams
Gender
Female 438a 226, 664 362, 296, 658
Male 176a 153b 329 152, 177b 329
Total 614a 379b 993 514, 473b 987
Race/Ethnicity
White 203a 141b 344 187, 154b 341
Black/African American 130a 65b 195 95, 100, 195
Hispanic/Latino 193a 115b 308 160, 145, 305
All other 80a 40b 120 61, 59, 120
Total 606a 361b 967 503, 458b 961
Grade Level
9th 194a 125b 319 157, 161b 318
10th 154, Xi o\ 00 243 124, 118b 242
11th 131, 78b 209 108, 100, 208
12th 130, 80, 210 120, 88, 208
Total 609, 372b 981 509, 467b 976
Each subscript letter denotes a subset of Exercise and Sports Participation categories whose column proportions do not differ significantly from each other at the .05 level. The column
proportions test table assigns a subscript letter to the categories of the column variable. For each pair of columns, the column proportions are compared using a z test. If a pair of values is significantly different, the values have different subscript letters assigned to them.
Although exercise and sports participation effect sizes were considered weak to very
weak, further investigations of gender, race/ethnicity, and grade level of physical activity and
sports team participation were analyzed for significant differences between column proportion z-
tests (Table 7). Results from the z-tests, per gender and physical activity chi-square, indicated that
of students who had attempted suicide in the previous year were reporting 0 to 4 days of at least
60 minutes of physical activity in the previous week (n=614) that was significantly higher than
those who were physically active for 5 or more days (n=379, p < 05). Overall suicide
attempters reported similarly across gender, race/ethnicity, and grade level. Across demographic
28


categories, all responses displayed significantly higher scores for 0 to 4 days of at least 60
minutes of physical activity in the previous week (Table 7).
As for sports participation, per gender and sports participation chi-square, z-tests
indicated that overall, students whom attempted suicide in the previous year had significantly
higher reports of participation in 0 sports teams during the previous year (n=514 0 teams,
n 473 1 or more,/? < 05). However, attempting suicide and participating on 0 teams over
the previous year were not consistently significant across demographics. When comparing gender
and sports participation, there was no significant difference between female students who had
previously attempted suicide and sports participation (n=362 0 teams, n 296 1 or more,/? <
05). Additionally, male students who had attempted suicide in the previous year indicated
significantly more sports participation (n 152 0 teams, n 177 1 or more,/? < 05). Race and
ethnicity identifiers also indicated differences. For White students who had attempted suicide, the
majority reported no sports participation during the previous year (n 187 0 teams, /? 154 1 or
more,/? < 05). However, Black/African American, Hispanic/Latino, and all other race/ethnicity
categories of students whom previously attempted suicide indicated no significant differences
between sports participation (Table 7). Additionally, for 9th grade students, significantly more
suicide attempters reported participation in sports over the previous year (n 157 "0 teams,
n 161" 1 or more,/? < 05). Tenth grade students were similar to the overall sample of student
attempters, where there was significantly more 10th grade attempters who did not participate on a
sports team during the previous year (n 124 0 teams, /? 1 18" 1 or more,/! < 05). However,
11th and 12th grade student attempters indicated no significant differences between sports
participation (Table 7).
29


CHAPTER IV
DISCUSSION
The purpose of the current study was to assess how well the following variables predicted
suicide attempts: gender, grade, race/ethnicity, bullied- on school property and electronically,
substance use- alcohol and marijuana, media exposure- television and video games/internet, and
exercise and sports participation. Examination of these particular variables was based on what is
currently known about risk and protective factors of suicidal high school students in order to
identify specific supports and behavioral risks for students among different groups. The results of
this study reveal that the relationship between suicidal students and known risk and protective
factors is complex. An analysis of data from a nationally representative sample of high school
students suggests that across demographic characteristics of suicidal students risk and protective
factors vary on being predictive of a suicide attempt.
Interpretation
Previous research indicates that student demographic variables of gender and
race/ethnicity were predictors of suicide attempt; specifically for students who are female,
Black/African American, and/or Hispanic/Latino (Pena et. al, 2012). Findings from this research
study supported previous findings that female, Black/African American, and/or Hispanic/Latino
students were demographic risk factors for predicting suicide attempt. Results of the present study
additionally found that students who were White were also at a higher risk for suicide.
Additionally, this study found that students who were in the 9th or 10th grade were also predictive
factors for suicide.
Bullying was previously established as a risk factor for adolescent suicide (Klomek et. al,
2010). Findings from this study supported previous research in which both being bullied on
school property and electronically were significant predictors of student suicide attempt. Across
gender, bullying was a moderate predictor of attempting suicide. Students who attempted suicide
who were White or in the 9th grade reported being bullied both electronically and on school
30


property significantly more than not; however, the majority of suicide attempters indicated they
had not been bullied within the past 12 months. These are not expected findings given previous
literature related to bullying as a suicide risk factor. Although bullying is moderate predictor of
suicide, it is an interesting finding that overall, more attempters reported they had not been
bullied.
Additionally, substance abuse variables of marijuana and alcohol were significantly
predictive of suicide attempt; however, findings indicated that alcohol and marijuana use were
only a weak to very weak effect size as predictors of adolescent suicide attempt. Although
significant, the majority of student attempters did not use marijuana or drink excess alcohol in the
previous month. These findings are surprising given the results of other studies have indicated
alcohol and drug use are indicators of suicidal high school students (Borowsky et al., 2001;
Esposi et al., 2004). This study also found that male students, who had attempted suicide in the
past year, also indicated significantly higher marijuana use. Also 12th grade students, who had
previously attempted suicide in the past year, also reported higher rates of marijuana use. These
findings indicate that marijuana and alcohol use may only be weak risk factors for a suicide
attempt; however, students who use marijuana and are male and/or in the 12th grade may be at
higher risk for a suicide attempt.
A relationship between video gaming, internet use, and suicide has been supported by
previous research (Kim et al, 2008; Mathers et al., 2009; Rehbein et al., 2010; Weaver et al.,
2009; Lam et al., 2010). Findings from the current study are aligned with previous findings that
across demographic variables, students who had previously attempted suicide reported
significantly higher video game/internet use and that the majority reported watching less than 3
hours of television on an average school night; both of which were representative of the sample
of suicide attempters. This was an interesting finding, indicating that overall, students who
attempted in the past year were also using the internet and/or playing video games for more than
3 hours and spending less than 3 hours per day watching television. These findings may
31


indicate that videogame/internet use may be a risk factor for suicidal students; however, since this
finding was representative of the overall sample, it may indicate that video game/intemet use and
television are neither a protective nor a risk factor for adolescent suicide.
Other studies also provided evidence of physical inactivity was a risk factor for suicide
(Simon et al., 2004; Chioqueta et al., 2007). Youth physical activity guidelines, according to the
U.S. Department of Health and Human Services, recommend that children and adolescents aged
6-17 years should have 60 minutes or more of physical activity each day (2008). Another
unexpected finding was that physical activity and sports participation, although significant,
maintained a weak to very weak effect size. Overall, significantly higher proportions of suicide
attempters reported being physically active for 0-4 days during the past school week than those
who reported physical activity of 5 or more days over the past school week. This was
representative of the overall sample of students. Furthermore, suicide attempters were more likely
to not participate on a sports team over the previous year; however, sports participation varied
across the demographic variables. Of female suicide attempters, there was no significant
difference between the level of sports participation, indicating that sports participation is not a
significant protective or risk factor for female students. The same finding was true for race/ethnic
groups of Black/African American, Hispanic/Latino, and all other race/ethnicities, which also
indicates that increased level of sports participation is not a significant protective or risk factor for
these groups of students. During the previous year, there were higher reports of male attempters
who participating in one or more sports teams than those who did not. Ninth grade attempters also
reported higher sports participation over the previous year, which may also indicate that higher
levels of sports participation were not protective factors for these groups. This is an unexpected
finding considering previous research supports involvement in certain extracurricular activities,
particularly sports, as a strong protective factor against adolescent suicide (Babiss & Gangwisch,
2009; Taliaferro, Rienzo, Miller, Pigg, & Dodd, 2008).
32


In conclusion, similar to currently known risk and protective factors for high school
students, this study found significant predictors of suicide attempt including: (1) female gender,
(2) race/ethnicity of White, Black/African American, and Hispanic/Latino, (3) lower grade level,
(4) experience of bullying of both on school property and electronically, (5) fewer than 5 days of
physical activity per week, and (6) video game/intemet use. Although data indicated that
demographic characteristics of gender, race/ethnicity, and grade level were moderately associated
with attempting suicide, specific group analysis indicated that there are unique subgroups of
students that are affected by risk and protective factors differently.
Limitations and Future Research Directions
There are a number of limitations of the current study. All data were self-reported and
accuracy of report cannot be examined. Data was then cross-sectionally analyzed and therefore,
any causality cannot be concluded. Another limitation relates to the fact that only participants
who attended school were included in the data; therefore, students who were not in attendance the
day of the survey and/or do not attend school could not be included in the sample. This is an
important limitation due to the suicide risk for adolescents in the justice system and minority
populations often do not attend public school or graduate (National Center for Education
Statistics, 2010). Another limitation is the measurement of the grade levels, which are not precise
due to students being retained and large age ranges.
A further limitation is that state-level data were not available for all 50 states, as some did
not participate and were therefore unable to obtain and present completely representative data.
Additionally, YRBS questions directly addresses behaviors that contribute to the leading causes
of illness and death among United States adolescents and adults, which limits school and
community focus on targeted proactive and positive interventions for suicidal behavior. Also,
substance abuse questions of the YRBSS were limited in scope, with questions only focused on
substance abuse over the prior month rather that into the scope and duration of use.
33


This study examined only the relationship between demographic characteristics and risk
and protective factors of suicide attempters. This study did not consider the differences between
the suicide attempters and non-attempters. An examination of these differences may indicate
potential significance in the effect of these risk and protective factors on each population.
Furthermore, this study did not include consideration of other known predictors of suicidal
behavior such as depression, anxiety, and other mental health concerns.
Additionally, this study also did not analyze the impact on subgroup geography, which
may have an impact on the effect of risk and protective factors. Research indicates that suicide
rates in rural areas are higher than urban areas (Hirsch, 2006), which may influence differences in
opportunities and risks.
Finally, there were many small sample sizes that found little to no significance due to
multiple cross-sectional analyses. For future research, combining additional YRBS data may
increase sample sizes and therefore statistical power of analysis.
Implications for Practice
It is critical that student risk factors be identified in order to appropriately assess for
suicide risk and provided appropriate support and intervention based on their unique
characteristics. The findings of the present study have a number of implications for practice.
School mental health professionals will need to be able to identify various suicide risk and
protective factors in order to appropriately assess students at-risk for suicide. When risk factors
for suicide come to the attention of school personnel, it is crucial for school personnel and mental
health to be aware of unique risk factors for different demographic groups of students in order to
appropriately address student suicide risks.
The results of this study have implications for the professional training and practice for
those who work with adolescent populations. In promoting proper identification and assessment
of students who are considering suicide or are at risk, there are many promising programs in
means to prevent adolescent suicide. Mental health professionals who understand suicide risk
34


and implement evidence-based suicide prevention programs are crucial to saving adolescent lives.
It is essential that school psychologists and other community-based mental health professionals
are current with evolution of effective assessment and intervention of the continuously changing
and evolution of the adolescent population. In order to maintain authority on effective
intervention procedures, training would be beneficial through post-secondary and continuing
education opportunities (Crepeau-Hobson, 2013).
Competing Interests
I confirm that there are no known conflicts of interest associated with this publication and
there has been no significant financial support for this work that could have influenced its
outcome.
35


REFERENCES
American Academy of Child and Adolescent Psychiatry, American Psychiatric Association
(2004). Joint statement from the American Academy of Child and Adolescent Psychiatry
and the American Psychiatric Association for the hearing on suicide prevention and
youth: saving lives. Senate Substance Abuse and Mental Health Services Subcommittee
of the Health, Education, Labor, and Pensions. http://www.aacap.org/galleries/Legislative
Action/SuicideH.pdf. Accessed 5 November 2015.
Aseltine, R., James, A., Schilling, E., & Glanovsky, J. (2007). Evaluating the SOS suicide
prevention program: A replication and extension. Bmc Public Health, 7(1), 161-161.
doi: 10.1186/1471-2458-7-161
Babiss, L., & Gangwisch, J. (2009). Sports participation as a protective factor against depression
and suicidal ideation in adolescents as mediated by self-esteem and social support.
Journal of Developmental and Behavioral Pediatrics, 30(5), 376-384.
doi:10.1097/DBP.0b013e3181b33659
Buda, B., Berman, A. L., Jobes, D. A., & Silverman, M. M. (2007). Adolescent suicide.
Assessment and intervention (2nd ed.): (2006). Washington, DC: American
Psychological Association. Crisis: The Journal of Crisis Intervention and Suicide
Prevention, 28(2), 105-105. doi: 10.1027/0227-5910.28.2.105
Bonanno, R. A., & Hymel, S. (2013). Cyber bullying and internalizing difficulties: Above and
beyond the impact of traditional forms of bullying. Journal of Youth and Adolescence,
42, 685-697. doi:10.1007/sl0964-013-9937-l
Bossarte, R. M., & Swahn, M. H. (2011). The associations between early alcohol use and suicide
attempts among adolescents with a history of major depression. Addictive Behaviors,
36(5), 532-535. doi: 10.1016/j.addbeh.2010.12.031
Borowsky, I., Ireland, M., & Resnick, M. (2001). Adolescent suicide attempts: risks and
protectors. Pediatrics, 107:485-93.
Brener, N. D., Collins, J. L., Kann, L., Warren, C. W., & Williams, B. I. (1995). Reliability of the
youth risk behavior survey questionnaire. American Journal of Epidemiology, 141, 575-
580.
Brener, N. D., Kann, L., McManus, T., Kinchen, S. A., Sundberg, E. C., & Ross, J. G. (2002).
Reliability of the 1999 youth risk behavior survey questionnaire. Journal of Adolescent
Health, 31(4), 336-342.
Brener, N. D., McManus, T., Galuska, D. A., Lowry, R., & Wechsler, H. (2003). Reliability and
validity of self-reported height and weight among high school students. Journal of
Adolescent Health, 32(4), 281-287.
Brown, D., Galuska, D., Zhang, J., Eaton, D., Fulton, J., Lowry, R., & Maynard, L. (2007).
Physical activity, sport participation, and suicidal behavior: US high school students.
Medicine and Science in Sports and Exercise, 39 (12), 2248-2257.
doi:10.1249/mss.0b013e31815793a3
36


Brown, C. H., Wyman, P. A., Brinales, J. M., & Gibbons, R. D. (2007). The role of randomized
trials in testing interventions for the prevention of youth suicide. International Review of
Psychiatry, 19(6), 617-631. doi: 10.1080/09540260701797779
Cash, S., & Bridge, J. (2009). Epidemiology of youth suicide and suicidal behavior. Current
Opinion in Pediatrics, 21(5), 613-619.
Centers for Disease Control and Prevention. (2012). Youth Risk Behavior Survey. Available at:
www. cdc. gov/yrbss.
Centers for Disease Control and Prevention. (2013). Youth Risk Behavior Survey. Available at:
www. cdc. gov/yrbss.
Centers for Disease Control and Prevention. (2014). Youth Risk Behavior Survey. Available at:
www. cdc. gov/yrbss.
Centers for Disease Control and Prevention. (2010a). Web-based injury statistics query and
reporting system (WISQARS) Retrieved November 6, 2015, from
www.cdc.gov/injury/wisqars/pdf/10LCID_All_Deaths_By _Age_Group_2010-a.pdf
Chioqueta, A. P., & Stiles, T. C. (2007). Cognitive factors, engagement in sport, and suicide risk.
Archives of Suicide Research, 11(4), 375-390. doi: 10.1080/13811110600897143
Chung, S. S., & Joung, K. H. (2012). Risk factors related to suicidal ideation and attempted
suicide: Comparative study of Korean and American youth. The Journal of School
Nursing, 28(6), 448-458. doi: 10.1177/1059840512446704
Chabrol, H., Melioli, T., & Goutaudier, N. (2014). Cannabis use and suicidal ideations in high-
school students. Addictive Behaviors, 39(12), 1766-1768.
doi:10.1016/j.addbeh.2014.06.008
Crepeau-Hobson, F. (2013). an exploratory study of suicide risk assessment practices in the
school setting. Psychology in the Schools, 50(8), 810-822. doi: 10.1002/pits.21705
Crone, E., Dahl, R. (2012) Understanding adolescence as a period of social-affective engagement
and goal flexibility. Nature reviews. Neuroscience 13: 636-50.
Dahl, R. (2004) Adolescent brain development: a period of vulnerabilities and opportunities.
Keynote address. AnnN Y Academic Science 1021:1-22
Department, A. C. (2001). Practice parameter for the assessment and treatment of children and
adolescents with suicidal behavior. Journal of the American Academy of Child &
Adolescent Psychiatry, 40(1), 24S-51S. doi: 10.1097/00004583-200107001-00003
Degenhardt, L., Hall, W., & Lynskey, M. (2003). Exploring the association between cannabis use
and depression. Addiction, 9<3(11), 1493-1504. doi: 10.1046/j. 1360-0443.2003.00437.x
Dunn, M. S., Goodrow, B., Givens, C., & Austin, S. (2008). Substance use behavior and suicide
indicators among rural middle school students. The Journal of School Health, 78, 26-31.
37


Epstein, J.A., & Spirito, A. (2010) Gender-Specific Risk Factors for Suicidality Among High
School Students, Archives of Suicide Research, 14(3), 193-205,
doi:10.1080/13811118.2010.494130
Fine, J., Sung, C. (2014) Neuroscience of child and adolescent health development. Counseling
Psychology 61:521-527
Freedenthal, S. (2010). Adolescent help-seeking and the yellow ribbon suicide prevention
program: An evaluation. Suicide and Life-Threatening Behavior, 40(6), 628-639.
doi:10.1521/suli.2010.40.6.628
Gauthier, J. M., Zuromski, K. L., Gitter, S. A., Witte, T. K., Cero, I. J., Gordon, K. H., & ...
Joiner, T. (2014). The Interpersonal-Psychological Theory of Suicide and Exposure to
Video Game Violence. Journal of Social & Clinical Psychology, 33(6), 512-535.
doi:10.1521/jscp.2014.33.6.512.
Heron, M. (2013). Deaths: Leading causes for 2010. National vital statistics reports, 62(6).
Hyattsville, MD: National Center for Health Statistics.
Hinduja, S., & Patchin, J. W. (2010). Bullying, cyberbullying, and suicide. Archives of Suicide
Research, 14(3), 206-221. doi: 10.1080/13811118.2010.494133
Hirsch, J. K. (2006). A review of the literature on rural suicide: Risk and protective factors,
incidence, and prevention. Crisis,27(4), 189-199. doi: 10.1027/0227-5910.27.4.189
Johnston, L. D., OMalley, P. M., Bachman, J. G., & Schulenberg, J. E. (2008). Demographic
subgroup trends for various licit and illicit drugs, 1975-2007 (Monitoring the Future
Occasional Paper No. 69) [Online]. Ann Arbor, MI: Institute for Social Research.
Retrieved from http:// www.monitoringthefuture.org/
Kerr, D., Owen, L., Pears, K., & Capaldi, D. (2008). Prevalence of suicidal ideation among boys
and men assessed annually from ages 9 to 29 years. Suicide and Life-Threatening
Behavior, 38(4), 390-402.
Kim, D. S., & Kim, H. S. (2010). Early initiation of alcohol drinking, cigarette smoking, and
sexual intercourse linked to suicidal ideation and attempts: Findings from the 2006
Korean youth risk behavior survey. Yonsei Medical Journal, 51, 18-26.
Kim, E. J., Namkoong, K., Ku, T., & Kim, S. J. (2008). The relationship between online game
addiction and aggression, self-control and narcissistic personality traits. European
Psychiatry, 23, 212-218.
Klaus, N. M., Mobilio, A., & King, C. A. (2009). Parent-adolescent agreement concerning
adolescents' suicidal thoughts and behaviors. Journal of Clinical Child & Adolescent
Psychology, 38(2), 245-255. doi: 10.1080/15374410802698412
Klomek, A., Sourander, A., & Gould, M. (2010). The association of suicide and bullying in
childhood to young adulthood: A review of cross-sectional and longitudinal research
findings. Canadian Journal of Psychiatry, 55(5), 282-288.
38


Konrad, K., Firk, C., & Uhlhaas, P. J. (2013). Brain Development During Adolescence:
Neuroscientific Insights Into This Developmental Period. Deutsches Arzteblatt
International, 110(25), 425431. http://doi.org/10.3238/arztebl.2013.0425
Lam, L. T., & Peng, Z. W. (2010). Effect of pathological use of the internet on adolescent mental
health: A prospective study. Archives of Pediatrics and Adolescent Medicine, 164, 901
906.
Fanslow, J. L. (1999). Making sense of suicide: An in-depth look at why people kill themselves.
Oxford: Pergamon Press Inc.
Lester, D. (Ed.). (1997). Making sense of suicide: An in-depth look at why people kill themselves.
Philadelphia: Charles PressPub.
Malley, E., Posner, M., & Potter, L. (2008). Suicide risk and prevention for lesbian, gay, bisexual,
and transgender youth. Newton: Suicide Prevention Resource Center, 35-36.
Mathers, M., Canterford, L., Olds, T., Hesketh, K., Ridley, K., & Wake, M. (2009). Electronic
media use and adolescent health and well-being: Cross-sectional community study.
Academic Pediatrics, 9, 307-314.
Messias, E., Castro, J., Saini, A., Usman, M., & Peeples, D. (2011). Sadness, Suicide, and Their
Association with Video Game and Internet Overuse among Teens: Results from the
Youth Risk Behavior Survey 2007 and 2009. Suicide & Life-Threatening Behavior,
41(3), 307-315. doi: 10.1111/j. 1943-278X.2011< 00130.x
McIntosh, J. L. (2012). USA Suicide: 2009 Official Final Data. Washington, DC: American
Association of Suicidology. Retrieved November 5, 2015 from
http: //www .suicidology. org
Miller, K., & Hoffman, J. (2009). Mental well-being and sport-related identities in college
students. Sociology of Sport Journal, 26(2), 335-356.
Murphy, S., Xu, J., & Kochanek, K. (2012). Deaths: preliminary data for 2012. National Vital
Statistics Reports, 60, 1-51.
National Center for Education Statistics. (2010). Status and trends in the education of racial and
ethnic minorities. Retrieved from
http://nces.ed.gov/pubs2010/2010015/indicator4_l 8.asp
Nickerson, A. B., & Slater, E. D. (2009). School and community violence and victimization as
predictors of adolescent suicidal behavior. School Psychology Review, 38(2), 218.
Pena, J. B., Matthieu, M. M., Zayas, L. H., Masyn, K. E., & Caine, E. D. (2012). Co-occurring
risk behaviors among white, black, and hispanic US high school adolescents with suicide
attempts requiring medical attention, 1999-2007: Implications for future prevention
initiatives. Social Psychiatry and Psychiatric Epidemiology, 47(1), 29-42.
doi: 10.1007/s00127-010-0322-z
39


Pedersen, W. (2008). Does cannabis use lead to depression and suicidal behaviours? A
population-based longitudinal study. Acta Psychiatrica Scandinavica, 118(5), 395-403.
doi: 10.1111/j. 1600-0447.2008.01259.X
Price, C., Hemmingsson, T., Lewis, G., Zammit, S., & Allebeck, P. (2009). Cannabis and suicide:
Longitudinal study. The British Journal of Psychiatry, 195(6), 492-497.
doi: 10.1192/bjp.bp. 109.065227
Rasic, D., Weerasinghe, S., Asbridge, M., & Langille, D. (2013). Longitudinal associations of
cannabis and illicit drug use with depression, suicidal ideation and suicidal attempts
among nova scotia high school students. Drug and Alcohol Dependence, 129(1-2), 49-53.
doi:10.1016/j.drugalcdep.2012.09.009
Rehbein, F., Kleimann, M., & Mossle, T. (2010). Prevalence and risk factors of video game
dependency in adolescence: Results of a German nationwide survey. CyberPsychology
and Behavior, 13, 269-277.
Roberts, R.E., Ramsay, C., Yun, X., (2010). One-year incidence of suicide attempts and
associated risk and protective factors among adolescents. Arch Suicide Res 14:66-78
Roy, A., Lamparski, D., DeJong, W., Moore, V., Linnoila, M. (1990). Characteristics of
alcoholics who attempt suicide. The American Journal of Psychiatry, 147, 761-765
Rylander, M., Valdez, C., & Nussbaum, A. M. (2014). Does the legalization of medical marijuana
increase completed suicide?. American Journal Of Drug & Alcohol Abuse, 40(4), 269-
273. doi: 10.3109/00952990.2014.910520
Shilubane, H. N., Ruiter, R. A. C., Bos, A. E. R., van den Borne, B., James, S., & Reddy, P. S.
(2012). Psychosocial determinants of suicide attempts among black South African
adolescents: A qualitative analysis. Journal of Youth Studies, 15(2), 177-189.
doi: 10.1080/13676261.2011.634400
Simon, T. R., Powell, K. E., & Swann, A. C. (2004). Involvement in physical activity and risk for
nearly lethal suicide attempts. American Journal of Preventive Medicine, 27(4), 310-315.
doi: 10.1016/j.amepre.2004.07.003
Taliaferro, L. A., Rienzo, B. A., Miller, M. D., Pigg, R. M., Jr., & Dodd, V. J. (2008). High
school youth and suicide risk: Exploring protection afforded through physical activity
and sport participation. The Journal of School Health, 78, 545-553.
Taliaferro, L. A., Rienzo, B. A., Pigg, R. M., Miller, M. D., & Dodd, V. J. (2009). Associations
between physical activity and reduced rates of hopelessness, depression, and suicidal
behavior among college students. Journal of American College Health, 57(4), 427-435.
doi: 10.3200/JACH.57.4.427-436
Undheim, A. M. (2013). Involvement in bullying as predictor of suicidal ideation among 12- to
15- year-old Norwegian adolescents. European Child & Adolescent Psychiatry, 22, 1-9.
doi: 10.1007/s00787-012-0373-7
U.S. Department of Health and Human Services (2008). Physical Activity Guidelines for
Americans. Washington, DC: U.S. Department of Health and Human Services. Retrieved
from http://www.cdc.gov/healthyschools/physicalactivity/guidelines.htm
40


Wang, J., Nansel, T. R., & Iannotti, R. J. (2011). Cyber bullying and traditional bullying:
Differential association with depression. The Journal of Adolescent Health : Official
Publication of the Society for Adolescent Medicine, 48(4), 415-417.
doi: 10.1016/j .j adohealth. 2010.07.012
Weaver, J. B., 3rd, Mays, D., Sargent Weaver, S., Kannenberg, W., Hopkins, G. L., Eroglu, D., et
al. (2009). Health-risk correlates of video-game playing among adults. American Journal
of Preventive Medicine, 37, 299-305.
Wolitzky-Taylor, K. B., Ruggiero, K. J., McCart, M. R., Smith, D. W., Hansen, R. F., Resnick,
H. S.,. Kilpatrick, D. G. (2010). Has adolescent suicidality decreased in the United
States? Data from two national samples of adolescents interviewed in 1995 and 2005.
Journal of Clinical Child and Adolescent Psychology, 39, 64-76.
Wyman, P. A., Brown, C. H., Inman, J., Cross, W., Schmeelk-Cone, K., Guo, J., & Pena, J. B.
(2008). Randomized trial of a gatekeeper program for suicide prevention: 1-year impact
on secondary school staff. Journal of Consulting and Clinical Psychology, 76(1), 104-
115. doi: 10.1037/0022-006X.76.1.104
41


Full Text

PAGE 1

RISK AND PROTECTIVE FACTORS OF SUICIDAL ADOLESCENTS by KELSEY STAPPERT B.G.S., University of Kansas, 2010 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the deg ree of Doctor of Psychology School Psychology Program 2016

PAGE 2

! ii 2016 KELSEY STAPPERT ALL RIGHTS RESERVED

PAGE 3

! iii This thesis for the Doctor of Psychology degree by Kelsey Stappert has been approved for the School Psychology Program by Fran ci Crepeau Hobson, Chair Bryn Harris Colette Hohnbaum March 31, 2016

PAGE 4

! iv Stappert, Kelsey (Psy D School Psychology) Risk and Protective Factors of Suicidal Adolescents Thesis directed by Associate Professor Franci Crepeau Hobson ABSTRACT Recognizing risk and protective factors for adolescent suicide attempts is an important factor for identification and suicide prevention strategies. The purpose of the study wa s to assess the relationship between adolescent risk factors and suicide. National ly represe ntative data from the 2013 Youth Risk Behavior Survey (YRBS) created by the Centers for Disease Control and Prevention (CDC) were used. The relationship of demographic characteristics, bullying, substance abuse, media exposure, and physical activity to sui cide attempt s was assessed. Results concluded distinguishable characteristic among demographics and suicide risk factors: gender, race/ethnicity, grade level, experience of bullying school property and electronically, exercise physical activity and sports participation, and media exposure video game/internet use. Implications for practice and future research are discussed. The form and content of this abstract are approved. I recommend its publication. Approved: Franci Crepeau Hobson

PAGE 5

! v ACKNOWLEDGEMENTS Life has been a spectacular journey thus far. I have been fortunate to have many supportive mentors who have provided me with the expertise and experience that lead me to constantly challenge my knowledge of the world. Without each of you, this wonderful j ourney may not have happened. Thank you for your patience and guidance along the way.

PAGE 6

! vi TABLE OF CONTENTS CHAPTER I. INTRODUCTION ............................................................ ............ ........... ............................... 1 Adolesc ent Suicide Prevalence and Risk ... .............. .. ........... ......... ............................... 1 Suicide Risk and Protective Factors .................. ...... .... ...... .............................. .......... 1 Demographics ..................... ................ .......... ......................................... ........ 1 Bullying ............................................ ........... ............................................ ....... 2 Substance Abuse .............................. ......... ............................................. ........ 2 Media Exposure .............................................................................. ...... .......... 3 Exercise and Sports Participation ........................... ...................... ........ .......... 4 Sp ecific Aims............................... .......................... ....................... ........... ... ...... ........... 4 II. METHODS .... .. ............................................................................. ........ ... .. .. .............. ......... 7 Sample ..... ............. ............................ ..................... .............. ......... ... .. ......................... 7 Measures ........................................ ............ ..................... ....... ......................... ......... .... 8 Demographics ........... ........... ................................................................... ........ 8 Bullying ........................ ........... .................................................... ............... .... 8 Substance Abuse .......... .......... ........................... ........................ .............. ....... 9 Media Exposure ............................................... .... ................. .......... .......... ..... 9 Exe rcise and Sports Participation ........................... ........................ ...............10 Suicide ........................................................................................ .................. 11 Data Analysis ................... ...................................................................... ................... 12 Assumptions ............................................................ ...................................... 13 II I RESULTS .................................. ................................................................... ..................... 14 Demographics as Predictors of Suicide Attempt ... ................................................ 14 Risk and Protective Factors as Predictors of Suicide Attem pt ................ .... .............. 15 Demographics with Risk and Protective Fa ctors Predicting Suicide Attempt........... 17

PAGE 7

! vii Bullying ........................ ..... ........................................................................... 17 Sub stance Abuse ..................................................................................... ..... 22 Media Exposure ........................................................................................ ....24 Exercise and Sports Participation ................................................... ..............27 IV. DISCUSSION.............................................................. .. ............ ................ ....... ............... 30 Interpretation..................................... ......................................................... ............... 30 Limitations and Future Research Directions ...................................................... ...... 33 I mplications for Practice .......................................... ........................................ ......... 3 4 Competing Interests .............................................................................................. .... 35 REFERENCES..................................................................... .......... ..................... ........... .................... 36

PAGE 8

! viii LIST OF TABLES TABLE 1a. Sample Characteristics of the Population ................................................................ ..................... 11 1b. Sample Characteristics of R isk and Protective Factors .................................................. .............. 12 2. Logistic Regression Results for Demographic Predictors of Gender Race and Grade Level: Suicide Attempt .. ... ......... .................. ....................... ............................... .................. ............................... 15 3. Logistic Regression Results for Predictors of Risk and Protective Factors: Suicide Attempt ....... 17 4. Chi Square Results for Predictors of Gender, Race, and Grade Lev el by Bullying Factors: Suicide Attempt ............... ....................................................................................... ...... ............................. 19 5. Chi Square Results for Predictors of Gender, Race, and Grade Level by S ubstance Use Factors: Suicide Attempt ................................................................................................ ............................. 22 6. Chi Square Results for Predictors of Gender, Race, and Grade Level by Media Exposure F actors: Suicide Attempt ................................................................................................ ............................. 25 7. Chi Square Results for Predictors of Gender, Race, and Grade Level by Exercise and Sports Participat ion Factors: Suicide Attempt ...................................... ............................ 28

PAGE 9

! 1 CHAPTER I INTRODUCTION Adolescent Suicide Prevalence and Risk According to the Center s for Disease Control (CDC), suicide is the third leading cause of death for people aged 15 to 24 years (CDC, 2010a; Buda, Berman, Jobes & Silverman, 2007 ; Murphy, Xu, & Kochanek, 2012; Heron, 2013). The American Association of Suicidology indicates increasing rates of depression and death by suicide for across a dolescent populations (McIntosh, 2012). With adolescent suicide rates on the rise, there has also been subsequent advancement in identifying and understanding fac t ors associated with adolescent suicide (Cash & Bridge, 2009). T he developmental stage of ad olescence, a phase of life when children transition to adulthood, is a crucial period of physical maturation and brain development (Crone & Dahl, 2012; Konrad, Firk, & Uhlhaas, 2013). W ith changes occurring in brain development during adolescence, there is often an increase in risk taking behaviors and sensation seeking (Dahl, 2004; Fine, & Sung, 2014) Some of these risk taking behaviors can lead to exposure to potential adverse life events and an increased vulnerability to mental illness (Dahl, 2004). By further examining the risk taking behaviors of suicidal adolescents, it may point to potential prevention initiatives and strategies to combat further adolescent death. Suicide Risk and Protective Factors Since suicide is one of the leading causes of adol escent death, identifying suicide related behaviors and both risk and protective factors is of critical importance. R isk and protective factors can affect the likelihood that a suicide attempt will occur. There are several suicide related factors that are associated with increased and decreased suicide attempts Demographics I n a study using a national adolescent sample in the United States, gender was identified as a major demographic risk factor for suicide ( Epstein & Spirito, 2010; Wolitzky Taylor, Rug giero, McCart, Smith, Hansen, Resnick, & Kilpatrick, 2010; Chung & Joung, 2012 ).

PAGE 10

! 2 In the United States, the rate of suicidal ideation and attempts is higher among females than for their male peers. Additionally, a ccording to research conducted by Pena and c olleagues suicide attempts are disproportionately female, B lack, and Hispanic (Pena, Matthieu, Zayas, Masyn, & Caine, 2012) They also found that of subpopulations, Hispanic females had higher rates of suicide attempts than other minority groups; w hereas, White males had lower r ates of suicide attempts than other subpopulations Additionally, research indicates that men die by suicide three and a half time more often than women (CDC, 2014). Since demographic factors such as gender and race /ethnicity are as sociated with increased suicide rates among adolescent youth, further analysis may indicate additional risk factors. Bullying Data from the 2011 National Youth Risk Behavior Survey (YRBS ) indicates over 20% of high school youth are bullied on school pr operty and 16% are victims of electro nic bullying or cyberbullying (CDC 2012 ) Victims of cyber bullying are more likely to report depressiv e symptoms than victims of other types of bullying (Wang, Nansel, & Iannotti, 2011). There is a large body of resea rch that supports the relationship between bullying and adolescent suicide (Klomek, Sourander, & Gould, 2010) This includes a wide range of bullying behaviors: being belittled about religion, race, looks, or speech; being the subject of lies, rumors, or s exual jokes or comments, and cyberbullying (Bonanno & Hymel, 2013; Hinduja & Patchin, 2010; Klomek et al., 2010; Undheim, 2013). Given the high prevalence of bullying and its strong association with adolescent suicide risk, it is necessary to further study this relationship. Substance Abuse Another risk factor linked with suicidal ideation and attempts among adolescents include substance use in the United States (Dunn, G oodrow, Givens, & Austin, 2008). Adolescent substance use is highly prevalent, particu larly among high school students (Johnston, O'Malley, Bachman, & Schulenberg, 2008) Research has identified that adolescents e xperiencing suicidal ideation are at a greater risk for substance abuse ( Department 2001; American Academy of Child and Adolesce nt Psych iatry, 2004)

PAGE 11

! 3 Studies have found associations between adolescent i llicit drug use and suicide although types of drugs are frequently not identified or compare d (Miller, 2011 ). Alcohol use has been associated with an increased risk for suicidal b ehaviors among adolescents (Roy, Lamparski, DeJong, Moore, & Linnoila, 1990; Borowsky, Ireland, & Resnick, 2001; Bossarte, & Swahn, 2011). However, the association o marijuana use and suicide has been inconsistent (Degenhardt, Hall, & Lynskey, 2003; Peders en, 2008; Price, Hemmingsson, Lewis, Zammit, & Allebeck, 2009; Rasic, Weerasinghe, Asbridge, & Langille, 2013; Chabrol, Melioli, & Goutaudier, 2014; Rylander, Valdez, & Nussbaum, 2014). A study by Rasic et al (2013) found and association between heavy mar ijuana use and depression but identified no association with suicidal ideation or attempts. Overall the use of alcohol and marijuana use have all been associated with suicidality (Kim, 2010; Roberts 2010) ; however extensive research into specific use and suicide attempts have not been consistent Media Exposure Currently, v ideo game s and i nternet use are common activi ties among children and adolescents (Lenhart & Madden, 2007; Marshall, Gorely, & Biddle, 2006) According to a study from Rideout and col leagues, approximately 83% of United States yo uth between 8 and 18 years ha ve video game consoles at home (Rideout, Roberts, & Foehr, 2005) According to this stu dy 86% have a home computer 35% have a computer in their bed room, and the average time of I n ternet use for recreation was estimated for over an hour. Additionally, compared to adolescents who report no video game or Internet use, adolescents reporting 5 or more hours of daily video game and Internet use were more likely to ha ve experienced suicid al idea tion or have made a suicidal plan ( Messias, Castro, Saini, Usman, & Peeples, 2011 ; Gauthier, Zuromski, Gitter, Witte, Cero, Gordon, & Joiner, 2014 ). Such results are consistent with other studies that found a relati onship between depression, video gaming, and suicide ( Kim, Namkoong, Ku, & Kim, 2008 ; Mathers, Canterford, Olds, Hesketh, Ridley, & Wake, 2009; Rehbein, Kleimann, & Mossle, 2010; Weaver, Mays, Sargent Weaver, Kannenberg, Hopkins Eroglu et al ., 2009) including Internet overuse (Lam & Pen g, 2010) A lt hough research has

PAGE 12

! 4 found associations between video game and internet use with suicide, there has been limited research that examines the relationship between number of hours spent watching television and suicide. Exercise and Sports Particip ation Research suggests that e xercise and sports participation might be protective factors in suicide attempts and suicide. Youth physical activity guidelines, according to the U.S. Department of Health and Human Services, recommend that children and adol escents aged 6 17 years should have 60 minutes or more of physical activity each day (2008). A number of studies with adolescents who had regular participation in athletic activity displayed lower reported instances of hopelessness, depression, suic idal id eation, and suicide (Babiss, & Gangwisch 2009; Brown, Galuska, Zhang, Eaton, Fulton, Lowry, & Maynard 2007; Miller & Hoffman, 2009 ; Taliaferro, Ri enzo, Miller, Pigg, & Dodd, 2009). Other studies also provided evidence of physical inactivity was a risk f actor for suicide For example, research indicates that individuals who attempted suicide were less physically active than those w ho did not attempt (Simon, Powell, & Swann, 2004 ; Chioqueta, & Stiles, 2007 ). Although many case studies have examined physica l activity as a protective factor of suicide in adolescents, further examination may clarify types of physical activity and its contribution to protective By contributing to research that further examines potential risk and protective factors in adolesc ent populations may assist in understanding and prevention efforts to combat adolescent suicide. Specific Aims Of adolescents who experience the distress of suicidal ideation, only a small p rop ortion of suicidal adolescents is recognized by adults, includ ing parents and school staff who may be the best prospects to help (Brown, Wyman, Brinales, & Gibbons, 2007). Additionally, the majority of suicidal adolescents do not seek adult support (Kerr, Owen, Pears, & Capaldi, 2008; Wyman, Brown, Inman, Cross, Schm eelk Cone, Guo, & Pena, 2008; Klaus, Mobilio, & King, 2009). Parent and school staff's inability to identify and detect adolescents at r isk of suicide is a major

PAGE 13

! 5 bar rier to reducing suicide deaths. Although an increasing number of school based suicide prev ention efforts are aimed at improving help seeking behavior from adolescents (Aseltine, James, Schilling, & Glanov sky, 2007; Freedenthal, 2010), there remains a gap in clear identification of adolescent suicide risk and protective factors The present stud y seeks to address this gap by investi gating the relationship between adolescent suicidal ideation a nd a variety of risk and protective factors for on a nationally representative sample of high school students in the United States Specifically, t his study examine d demographics (i.e., gender, race /ethnicity and grade), bullying (i.e., on school property and electronically), substance abuse (i.e., alcohol and marijuana) media exposure (i.e., hours per school day spent watching television and video game/Int ernet usage), as well as exercise and physical activity (i.e., days per week of physical activity and sports participation) This study utilized data from Youth Risk Behavior Surveillance System (YRBS; Brener et al., 2013). Data from a nationally represe ntative sample such as the YRBS can be used to determine the degree of the relation between demographics of risk and protective factors and suicide attempts to inform the efforts of early identification and intervention of youth at risk for suicide. The pu rpose of the current study was to assess how well the following varia bles predicted suicide attempts : gender, grade, race/ethnicity, bullied on school property and electronically, substance use alcohol and marijuana, media exposure television and video games/internet, and exercise and sports participation The following research questions were investigated: RQ1: Do the demographic characteristics of gender, grade, or race /ethnicity predict whether a student attempted suicide ? RQ2: Does the experience of being bullied on school property and electronically, substance use alcohol and marijuana, media exposure television and video games/internet, and exercise and sports participation that best predicts whether a student attempted suicide?

PAGE 14

! 6 RQ3 : Is there a c ombination gender, grade, race /ethnicity gender, experience of being bullied on school property and electronically, substance use alcohol and marijuana, media exposure television and video games/internet, and exercise and sports participation that best predicts whether a student attempted suicide? Knowing the answers to these questions is important for identifying students at risk for suicide attempts in order to intervene and save adolescent lives.

PAGE 15

! 7 CHAPTER II METHODS Sample The Youth Risk Behavi or Survey (YRBS) is a scale created by the Centers for Disease Control and Prevention (CDC) to assess the prevalence of adolescent behaviors that contribute to the leading causes of death, disability, and social problems in the United States. The survey co ntai ns about 98 self report items designed to measure the frequency and the severity of behaviors within six categories: (1) violent and self injurious behavior, (2) tobacco use, (3) alcohol and other drug use, (4) sexual behavior, (5) dietary behaviors, a nd (6) physical inactivity (CDC 2013). The YRBS utilizes a three stage cluster sample design and produce s a representative sam ple of 9th through 12th grade United States high school students aged 14 18 years. YRBS s urveys were distributed to all regular members of public, Catholic, and private school s between grades 9 and 12 across 50 states A weighting factor was applied to each student record to adjust for a nonresponse and the oversampling of students in the sample. According to the YRBS Combined Set s User's Guide, "overall weights were scaled so the weighted count of students was equal to the total sample size, and the weighted proportions of students in each grade matched population projections for each survey year." Data from the most recent (2013 ) Youth Risk Behavior Surveillance System was utilized for this study. The final sample for the 2013 YRBSS was collected from 148 of the 193 high schools sampled across the country (77% school response rate) Of the 15,480 students sampled, 13,633 submitte d questionnaires. Surveys that contained incomplete and missing data were omitted during the YRBS data scrub, which left 13,583 usable questionnaires Several studies have examined and determined the psychometric properties of the YRBS, indicating that the YRBS has sufficient levels of test retest reliability ( e.g., Brener et al. 1995, 2002, 2003).

PAGE 16

! 8 Measures Due to the infrequency of multiple occurrences of the independent variables under consideration and the strong positively skewed distributions (i.e., majority of the scores in the lower range of the scales), predictor variables that were ordinal in nature (i.e., 0 times/days, 1 time/day, 2 or 3 times/days, 4 or 5 times/days, 6 or more times/days) were dichotomized using dummy variables. This was also do ne to meet the condition of logistic regression that all predictors be either scale or dichotomous in nature. Other researchers have used similar procedures (e.g., Nickerson & Slater, 2009; Shilubane et al., 2012). Demographics Gender, grade level, year and race /ethnicity were provided by the data. In the 2013 survey gender included the categories of "Female" or "Male." Grade subscale was measured by providing response answers of "9 th ", "10 th ", "11 th ", "12 th and "u ngraded or other." For the purpose of this study the "ungraded or other" grade level category contained only 23 responses, and was omitted from the analysis. The r ace /ethnicity was broken down into two questions: Are you Hispanic or Latino?'' and What is your race?'' Although there was an extensive race/ethnicity category of eight levels the race /ethnicity category was modified for utilization for the present study. The race /ethnicity category was broken down from the previous eight categories and abbreviated into four: White "Black or African American Hispanic/Latino and "All other race/ethnicities." Of the previous eight race/ethnicity categories, White and Black or African American" remained the same, whereas "Hispanic/Latino" was combined with "Multiple Races Hispanic" and "A ll other race/ethnicities" included "American Indian/Alaska Native," "Asian," "Native Hawaiian /Other Pacific Islander," and "Multiple Race Non Hispanic." Bullying. Bullying was measured based on 2 items from the Youth Risk Behavior Survey (CDC, 2013) in order to measure how frequently adolescents were victims of bullying at school and electronically Bullying on school property was examined by the question, "During the past 12 months, have you ever been bullied on school property?" This question provided responses

PAGE 17

! 9 from "Yes" or "No." The electronic bullying measurement consisted of 1 item intended to measure how frequently adolescents were bullied by peers through electronic communication This item was also binary to responses of "Yes" or "No" to the qu estion, During the past 12 months, have you ever been electronically bullied? (Include being bullied through e mail, chat rooms, instant messaging, Websites, or texting) Substance Abuse. Alcohol and Marijuana abuse were measured based on 2 items from the Youth Risk Behavior Survey (CDC, 2013). Alcohol abuse was examined by the question, During the past 30 days, on how many days did you have 5 or more drinks of alcohol in a row, that is, within a couple of hours?" This quest ion provided responses from "0 days," "1 day," "2 days," "3 to 5 days," "6 to 9 days," "10 to 19 days," and "20 or more days." For this study, the alcohol abuse variable examined all responses of consuming 5 or more drinks of alcohol in the last 30 days, for "1 or more days" as compa red to "0 days." Marijuana use was analyzed similarly to alcohol abuse. Marijuana abuse was examined by the question, During the past 30 days, how many times did you use marijuana?" Question responses included, "0 times," "1 or 2 times," "3 to 9 times ," 10 to 19 times ," 20 to 39 times ," and 40 or more times Additionally, for this study, marijuana use was analyzed by separating responses into "0 times" and "1 or more times" of marijuana use in the last 30 days. This variable was analyzed t o compare tho se who are using marijuana 1 or more time to the en tire sample of questionnaires. This variable was modified to display marijuana use, the dummy variable of "1" was used for "0 days" and "2" was used for" 1 or more days." Media Exposure. In order to asses s media exposure, variables that measured hours of watching television and video game/internet use were examined. This YRBS item asked On an average school day, how many hours do you watch TV? Item responses included, "I do not watch TV on an average s chool day ," Less than 1 hour per day ," 1 hour per day ," 2 hours per day ," 3 hours per day ," 4 hours per day ," and 5 or more hours per day ." For the purpose of the study, only responses of 3 or more hours of television on an average school day were ex amined.

PAGE 18

! 10 Additionally, average hours per school day and video game/internet use were examined. The YRBS item question inquired, On an average school day, how many hours do you play video or computer games or use a computer for something that is not schoolw ork? (Count time spent on things such as Xbox, PlayStation, an iPod, an iPad or other tablet, a smartphone, YouTube, Facebook or other social networking tools, and the Internet.) Respondents were given the following responses to choose from, I do not pla y video or computer games or use a computer for something that is not school work ," Less than 1 hour per day ," 1 hour per day ," 2 hours per day ," 3 hours per day ," 4 hours per day ," and 5 or more hours per day ." In order to examine excess use of vide o game/internet, only responses of 3 or more hours on an a verage school day were examined, specifically the number of hours per average school day was coded as "1" for less than "3 hours" and more than "3 hours" was coded as "2" Exercise and Sports Partic ipation. Examining adolescent inactivity or frequency in exercise and sports participation, two survey items were analyzed. Regular physical activity was measured by responses of "0 days," "1 day," "2 days," "3 days," "4 days," "5 days," "6 days," and "7 d ays" to the survey item, During the past 7 days, on how many days were you physically active for a total of at least 60 minutes per day? (Add up all the time you spent in any kind of physical activity that increased your heart rate and made yo u breathe ha rd some of the time. ) The p hysical activity item was analyzed based on self response to questions of 5 days or more. This variable was modified with the dummy code "1" for "less than 5 days" and "2" was used for "more than 5 days" of physical activity. S ports participation was measured by self report to the YRBS item, During the past 12 months, on how many sports teams did you play? (Count any teams run by your school or community groups.) Responders were provided responses of 0 teams ," 1 team ," 2 t eams ," and 3 or more teams ." Sports participation is examined by responses of one or more sports teams in the past 12 months. Similarly, for number of sports teams, "1" was used for being on "0 teams" and "2" was used for "1 or more teams"

PAGE 19

! 11 Suicide. The suicidal behavior subscale contained four items from the YRBS although items were only measured based on self reported suicide attempts. This item asked adolescents to respond to items of "0 times "1 time," "2 or 3 times," "4 or 5 times," "6 or more tim es" measuring suicide attempts (e.g., During the past 12 months, how many times did you actually attempt suicide?''). From this question, a binary categorical variable was created: attempted suicide as "yes" which indicated a response of "one or more ti mes" and "no" which indicated a response of "0 times Table 1a and Table 1b present the n for all variables included in the analysis. Table 1a Sample Characteristics of the Population ( N =13,583) Characteristics n (%) Gender Female 6,621 48.8 Male 6,950 51.2 Grade Level 9th 3,588 26.6 10 th 3,152 23.4 11 th 3,184 23.6 12 th 3,557 26.4 Ethnicity White 5,449 41.1 Black or African American 2,993 22.6 Hispanic/Latino 3,395 25.6 All Other 1,428 10.8 Suicide Attempt in past 12 months No 10,967 91.5 Yes 1,015 8.5

PAGE 20

! 12 Table 1b Sample Characteristics of Risk and Protective Factors ( N =13,583) Variable n (%) Bullied Bullied on School Property Yes 2,508 18.6 No 11,007 81.4 Bullied Electronically Yes 1,878 13.9 No 11,623 86.1 Substance Use Alcohol Abuse (5+ drinks in a row in past 30 days) 0 days 10,440 79.9 1+ day 2,620 20.1 Marijuana Abuse (Use i n last 30 days) 0 times 9,947 74.7 1+ times 3,367 25.3 Media Exposure TV (Average hours/school day) Less than 3 8,463 63.9 More than 3 4,782 36.1 Video Game/Internet Use (Average hours/school day) Less than 3 7,586 57.2 More than 3 5,669 4 2.8 Exercise and Sports Participation Physical Activity (60 min/day) 0 4 days 7,120 53.5 5+ days 6,190 46.5 Sports Team Participation (Previous 12 months) 0 teams 6,151 46.6 1+ teams 7,044 53.4 Data Analysis A nalyses were conducted using SP SS Version 23 YRBS data provides weights based on gender, race/ethnicity, and grade level. These weights were applied to provide representative data to adolescents in the United States. Descriptive statistics were used to describe the youth demographics a nd their risk and protective factors Logistic regression analyses were used to describe the relationship(s) among risk and protective factors. The logistic regressions were performed to assess how well the following variables predicted suicide attempts: gender, grade, race/ethnicity, bullied on school property and

PAGE 21

! 13 electronically, substance use alcohol and marijuana, media exposure television and video games/internet, and exercise and sports participation. Prevalence of student suicide attempt s w as ca lculated based on the three demographic variables: gender, grade and race/ethnicity SPSS Crosstabs function was applied to each risk characteristic with suicide attempt with chi square analyses were utilized. Pearson chi square was conducted to determine strength of association between risk and protective characteristics and suicide attempts within demographics Additionally, pairwise comparisons were made for each variable and Cramer's V and Phi were used to examine effect size and strength of associatio ns. Column proportions were analyzed for significant differences at the .05 level to determine differences within risk and protective factors by demographics. These variables were included in the analysis based on what is known about risk factors for suici dal behavior and to determine specific risk and protective factors for different groups. Assumptions Assumptions of logistic regression were checked and met, including observations being independent and in dependent variables be ing linearly related to th e logit of the dependent variable. Odds ratios (OR) were checked to assess the ratio of the probability of the occurrence of the outcome variable based on the predictor variable. Pearson chi square is used when your research question includes two or more n ominal or ordinal variables. The assumptions of (1) having a large sample size and (2) having at least 5 for expected frequencies in each cell were checked and met.

PAGE 22

! 14 CHAPTER II I RESULTS Demographics as Predictors of Suicide Attempt A logistic regressi on was conducted to answer the first research question: Do the demographic characteristics of gender, grade, or race/ethnicity predict whether a student attempted suicide? When all predictor variables were considered together, they significantly distinguis hed between demographic characteristics of students who attempted suicide in the previous 12 months 2 = 78.76 df = 1 n = 11,689 p < .001. This model predicted 4% of the variance in the outcome variable, Nagelkerke R 2 = .03 7. Table 2 presents the demographic characteristic predictors for students who attempted suicide Wald 2 OR, and 95% confidenc e intervals (CI) for the ORs. Gender, race/ ethnicity and grade level were statis tically significant predictors for suicide attempt, specific predictors included females ( = # .77, Wald 2 (1) = 120.95 p <.001, OR = 2.17 ) White ( = .36 Wald 2 (1) = 10 .38 p <.001, OR = .61 ) Black/African American ( = .11 Wald 2 (1) = .88 p <.001, OR = .71 ), Hispanic/Latino ( = # .13 Wald 2 (1) = 1.37 p =. 1 0, OR = .84 ), 9 th grade ( = # .50 Wald 2 (1) = 28.60 p <.001, OR = .61 ), and 10 th grade ( = # .35 Wald 2 (1) = 12.26 p <.001, OR = .71 ) Although 12 th grade and all other race/ethnicity were not found significant these factors were omitted due to redundancy This finding is likely related to the differences in the sample sizes for the all other race/et hnic group and 12 th grade, leading to likely insufficient statistical power to detect significance for these two groups. Results of these analyses are presented in Table 2.

PAGE 23

! 15 Risk and Protective Factors as Predictors of Suicide Attempt A logistic regression w as conducted to answer the second research questio n for students who attempted suicide in the past 12 months : Does the experience of being bullied on school property and electronically, substance use alcohol and marijuana, media exposure television and video games/internet, and exercise and sports parti cipation that best predicts whether a student attempted suicide? Each risk and protective factor was analyzed by variable group, therefore by running four logistic regressions by variable group Table 3 presents the risk and protective predictors for stude nts who attempted suicide, Wald 2 OR, and 95% confidence intervals (CI) for the ORs. When predictor variables of bullying on school property and electronically were considered together, they significantly predicted whet her or a student attempte d suicide 2 = 561.19, df = 2 N = 11,945 p < .001. This model predicted 10 % of the variance in the outco me variable, Nagelkerke R2 = .10 ( Table 3 ). Table 3 presents odds ratios, which suggests being Table 2 Logistic Regression Results for Demographic Predictors of Gender, Race, a nd Grade Level: Suicide Attempt Variable B Wald Sig. OR 95% CI for OR Gender .77 120.27 <.001 2.17 1.89 2.49 Female .77 120.95 <.001 .46 .40 .53 Male 0 b . Race/Ethnicity .17 29.08 <.001 .85 .80 .90 White .36 10.38 <.001 .61 .50 .73 Blac k/African American .11 .88 <.001 .71 .58 .86 Hispanic/Latino .13 1.37 .10 .84 .69 1.03 All other 0 b . Grade Level .17 32.57 <.001 1.18 1.12 1.26 9th .50 28.60 <.001 .61 .50 .73 10th .35 12.26 <.001 .71 .58 .89 11th .17 2.73 .10 .84 .69 1 .03 12th 0 b . Note. a The reference category is: Yes. b This parameter is set to zero because it is redundant.

PAGE 24

! 16 bullied on school property ( = # .89 Wald 2 (1) = 123.29 p < .001, OR = .41 ), and electronically ( = # 1.14 Wald 2 (1) = 191.86 p <.001, OR = .32 ) were predictors of suicide attempt. When substance use variables of alcohol and marijuana were considered together, they signifi cantly predicted whether a student att empted suicide, 2 = 250.13, df = 2, N = 11,500 p < .001. This model predicted 5% of the variance in the outcome variable, Nagelkerke R2 = .05 (Table 3 ). Odds ratios, which suggests abusing alcohol "0 times" in the past 30 days ( = .55 Wald 2 (1) = 43. 82 p <.001, OR = 1.73 ), and using marijuana "0 times" ( = .82 Wald 2 (1) = 105.68 p <.001, OR = 2.26 ) were predictors of suicide attempt. When media exposure variables of television and video games/internet use were considered together, they signifi cantly predicted whether or not a student attempte d suicide, 2 = 50.18 df = 2, N = 11,708 p < .001. This model predicted 1 % of the variance in the outco me variable, Nagelkerke R2 = .01 (Table 3 ). Odds ratios, which suggests wat ching TV "less than three hours on an average school day ( = .55, Wald 2 (1) = 43.82, p <.001, OR = 1.73) and spending on average "less than three hours playing video games or internet on a school day ( = .19 Wald 2 (1) = 7.48 p =.006, OR = 1.21 ) were predictive of a stud ent suicide attempt. Additionally, when exercise and sports participation were considered together, they signifi cantly predicted whether a student attempted suicide, 2 = 36.66 df = 2, N = 11,654, p < .001. This model predicted 1% of the variance in the outcome variabl e, Nagelkerke R2 = .007 (Table 3 ). Odds ratios, which suggests being physically active for 60 minutes for an average of "0 4 days per week" ( = .34 Wald 2 (1) = 22.16 p <.001, OR = .71 ) were also predictive of a student suicide attemp t, although sports participation was not. Some of the significant findings were similar to the overall population's report of less substance abuse, media exposure, and video game/internet use potentially creating a higher statistical power to reported area s (See Table 3 )

PAGE 25

! 17 Demographics with Risk and Protective Factors Predicting Suicide Attempt Pearson chi square was conducted to answer the third research question: Is there a co mbination gender, grade, race/ethnicity, gender, experience of being bullied on school property and electronically, substance use alcohol and marijuana, media exposure ( television and video Table 3 Logistic Regression Results for Predictors of Risk and Protective Factors: Suicide Attempt Variable B Wald Sig. OR 95% CI for OR Bullying On School Property .89* 123.29 <.001 2.42 2.07 2.83 Yes .89 123.29 <.001 .41 .35 48 No 0 b . Electronically 1.14* 191.86 <.001 3.12 2.66 3.66 Yes 1.14 191.86 <.001 .32 .27 .38 No 0 b . Substance Use Alcohol .55 43.82 <.001 .58 .49 .68 0 days .55 43.82 <.001 1.73 1.47 2.03 1+ day 0 b . Marijuana .8 2 105.68 <.001 .44 .38 .52 0 times .82 105.68 <.001 2.26 1.94 2.64 1+ time 0 b . Media Exposure TV Use (Average hours/school day) .19 7.48 .006 .83 .72 .95 Less than 3 .41 35.50 <.001 1.50 1.31 1.71 More than 3 0 b . Video Game/ Internet Use (Average hours/school day) .41 35.50 <.001 .67 .58 .76 Less than 3 .19 7.48 .006 1.21 1.06 1.38 More than 3 0 b . Exercise and Sports Participation Physical Activity (60 min/day) .34 22.16 <.001 1.40 1.22 1.61 0 4 days .34 22.16 <.001 .71 .62 .82 5+ days 0 b . Sports Team Participation (Previous 12 months) .15 4.56 .033 1.16 1.01 1.33 0 teams .15 4.56 .033 .86 .75 .99 1+ teams 0 b . Note a The reference category is: Yes. b This parameter is set to zero b ecause it is redundant. *Did not meet assumptions of logistical regression when run as factors

PAGE 26

! 18 games/internet ) and exercise and sports participation that best predicts whether a student attempted suicide? Each risk and protective factor was analyzed by variable group with demographics ; therefore four chi squares were conducted with z test of column portion significant differences were applied to examine data. A ssumptions were checked and met. Table s 4 7 presents the risk and protective predictors combined with student demographics who attempted suicide, with Cramer's V, Phi, and column z test. Bullying. To investigate whether student demographics and bullying o n school property were related to student suicide attempt, a chi square was used (Table 4 ) Assumptions were checked and met Pearson ch i square results in dicate that suicide attempt, gender and being bullied on school property are significantly related t o one another ( 2 = 464.55 df = 1 N = 11 ,953 ), p <. 001 ). Cramer's V which indicates the strength of the association be tween the two variables, was .2 0 and, thus, the eff ect size is considered moderate The Pearson chi square results in dicate that suicide attempt, ra ce/ethnicity and being bullied on school property were also significantly related to one another ( 2 = 565.39 df = 1, N = 11 ,702 ), p <. 001). Cramer's V which indicates the strength of the association between the two variables, was .2 2 and, thus, the ef fect size is considered moderate. Lastly the Pearson chi square results also ind icated that suicide attempt, grade level and school property bullying were also significantly related to one another ( 2 = 576.649, df = 1, N = 11,869), p <. 001). Cramer's V which indicates the strength of the association between the two variables, was .22 and, thus, the effect size is considered moderate.

PAGE 27

! 19 Table 4 Chi Square Results for Predictors of Gender, Race, and Grade Level by Bullying Factors: Suicide Attempt Variable Bullied on School Property Total Bullied Electronically Total Yes No Yes No Gender Female 322 a 346 b 668 295 a 373 b 668 Male 128 a 211 b 339 107 a 232 b 339 Total 450 a 557 b 1007 402 a 605 b 1007 Race/Ethnicity White 194 a 150 b 344 187 a 158 b 3 45 Black/African American 65 a 136 b 201 41 a 159 b 200 Hispanic/Latino 124 a 191 b 315 117 a 198 b 315 All other 59 a 64 b 123 45 a 78 b 123 Total 442 a 541 b 983 390 a 593 b 983 Grade Level 9th 172 a 152 b 324 128 a 196 b 324 10th 118 a 131 b 249 103 a 147 b 250 11th 83 a 130 b 213 76 a 135 b 211 12th 72 a 138 b 210 88 a 123 b 211 Total 445 a 551 b 996 395 a 601 b 996 Note. Each subscript letter denotes a subset of Bully categories whose column proportions do not differ significantly from each other at the .05 level. The c olumn proportions test table assigns a subscript letter to the categories of the column variable. For each pair of columns, the column proportions are compared using a z test. If a pair of values is significantly different, the values have different subscr ipt letters assigned to them. To further investigate the relationship a z tes t as applied to column proportions of students who attempted suicide and were bullied on school property Potential differences based on gender, race/ethnicity, and grade leve l of students who attempted suicide and were bullied on school property were also examined Of all students who reported attempting suicide in the previous year, there were significantly more who were not bullied on school property ( n= 557 p =.05) than stud ents who were bullied on school property ( n= 450 p =.05 ; Table 4) Of female students who had attempted suicide in the previous year there w ere si gnificantly more who were not bullied on school property ( n= 346 p =.05) than female students who were bullied on school property ( n= 346 p =.05). Additionally, of male students who had attempted suicide in the

PAGE 28

! 20 previous year there was significantly more who were not bullied on school property ( n= 211 p =.05) than males students who were bullied on school property ( n = 128 p =.05). When considering race/ethnicity, being bullied on school property, and suicide attempt, z tests indicated that of White students who attempted suicide, there were significantly more who were bullied on school property ( n= 194 p =.05) than t hose who were not ( n= 150 p =.05). Although the race/ethnicity groups of Black/African American, Hispanic/Latino, and all other races did indicate a significant difference between column proportions of those who were bullied on school property and those who were not, these race/ethnicity groups of students reflected significantly more students were not bullied on school property. When considering grade level, being bullied on school property, and suicide attempt, z tests indicated that among 9 th grade studen ts who attempted suicide, there were significantly more who were bullied on school property ( n= 172 p =.05) than those who were not ( n= 152 p =.05). Whereas of 10 th 11 th and 12 th gr ade students who reported a previous suicide attempt they reported signif icantly less bullying on school property (Table 4). To investigate whether student demographics and electronic bullying were related to student suicide attempt a chi square was used (Table 4 ). Assumptions were checked and met. Pearson chi square results indi cate that suicide attempt, gender and electronic bullying are significantly related to one another ( 2 = 592.68 df = 1, N = 11 ,947 ), p <. 001). Cramer's V which indicates the strength of the association between the two variables, was .2 2 and, thus, the effect size is considered moderate. S uicide attempt, race/ethnicity and electronic bullying were also significantly related to one another ( 2 = 565.39, df = 1, N = 11,702), p <. 001). Cramer's V which indicates the strength of the association betwee n the two variables, was .22 and, thus, the effect size is considered moderate. Lastly Pearson ch i square results indicated that suicide attempt grade level and electronic bullying were also significantly related to one another ( 2 = 576.649, df = 1, N = 11,869), p <. 001). Cramer's V which indicates the strength of the association between the two variables, was .22 and, thus, the effect size is considered moderate.

PAGE 29

! 21 The relationship between student demographics and electronic bullying with suicide attem pt was further investigated by applying a z test to column proportions of students who attempted suicide. This was done in order to examine if there was a signifi cant difference between demographic characteristics of students who attempted suicide and were bullied electronically and those who were not. Of all students who reported attempting suicide in the previous year (results from gender and electronic bullying chi square), there were significantly more who wer e not bullied electronically ( n= 605 p =.05) than students who were bullied electronically ( n= 402 p =.05). Of female students who had attempted suicide in the previous year there was significantly more who were not bullied electronically ( n= 295 p =.05) than femal es students who were bullied electroni cally ( n= 373 p =.05). Additionally of male students who had attempted suicide in the previous year there were also significantly more who were not bullied electronically ( n= 232 p =.05) than males students who were bullied electronically ( n= 107 p =.05). A lthough there were significantly fewer reports of being bullied electronically among students who had attempted suicide in the previous year z tests indicated some differences by race/ethnicity Similarly to the on school property bullying results from the z test, White students who had previously attempted suicide had significantly higher proportions of students who had been bullied electronically ( n= 187 p =.05) than those who were not ( n= 158 p =.05). Compared to other race/ethnicity categories, only th e White students reported higher proportions of electronic bullying. Additionally, students who had previously attempted suicide and were bullied electronically and were analyzed by grade level. Z test results from co lumn comparisons indicated that stud ents from all grade levels reported significantly less proportions of electronic bullying ( p =.05) Overall each grade level was representative of the overall sample of students who had attempted suicide, where there were significantly higher reports of not being bullied electronically ( n= 601) than those who were ( n= 395 p =.05)

PAGE 30

! 22 Substance Use. To investigate wheth er student demographics and substance use were related to student suicide attempt, a chi square statistic was used to compare alcohol use and mar ijuana use with student demographic characteristics (Table 5) Assumptions for alcohol use chi squares were checked and met. Pearson chi square results indicate that suicide attempt, gender, and alcohol use are significantly related to one another ( 2 = 17 0.98 df = 1 N = 11,599 p <. 05). The Pearson chi square results for suicide attempt, race/ethnicity, and alcohol use are significantly related to one another ( 2 = 168.86 df = 1, N = 11,365 p <. 05). Lastly, chi square results for suicide attempt, grad e level, and alcohol use are significantly related to one another ( 2 = 165.89 df = 1, N = 11,527, p <. 05). Cramer's V which indicates the strength of the association between the two variables was .12 for each chi square and, thus the effect size is cons idered very weak. Table 5 Chi Square Results for Predictors of Gender, Race, and Grade Level by Substance Use Factors: Suicide Attempt Variable Alcohol Total Marijuana Total 0 day 1+ day 0 times 1+ times Gender Female 409 a 225 b 634 37 6 a 277 b 653 Male 187 a 121 b 308 147 a 163 b 310 Total 596 a 346 b 942 523 a 440 b 963 Race/Ethnicity White 201 a 126 b 327 179 a 158 b 337 Black/African American 138 a 49 b 187 100 a 93 b 193 Hispanic/Latino 158 a 130 b 288 157 a 143 b 300 All other 83 a 35 b 118 73 a 38 b 111 Total 580 a 340 b 920 509 a 432 b 941 Grade Level 9th 211 a 96 b 307 185 a 127 b 312 10th 152 a 87 b 239 132 a 109 b 241 11th 127 a 72 b 199 106 a 94 b 200 12th 102 a 86 b 188 95 a 105 b 200 Total 592 a 341 b 933 518 a 435 b 9 53 Note. Each subscript letter denotes a subset of Substance Use categories whose column proportions do not differ significantly from each other at the .05 level. The column proportions test table assigns a subscript letter to the categories of the colum n variable. For each pair of columns, the column proportions are compared using a z test. If a pair of values is significantly different, the values have different subscript letters assigned to them.

PAGE 31

! 23 The association between student demographics and mari juana use were also investigated through the Pearson chi square statistic. Assumptions were checked and met. Pearson chi square results indicate that suicide attempt, gende r, and marijuana use are significantly related to one another ( 2 =246.37, df = 1, N = 11,790, p <. 05) and the effect size is considered weak (Cramer's V = .15). Chi square results for suicide attempt, race/ethnicity, and marijuana use were also significantly related to one another ( 2 =243.46, df = 1, N = 11,548, p <. 05) although also considered a weak effect size ( Cramer's V = .15 ). Lastly, chi square results for suicide attempt, grade level, and marijuana use are significantly related to one another ( 2 =243.77, df = 1, N = 11,714, p <. 05) and the effect size is considered very weak (Cramer's V =.14). Although each association was considered weak or very weak, further investigations of gender, race/ethnicity, and grade level of alcohol and marijuana use were a nalyzed for significant differences between column pro portion z tests. Results from the z tests indicated that both female and male students who had attempted suicide had significantly less alcohol use of "0 days" drinking 5 or more drinks in the past month (female n= 409, male n= 187) than those who reported they had "1 or more days" (female n= 225, male n= 121, p <. 05) Although gender was not a factor indicator of higher reports of prior suicide attempt and alcohol use, gender was a factor for significantly higher reports of marijuana use and previous suicid e attempt. C olumn proportions for female students who attempted suicide and used marijuana in the prior month indicated significantly less marijuana use ( n= 376 used "0 times," n= 277 used "1+ time"), while their male counterparts indicated significantly hi gher marijuana use in the past month ( n= 147 used "0 times," n= 163 used "1+ time", p <. 05). When comparing race/ethnicity characteristics with alcohol use and previous suicide attempt, as presented in Table 5, indicated significant reports of "0 days" o f consuming 5 or more drinking in the past month. This was also the case across grade levels, where all students who reported attempting suicide had sig nificantly higher reports of "0 days" of using alcohol in excess during the previous month. Additionally marijuana use was significantly lower across all

PAGE 32

! 24 race/ethnicities of students who had atte mpte d suicide in the previous year (Table 5). However, when considering grade level marijuana use in the previous month, 12 th grade students who had attempted suici de in the previous year also reported significantly higher marijuana use in the past month ( n= 105) than those who reported they had not ( n= 95, p <. 05 ); w hereas Table 5 presents that all other grade levels had significantly higher reports for using marij uana "0 times" over the past month. Media Exposure. Pearson chi square statistic was used to compare student demographics and media exposure as related to student suicide attempt (Table 6 ). Assumptions for television use chi squares were checked and met Pearson chi square results indicate that suic ide attempt, gender, and television use are significantly related to one another ( 2 = 14.18 df = 1, N = 11,729 p < 05 ) although Cramer's V was .04 indicating the effect size is considered very weak. The Pearson chi square results for suicide atte mpt, race/ethnicity, and television use are significantly related to one another ( 2 = 12 .81 df = 1, N = 11,490 p <. 05) although Cramer's V was .03 displaying another very weak effect size Lastly, chi square results for suicide attempt, grade level, and television use are significantly related to one another ( 2 = 12.432 df = 1, N = 11,65 2 p <. 05). Cramer's V which indicates the strength of the association bet ween the two variables, was also .03 and, thus, the effect size is considered very weak.

PAGE 33

! 25 Table 6 Chi Square Results for Predictors of Gender, Race, and Grade Level by Media Exposur e Factors: Suicide Attempt Variable TV Total Video Game/Internet Total Less than 3 More than 3 Less than 3 More than 3 Gender Female 407 a 254 b 661 313 a 349 b 662 Male 180 a 143 b 323 152 a 173 b 325 Total 587 a 397 b 984 465 a 522 b 987 Race/Ethnicity White 245 a 93 a 338 174 a 165 b 339 Black/African American 81 a 114 a 195 93 a 102 a 195 Hispanic/Latino 171 a 135 b 306 136 a 171 b 307 All other 77 a 43 a 120 48 a 72 b 120 Total 574 a 385 b 959 451 a 510 b 961 Grade Level 9 th 194 a 123 a 317 143 a 176 b 319 10th 142 a 99 b 241 117 a 124 b 241 11th 133 a 74 a 207 103 a 104 b 207 12th 114 a 93 b 207 98 a 109 b 207 Total 583 a 389 b 972 461 a 513 b 974 Note. Each subscript letter denotes a subset of Media Exposure categories whose c olumn proportions do not differ significantly from each other at the .05 level. The column proportions test table assigns a subscript letter to the categories of the column variable. For each pair of columns, the column proportions are compared using a z t est. If a pair of values is significantly different, the values have different subscript letters assigned to them. The relation ship between student demographics and video game/internet use w as also investigated through the Pearson chi square statistic. Assumptions were checked and met. Pearson chi square results indicate that suicide attempt, gender, and video game/internet use are significantly related to one another ( 2 =43.39 df = 1, N = 11,731 p <. 05), and the effect size is considered very weak (Cramer's V = .06 ). Chi square results for suicide attempt, race/ethnicity, and video game/internet use were also significantly related to one another ( 2 =43.97 df = 1, N = 11,492 p <. 05), although also considered a very weak effect size (Cramer's V = .0 6 ). Lastly, chi square results for suicide attempt, grade level, and video game/internet use are significantly

PAGE 34

! 26 related to one another ( 2 =41.00 df = 1, N = 11,654 p <. 05), and the effect size is considered very weak (Cramer's V =.06 ). Although media ex posure effect sizes were considered weak to very weak, further investigations of gender, race/ethnicity, and grade level of television and video game/internet use were analyzed for significant differences between column proportion z tests. Results from the z tests indicated both female and male studen ts who had attempted suicide reported television use of "less than 3 hours on an average school day (female n= 407 male n= 180 ) which was significantly less than those who reported they watched more than 3 ho urs of television on an average school day (female n= 254 male n= 143 p <. 05). Average television use on a school night of students who had attempted suicide produced similar results for race/ethnicity and grade levels. Overall, there were significantly higher reports of "less than 3 hours" on an average school night across all grade levels and race/ethnicities, except for Black/African American and all other races categories (Table 6). For Black/African American and a ll other races, results indicated th at for students who had attempted suicide in the prior year there was no significant difference between those who watched less than or more than three hours of television on an average school night. As for video game/internet use, z tests indicated that overall students who had previously attempted suicide reported significantly higher video game/internet use (per gender and video game/internet use chi square statistics, Table 6). Gender and video game/internet use indicated that b oth female and male stu dents who had attempted suicide reported video game/internet use of "more than 3 hours" on an average school day (female n= 349, male n= 173 ) was significantly higher than those who reported they watched "less than 3 hours" of television on an average school day (female n= 313, male n= 152, p <. 05). Grade level comparisons indicated similar findings, where there were significantly higher findings for students who attempted suicide in the past year reported and "more than 3 hours" of video game/internet use on an average school night. When comparing video game/internet use across race/ethnicities there was no significant

PAGE 35

! 27 difference between video game/internet use for Black/African American students who had previously attempted su i cide. Presented on Table 6, Whi te students who had previously attemp ted suicide reported significantly less video game/internet use on an average school night ( n= 174) of "less than 3 hours which is different than the collective sample. Whereas Hispanic/Latino and all other race/ethnic ity categories indicated significantly higher video game/internet use of "more than 3 hours" on an average school night (Hispanic/Latino n= 171, All other race/ethnicities n=72, p <. 05 ) Exercise and Sports Participation. Pearson chi square statistic was used to compare student demographics with exercise and sports participation as related to student suicide attempt (Table 7 ) Assumptions for physical activity chi squares were checked and met. Pearson chi square results indicate that suicide attempt, gende r, and physical activity are significantly related to one another ( 2 = 33.05 df = 1, N = 11,777 p <. 05), although Cramer's V was .05 indicating the effect size is considered very weak. The Pearson chi square results for suicide attempt race/ethnicity, and physical activity are significantly related to one another ( 2 = 37.26 df = 1, N = 11,534 p <. 05), although Cramer's V was .06 displaying another very weak effect size Lastly, chi square results for suicide attempt, grade level, and physical activit y are significantly related to one another ( 2 = 34.45 df = 1, N = 11,700 p <. 05). Cramer's V which indicates the strength of the association between the two variables, was also .05 and, thus, the effect size is considered very weak. The association be tween student demographics and sports participation w as also investigated through the Pearson chi square statistic. Assumptions were checked and met. Pearson chi square results indicate that suicide attempt, gender, and sports participation use are signif icantly related to one another ( 2 =43.39, df = 1, N = 11,731, p <. 05), and the effect size is considered very weak (Cramer's V = .04 ). Chi square results for suicide attempt, race/ethnicity, and sports participation were also significantly related to one another ( 2 =14.79 df = 1, N = 11,428 p <. 05), although also considered a very weak effect size (Cramer's V = .0 4 ). Lastly, chi

PAGE 36

! 28 square results for suicide attempt, grade level, and sports participation are significantly related to one another ( 2 =14.59 df = 1, N = 11,592 p <. 05), a nd the effect size is considered very weak (Cramer's V =.06). Table 7 Chi Square Results for Predictors of Gender, Race, and Grade Level by Exercise and Sports Participation Factors: Suicide Attempt Variable Physical Activity Total Sports Team Participat ion Total 0 4 days 5+ days 0 teams 1+ teams Gender Female 438 a 226 a 664 362 a 296 a 658 Male 176 a 153 b 329 152 a 177 b 329 Total 614 a 379 b 993 514 a 473 b 987 Race/Ethnicity White 203 a 141 b 344 187 a 154 b 341 Black/African A merican 130 a 65 b 195 95 a 100 a 195 Hispanic/Latino 193 a 115 b 308 160 a 145 a 305 All other 80 a 40 b 120 61 a 59 a 120 Total 606 a 361 b 967 503 a 458 b 961 Grade Level 9th 194 a 125 b 319 157 a 161 b 318 10th 154 a 89 b 243 124 a 118 b 242 11th 1 31 a 78 b 209 108 a 100 a 208 12th 130 a 80 a 210 120 a 88 a 208 Total 609 a 372 b 981 509 a 467 b 976 Each subscript letter denotes a subset of Exercise and Sports Participation categories whose column proportions do not differ significantly from each other at the .05 level. The column proportions test table assigns a subscript letter to the categories of the column variable. For each pair of columns, the column proportions are compared using a z test. If a pair of values is significantly different, the value s have different subscript letters assigned to them. Although exercise and sports participation effect sizes were considered weak to very weak, further investigations of gender, race/ethnicity, and grade level of physical activity and sports team partic ipation were analyzed for significant differences between column proportion z tests (Table 7) Results from the z tests per gender and physical activity chi square, indicated that of students who had attempted suicide in the previous year were reporting 0 to 4 days" of at least 60 minutes of physical activity in the previous week ( n= 614) that was significantly higher than those who were physically active for "5 or more days" ( n= 379, p <. 05). Overall suicide attempters reported similarly across gender, ra ce/ethnicity, and grade level. Across demographic

PAGE 37

! 29 categories, all re s ponses displayed significantly higher scores for "0 to 4 days" of at least 60 minutes of physical activity in the previous week (Table 7). As for sports participation per gender and sp orts participation chi square z tests indicated that overall students whom attempted suicide in the previous year had significantly higher reports of participation in "0 sports teams" during the previous year ( n= 514 "0 teams, n= 473 "1 or more", p <. 05). However, attempting suicide and participating on "0 teams" over the previous year were not consistently significant across demographics. When comparing gender and sports participation, there was no significant difference between female students who had pr eviously attempted suicide and sports participation ( n= 362 "0 teams, n= 296 "1 or more", p <. 05). Additionally, male students who had attempted suicide in the previous year indicated significantly more sports participation ( n= 152 "0 teams, n= 177 "1 or more ", p <. 05) Race and ethnicity identifiers also indicated differences. For White students who had attempted suicide, the majority reported no sports particip ation during the previous year ( n= 187 "0 teams, n= 154 "1 or more", p <. 05). However, Black/Africa n American, Hispanic/Latino, and all other race/ethnicity categories of students whom previously attempted suicide indicated no significant difference s between sports participation (Table 7). Additionally, f or 9 th grade students, significantly more suicide attempters reported participation in sports over the previous year ( n= 157 "0 teams, n= 161"1 or more", p <. 05). Tenth grade students were similar to the overall sample of student attempters, where there was significantly more 10 th grade attempters who did not participate on a sports team during the previous year ( n= 124 "0 teams, n= 118 "1 or more", p <. 05). However, 11 th and 12 th grade student attempters indicated no significant difference s between sports participation (Table 7).

PAGE 38

! 30 CHAPTER IV DISCUSSION The purpose of the current study was to assess how well the following variables predicted suicide attempts: gender, grade, race/ethnicity, bullied on school property and electronically, substance use alcohol and marijuana, media exposure television and vid eo games/internet, and exercise and sports participation. Examination of these particular variables was based on what is currently known about risk and protective factors of suicidal high school students in order to identify specific supports and behaviora l risks for students among different groups. The results of this study reveal that the relationship between suicidal students and known risk and protective factors is complex. An analysis of data from a nationally representative sample of high school stude nts suggests that across demographic characteristics of suicidal students risk and protective factors vary on being predictive of a suicide attempt. Interpretation Previous research indicates that student demographic variables of gender and race/ethnicit y were predictors of suicide attempt; specifically for students who are female, Black/African American, and/or Hispanic/Latino (Pena et. al, 2012). Findings from this research study supported previous findings that female, Black/African American, and/or Hi spanic/Latino students were demographic risk factors for predicting suicide attempt Results of the present study additionally found that students who were White were also at a higher risk for suicide Additionally, this study fo und that students who were in the 9 th or 10 th grade were also predictive factors for suicide. Bullying was previous ly established as a risk factor for adolescent suicide (Klomek e t. al, 2010). Findings from this study supported previous research in which both being bullied on scho ol property and electronically were significant predictors of student suicide attempt. Across gender, bullying was a moderate predictor of attempting suicide. Student s who attempted suicide who were White or in the 9 th grade reported being bullied both ele ctronically and on school

PAGE 39

! 31 property significantly more than not; however, the majority of suicide attempters indicated they had not been bullied within the past 12 months. These are not expected findings given previous literature related to bullying as a su icide risk factor. Although bullying is moderate predictor of suicide, it is an interesting finding that overall more attempters reported they had not been bullied. Additionally, substance abuse variables of marijuana and alcohol were significantly pred ictive of suicide attempt; however, findings indicated that alcohol and marijuana use were only a weak to very weak effect size as predictors of adolescent suicide attempt. Although significant, the majority of student attempters did not use marijuana or d rink excess alcohol in the previous month. These findings are surprising given the results of other studies have indicate d alcohol and drug use are indicators of suicidal high school students (Borowsky et al., 2001; Esposi et al., 2004). This study also fo und that male students, who had attempted suicide in the past year, also indicated significantly higher marijuana use. Also 12 th grade students, who had previously attempted suicide in the past year, also reported higher rates of marijuana use. These findi ngs indicate that marijuana and alcohol use may only be weak risk factors for a suicide attempt ; however students who us e marijuana and are male and/or in the 12 th grade may be a t higher risk for a suicide attempt. A relationship between video gaming, i nternet use, and suicide has been supported by previous research (Kim et al, 2008; Mathers et al., 2009; Rehbein et al., 2010; Weaver et al., 2009; Lam et al. 2010). Findings from the current study are aligned with previous findings that across demographi c variables students who had previously attempted suicide reported significantly higher video game/internet use and that the majority reported watching "less than 3 hours" of television on an average school night; both of which were representative of the sample of suicide attempters. This was an interesting finding, indicating that overall students who attempted in the past year were also using the internet and/or playing video games for "more than 3 hours and spending "less than 3 hours" per day watchin g television. These findings may

PAGE 40

! 32 indicate that videogame/internet use may be a risk factor for suicidal student s ; however, since this finding was representative of the overall sample, it may indicate that video game/internet use and television are neither a protective n or a risk factor for adolescent suicide. Other studies also provided evidence of physical inactivity was a risk factor for suicide (Simon et al., 2004; Chioqueta et al., 2007). Youth physical activity guidelines, according to the U.S. Depa rtment of Health and Human Services, recommend that children and adolescents aged 6 17 years should have 60 minutes or more of physical activity each day (2008). Another unexpected finding was that physical activity and sports participation, although signi ficant, maintained a weak to very weak effect size. Overall, significantly higher proportions of suicide attempters reported being physically active for "0 4 days" during the past school week than those who reported physical activity of "5 or more days" ov er the past school week. Thi s was representative of the overall sample of students Furthermore suicide attempters were more likely to not participate o n a sports team over the previous year; however, sports participation varied across the demographic var iables. Of female suicide attempters, there was no significant difference between the level of sports participation, indicating that sports participation is not a significant protective or risk factor for female students. The same finding was true for race /ethnic groups of Black/African American, Hispanic/Latino, and all other race/ethnicities, which also indicates that increased level of sports participation is not a significant protective or risk factor for these groups of students. During the previous ye ar, there were higher reports of m ale attempters who participating in one or more sports teams than those who did not Ninth grade attempters also reported higher sports participation over the previous year, which may also indicate that higher levels of sp orts participation were not protective factors for these groups. This is an unexpected finding considering previous research supports involvement in certain extracurricular activities, particularly sports, as a strong protective factor against adolescent s uicide (Babiss & Gangwisch, 2009; Taliaferro, Rienzo, Miller, Pigg, & Dodd, 2008).

PAGE 41

! 33 In conclusion, s imilar to currently known risk and protective factors for high school students, this study found significan t predictors of suicide attempt including: (1) fe male gender, (2) race/ethnicity of White, Black/ African American, and Hispanic/Latino, (3) lower grade level (4) experience of bullying of both on school property and electronically, (5) fewer than 5 days of physical activity per week and (6 ) video game/ internet use. Although data indicated that demographic characteristics of gender, race/ethnicity, and grade level were moderately associated with attempting suicide specific group analysis indicated that there are unique subgroups of students that are af fected by risk and protective factors differently. Limitations and Future Research Directions There are a number of limitations of the current study. All data w ere self reported and accuracy of report cannot be examined. Data was then cross sectional ly an alyzed and therefore any causality cannot be concluded Another limitation relates to the fact that only participants who attended school were included in the data ; therefore students who were not in attendance the day of the survey and/or do not attend school could not be included in the sample. This is an important limitation due to the suicide risk for adolescents in the justice system and minority populations often do not attend public school or graduate (National Center for Education Statistics, 2010 ). Another limitation is the measurement of the grade levels, which are not precise due to students being retained and large age ranges A f urther limitation is that state level data w ere not available for all 50 states, as some did not participate and w ere therefore unable to obtain and present completely representative data. Additionally, YRBS questions directly addresses behaviors that contribute to the leading causes of illness and death among United States adolescents and adults w hich limits school and community focus on targeted proactive and positive interventions for suicidal behavior. Also, substance abuse questions of the YRBSS were limited in scope, with questions only focused on substance abuse over the prior month rather that into the scope a nd duration of use.

PAGE 42

! 34 T his study examined only the relationship between demographic characteristics and risk and protective factors of suicide attempters. This study did not consider the differences between the suicide attempters and non attempters. An ex amination of these differences may indicate potential significance in the e ffect of these risk and protective factors on each population. Further more this study did not include consideration of other known predictors of suicidal behavior such as depressio n, anxiety, and other mental health concerns Additionally, this study also did not analyze the impact on subgroup geography which may have an impact on the effect of risk and protective factors Research indicates that sui cide rates in rural areas are higher than urban areas (Hirsch, 2006 ), which may influence differences in opportunities and risks. Finally, there were many small sample sizes that found little to no significance due to multiple cross sectional analyses For future research, combining additional YRBS data may increase sample sizes and therefore statistical power of analysis. Implications for Practice It is critical that student risk factors be identified in order to appropriately assess for suicide risk and provided appropriate suppor t and intervention based on their unique characteristics. The findings of the present study have a number of implications for practice. School mental health professionals will need to be able to identify various suicide risk and protective factors in order to appropriately assess students at risk for suicide. When risk factors for suicide come to the attention of school personnel, it is crucial for school personnel and mental health to be aware of unique risk factors for different demographic groups of stud ents in order to appropriately address student suicide ris ks. The results of this study have implications for the professional training and practice for those who work with adolescent populations. In promoting proper identification and assessment of stud ents who are considering suicide or are at risk, there are many promising programs in means to prevent adolescent suicide. Mental health professionals who understand suicide risk

PAGE 43

! 35 and implement evidence based suicide prevention programs are crucial to savi ng adolescent lives. It is essential that school psychologists and other community based mental health professionals are current with evolution of effective assessment and intervention of the continuously changing and evolution of the adolescent population In order to maintain authority on effective intervention procedures, training would be beneficial through post secondary and continuing education opportunities (Crepeau Hobson, 2013) Competing Interests I confirm that there are no known conflicts of i nterest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.

PAGE 44

! 36 REFERENCES American Academy of Child and Adolescent Psychiatry, American Psychiatric Association (2004). Jo int statement from the American Academy of Child and Adolescent Psychiatry and the American Psychiatric Association for the hearing on suicide prevention and youth: saving lives Senate Substance Abuse and Mental Health Services Subcommittee of the Health, Education, Labor, and Pensions. http://www.aacap.org/galleries/Legislative Action/SuicideH.pdf. Accessed 5 November 2015. Aseltine, R., James, A., Schilling, E., & Glanovsky, J. (2007). Evaluating the SOS suicide prevention program: A replication and exte nsion. Bmc Public Health, 7 (1), 161 161. doi:10.1186/1471 2458 7 161 Babiss, L., & Gangwisch, J. (2009). Sports participation as a protective factor against depression and suicidal ideation in adolescents as mediated by self esteem and social support. Jour nal of Developmental and Behavioral Pediatrics, 30 (5), 376 384. doi:10.1097/DBP.0b013e3181b33659 Buda, B., Berman, A. L., Jobes, D. A., & Silverman, M. M. (2007). Adolescent suicide. Assessment and intervention (2nd ed.): (2006). Washington, DC: American P sychological Association. Crisis: The Journal of Crisis Intervention and Suicide Prevention, 28 (2), 105 105. doi:10.1027/0227 5910.28.2.105 Bonanno, R. A. & Hymel, S. (2013). Cyber bullying and internalizing diff iculties: Above and beyond the impact of t raditional forms of bullying. Journal of Youth and Adolescence 42 685 697. doi:10.1007/s10964 013 9937 1 Bossarte, R. M., & Swahn, M. H. (2011). The associations between early alcohol use and suicide attempts among adolescents with a history of major d epression. Addictive Behaviors, 36 (5), 532 535. doi:10.1016/j.addbeh.2010.12.031 Borowsky, I., Ireland, M., & Resnick, M. (2001). Adolescent suicide attempts: risks and protectors. Pediatrics 107 :485 93. Brener, N. D., Collins, J. L., Kann, L., Warren, C. W., & Williams, B. I. (1995). Reliability of the youth risk behavior survey questionnaire. American Journal of Epidemiology, 141 575 580. Brener, N. D., Kann, L., McManus, T., Kinchen, S. A., Sundberg, E. C., & Ross, J. G. (2002). Reliability of the 1999 youth risk behavior survey questionnaire. Journal of Adolescent Health, 31 (4), 336 342. Brener, N. D., McManus, T., Galuska, D. A., Lowry, R., & Wechsler, H. (2003). Reliability and validity of self reported height and weight among high school students. J ournal of Adolescent Health, 32 (4), 281 287. Brown, D., Galuska, D., Zhang, J., Eaton, D., Fulton, J., Lowry, R., & Maynard, L. (2007). Physical activity, sport participation, and suicidal behavior: US high school students. Medicine and Science in Sports a nd Exercise, 39 (12), 2248 2257. doi:10.1249/mss.0b013e31815793a3

PAGE 45

! 37 Brown, C. H., Wyman, P. A., Brinales, J. M., & Gibbons, R. D. (2007). The role of randomized trials in testing interventions for the prevention of youth suicide. International Review of Psyc hiatry, 19 (6), 617 631. doi:10.1080/09540260701797779 Cash, S., & Bridge, J. (2009). Epidemiology of youth suicide and suicidal behavior Current Opinion in Pediatrics, 21 (5), 613 619. Centers for Disease Control and Prevention. (2012). Youth Risk Behavior Survey Available at: www.cdc.gov/yrbss Centers for Disease Control and Prevention. (2013). Youth Risk Behavior Survey Available at: www.cdc.gov/yrbss Centers for Disease Control and Prevention. (2014). Youth Risk Behavior Survey Available at: www.cdc .gov/yrbss Centers for Disease Control and Prevention. (2010a). Web based injury statistics query and reporting system (WISQARS ) Retrieved November 6, 2015, from www.cdc.gov/injury/wisqars/pdf/10LCID_All_Deaths_By_Age_Group_2010 a.pdf Chioqueta, A. P., & Stiles, T. C. (2007). Cognitive factors, engagement in sport, and suicide risk. Archives of Suicide Research, 11 (4), 375 390. doi:10.1080/13811110600897143 Chung, S. S., & Joung, K. H. (2012). Risk factors related to suicidal ideation and attempted suicide : Comparative study of Korean and American youth. The Journal of School Nursing, 28 (6), 448 458. doi:10.1177/1059840512446704 Chabrol, H., Melioli, T., & Goutaudier, N. (2014). Cannabis use and suicidal ideations in high school students. Addictive Behavior s, 39 (12), 1766 1768. doi:10.1016/j.addbeh.2014.06.008 Crepeau Hobson, F. (2013). an exploratory study of suicide risk assessment practices in the school setting. Psychology in the Schools, 50 (8), 810 822. doi:10.1002/pits.21705 Crone, E., Dahl, R. (2012) Understanding adolescence as a period of social affective enga gement and goal flexibility. Nature reviews. Neuroscience 13 : 636 50. Dahl, R. (2004) Adolescent brain development: a period of vulnerabilities and opportunities Keynote address. Ann N Y Academic Science 1021:1 22 Department, A. C. (2001). Practice parame ter for the assessment and treatment of children and adolescents with suicidal behavior. Journal of the American Academy of Child & Adolescent Psychiatry, 40 (7), 24S 51S. doi:10.1097/00004583 200107001 00003 Degenhardt, L., Hall, W., & Lynskey, M. (2003). Exploring the association between cannabis use and depression. Addiction 98 (11), 1493 1504. doi:10.1046/j.1360 0443.2003.00437.x Dunn, M. S., Goodrow, B., Givens C., & Austin, S. (2008). Sub stance use behavior and suicide indicators among rural middle sc hool students. The Journal of School Health, 78 26 31.

PAGE 46

! 38 Epstein, J.A., & Spirito, A. (2010) Gender Specific Risk Factors for Suicidality Among High School Students, Archives of Suicide Research, 14 ( 3 ) 193 205, doi: 10.1080/13811118.2010.494130 Fine, J., Su ng, C. (2014) Neuroscience of child and a dolescent health development. Counseling Psychology 61 :521 527 Freedenthal, S. (2010). Adolescent help seeking and the yellow ribbon suicide prevention program: An evaluation. Suicide and Life Threatening Behavior, 40 (6), 628 639. doi:10.1521/suli.2010.40.6.628 Gauthier, J. M., Zuromski, K. L., Gitter, S. A., Witte, T. K., Cero, I. J., Gordon, K. H., & ... Joiner, T. (2014). The Interpersonal Psychological Theory of Suicide and Exposure to Video Game Violence. Journa l o f Social & Clinical Psychology, 33 (6), 512 535. doi:10.1521/jscp.2014.33.6.512. Heron M. (2013). Deaths: Leading causes for 2010. National vital statistics reports 62 ( 6 ) Hyattsville, MD: National Center for Health Statistics. Hinduja, S., & Patchin, J. W. (2010). Bullying, cyberbullying, and suicide. Archives of Suicide Research 14 (3), 206 221. doi:10.1080/13811118.2010.494133 Hirsch, J. K. (2006). A review of the literature on rural suicide: Risk and protective factors, incidence, and prevention. Crisis 27 (4), 189 199. doi:10.1027/0227 5910.27.4.189 Johnston, L. D., O'Malley, P. M., Bachman, J. G., & Schulenberg, J. E. (2008). Demographic subgroup trends for various licit and illicit drugs, 1975 2007 (Monitoring the Future Occasional Paper No. 69) [Online]. Ann Arbor, MI: Institute for Social Research. Retrieved from http:// www.monitoringthefuture.org/ Kerr, D., Owen, L., Pears, K., & Capaldi, D. (2008). Prevalence of suicidal ideation among boys and men assessed annually from ages 9 to 29 years. Suicide and Life Threatening Behavior, 38 (4), 390 402. Kim, D. S., & Kim, H. S. (2010). Early initiation of alcohol drinking, cigarette smoking, and sexual intercourse linked to suicidal ideation and attempts: Findings from the 2006 Korean youth risk beha vior survey. Yonsei Medical Journal, 51 18 26. Kim, E. J., Namkoong, K., Ku, T., & Kim, S. J. (2008). The relationship between online game addiction and aggression, self control and narcissistic personality traits. European Psychiatry 23 212 218. Klaus, N. M., Mobilio, A., & King, C. A. (2009). Parent adolescent agreement concerning adolescents' suicidal thoughts and behaviors. Journal of Clinical Child & Adolescent Psychology, 38 (2), 245 255. doi:10.1080/15374410802698412 Klomek, A., Sourander, A., & Go uld, M. (2010). The association of suicide and bullying in childhood to young adulthood: A review of cross sectional and longitudinal research findings. Canadian Journal of Psychiatry, 55 (5), 282 288.

PAGE 47

! 39 Konrad, K., Firk, C., & Uhlhaas, P. J. (2013). Brain De velopment During Adolescence: Neuroscientific Insights Into This Developmental Period. Deutsches €rzteblatt International, 110 (25), 425 431. http://doi.org/10.3238/arztebl.2013.0425 Lam, L. T., & Pen g, Z. W. (2010). Effect of pathological use of the internet on adolescent mental health: A prospective study. Archives of Pediatrics and Adolescent Medicine 164, 901 906. Fanslow, J. L. (1999). Making sense of suicide: An in depth look at why people kill themselves Oxford: Pergamon Press Inc. Lester, D. (Ed.). (1997). Making sense of suicide: An in depth look at why people kill themselves Philadelphia: Charles PressPub. Malley, E., Posner, M., & Potter, L. (2008). Suicide risk and prevention for lesbian, gay, bisexual, and transgender youth. Newton: Suicide Prevention Resource Center 35 36. Mathers, M., Canterford, L., Olds, T., Hesketh, K., Ridley, K., & Wake, M. (2009). Electronic media use and adolescent health and well being: Cross sectional communit y study. Academic Pediatrics 9, 307 314. Messias, E., Castro, J., Saini, A., Usman, M., & Peeples, D. (2011). Sadness, Suicide, and Their Association with Video Game and Internet Overuse among Teens: Results from the Youth Risk Behavior Survey 2007 and 20 09. Suicide & Life Threatening Behavior, 41 (3), 307 315. doi:10.1111/j.1943 278X.2011 <.001 30.x McIntosh, J. L. (2012). USA Suicide: 2009 Official Final Data Washington, DC: American Association of Suicid ology. Retrieved November 5 2015 from http://www.su icidology.org Miller, K., & Hoffman, J. (2009). Mental well being and sport related identities in college students. Sociology of Sport Journal, 26 (2), 335 356. Murphy, S., Xu, J., & Kochanek, K. (2012). Deaths: preliminary data for 2012 National Vital Sta tistics Reports, 60, 1 51. National Center for Education Statistics. (2010). Status and trends in the education of racial and ethnic minorities Retrieved from http://nces.ed.gov/pubs2010/2010015/indicator4_18.asp Nickerson, A. B., & Slater, E. D. (2009). School and community violence and victimization as predictors of adolescent suicidal behavior. School Psychology Review,38 (2), 218. Pena, J. B., Matthieu, M. M., Zayas, L. H., Masyn, K. E., & Caine, E. D. (2012). Co occurring risk behaviors among white, bl ack, and hispanic US high school adolescents with suicide attempts requiring medical attention, 1999 2007: Implications for future prevention initiatives. Social Psychiatry and Psychiatric Epidemiology 47(1), 29 42. doi:10.1007/s00127 010 0322 z

PAGE 48

! 40 Pedersen, W. (2008). Does cannabis use lead to depression and suicidal behaviours? A population based longitudinal study. Acta Psychiatrica Scandinavica, 118 (5), 395 403. doi:10.1111/j.1600 0447.2008.01259.x Price, C., Hemmingsson, T., Lewis, G., Zammit, S., & Alle beck, P. (2009). Cannabis and suicide: Longitudinal study. The British Journal of Psychiatry, 195 (6), 492 497. doi:10.1192/bjp.bp.109.065227 Rasic, D., Weerasinghe, S., Asbridge, M., & Langille, D. (2013). Longitudinal associations of cannabis and illicit drug use with depression, suicidal ideation and suicidal attempts among nova scotia high school students. Drug and Alcohol Dependence, 129 (1 2), 49 53. doi:10.1016/j.drugalcdep.2012.09.009 Rehbein, F., Kleimann, M., & Mossle, T. (2010). Prevalence and risk factors of video game dependency in adolescence: Results of a German nationwide survey. CyberPsychology and Behavior 13 269 277. Roberts R E Ramsay C Yun X ., (2010) One year incidence of suicide attempts and associated risk and protective factor s among adolescents. Arch Suicide Res 14:66 78 Roy, A., Lamparski, D., DeJong, W., Moore, V., Linnoila, M. (1990). Characteristics of alcoholics who attempt suicide. The American Journal of Psychiatry 147 761 765 Rylander, M., Valdez, C., & Nussbaum, A. M. (2014). Does the legalization of medical marijuana increase completed suicide?. American Journal Of Drug & Alcohol Abuse, 40 (4), 269 273. doi:10.3109/00952990.2014.910520 Shilubane, H. N., Ruiter, R. A. C., Bos, A. E. R., van den Borne, B., James, S., & Reddy, P. S. (2012). Psychosocial determinants of suicide attempts among black S outh A frican adolescents: A qualitative analysis. Journal of Youth Studies, 15 (2), 177 189. doi:10.1080/13676261.2011.634400 Simon, T. R., Powell, K. E., & Swann, A. C. (2004 ). Involvement in physical activity and risk for nearly lethal suicide attempts. American Journal of Preventive Medicine, 27 (4), 310 315. doi:10.1016/j.amepre.2004.07.003 Taliaferro, L. A., Rienzo, B. A., Miller, M. D., Pigg, R. M., Jr., & Dodd, V. J. (200 8). High school youth and suicide risk: Exploring protection afforded through physical activity and sport participation. The Journal of School Health, 78 545 553. Taliaferro, L. A., Rienzo, B. A., Pigg, R. M., Miller, M. D., & Dodd, V. J. (2009). Associat ions between physical activity and reduced rates of hopelessness, depression, and suicidal behavior among college students. Journal of American College Health, 57 (4), 427 435. doi:10.3200/JACH.57.4.427 436 Undheim, A. M. (2013). Involvement in bullying as predictor of suicidal ideation among 12 to 15 year old Norwegian adolescents. European Child & Adolescent Psychiatry 22 1 9. doi:10.1007/s00787 012 0373 7 U.S. Department of Health and Human Services (2008). Physical Activity Guidelines for Americans Washington, DC: U.S. Department of Health and Human Services. Retrieved from http://www.cdc.gov/healthyschools/physicalactivity/guidelines.htm

PAGE 49

! 41 Wang, J., Nansel, T. R., & Iannotti, R. J. (2011). Cyber bullying and traditional bullying: Differential assoc iation with depression. The Journal of Adolescent Health : Official Publication of the Society for Adolescent Medicine, 48 (4), 415 417. doi:10.1016/j.jadohealth.2010.07.012 Weaver, J. B., 3rd, Mays, D., Sargent Weaver, S., Kannenberg, W., Hopkins, G. L., Eroglu, D., et al. (2009). Health risk correlates of video game playing among adults. American Journal of Preventive Medicine 37 299 305. Wolitzky Taylor, K. B., Ruggiero, K. J., McCart, M. R., Smith, D. W., Hansen, R. F., Resnick, H. S., Kilpatric k, D. G. (2010). Has adolescent suicidality decreased in the United States? Data from two national samples of adolescents interviewed in 1995 and 2005. Journal of Clinical Child and Adolescent Psychology, 39 64 76. Wyman, P. A., Brown, C. H., Inman, J., C ross, W., Schmeelk Cone, K., Guo, J., & Pena, J. B. (2008). Randomized trial of a gatekeeper program for suicide prevention: 1 year impact on secondary school staff. Journal of Consulting and Clinical Psychology, 76 (1), 104 115. doi:10.1037/0022 006X.76.1. 104