Academic journal article Journal of School Health

Using Path Analysis to Examine Adolescent Suicide Attempts, Life Satisfaction, and Health Risk Behavior. (Research Papers)

Academic journal article Journal of School Health

Using Path Analysis to Examine Adolescent Suicide Attempts, Life Satisfaction, and Health Risk Behavior. (Research Papers)

Article excerpt

Suicide currently ranks as the second-leading cause of death for youth aged 15 to 19 years. (1) The suicide rate for this group has increased more than 300% since 1950 while the overall suicide rate has remained fairly stable. (2) In addition, for each youth suicide, an estimated 200 youth attempt suicide. (3)

Research has determined risk behaviors associated with adolescent suicide attempts. (4-6) Mass population programs to reduce associated risk behaviors generally have been ineffective and even detrimental to program participants. (7-9) One hypothesis suggests that programs focus on specific race/gender behaviors. This study examined two questions: What risk behaviors predict adolescent suicide attempts? Do the risk behaviors differ between specific race/gender groups?



The sampling frame included 215 public high schools, in South Carolina, selected systematically using a random start within each stratum, proportional to minority enrollment. Eighty-seven schools in 53 school districts were selected randomly to obtain a 6,500 student sample. During 1997, the survey was administered to 63 schools in the sampling frame (72%), totaling 5,544 of the proposed 6,500 public high school students (85.2%).

Study protocol required that passive consent, parental notification forms be distributed. Less than 1% of the study population (n = 58) declined to participate in the study due to parental decline. The survey was administered anonymously, and students were informed that participation was voluntary. Completed answer sheets were placed by the student in an envelope and sealed before the administrator left the survey area. The University of South Carolina review board accepted these methods.


A modified version of the 1997 Centers for Disease Control and Prevention's (CDC) Youth Risk Behavior Survey (YRBS) was used for the study. This version contained CDC's 84 core questions measuring demographics and the original six risk-taking behavior categories: intentional and unintentional injuries, tobacco use, alcohol and other drug use, sexual activity, physical activity, and nutrition habits. Two additional constructs were added: life satisfaction and quality of life. Life satisfaction included a six-item construct gauging one's satisfaction with self, school, where they live, friendships, family, and overall life. These questions used a 7-point, Likert-type scale ranging from "terrible" to "delighted." Quality of life involved a 4-item construct consisting of a 5-point, Likert-type scale (excellent to poor); a perception of overall health question; and three, 7-point, past 30 day questions gauging subjects' perceived physical health not being good, mental health not being good, and number of days these physical and/or mental health problems kept them from doing their usual activities.

Data Analysis

All statistical calculations were performed using PC SAS.[R] All survey data were inspected for consistency with the answering schema. This procedure used 16 different combinations for verification of consistent answering; inconsistent data were removed.

Because a path analytic approach using a self-report questionnaire has rarely, if ever, been used in research on attempted suicide, a priori categorization of variables was conducted for the antecedent (health risk-taking behaviors) and mediator variables (life satisfaction, quality of life, and situational variables) based on logical assumption and review of the literature. Logistic regression analyses confirmed categorization by determining direct associations between the antecedent, mediating, and outcome variables of interest and intercorrelations among the antecedent variables.

Logistic regression analyses were performed on each race/gender group to determine initial associations among the variables of interest. Specifically, logistic regressions were performed on antecedent variables to determine associations with the outcome variable of attempted suicide. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.