Validity of self-report in assessing school climate and risk behavior

This dissertation presents a line of research aimed at improving the assessment of school climate and student risk behaviors by student self-report. Research suggests that school climate, broadly described as the quality of interaction between adults and students at school, may influence student behavior. Over the last decade, adolescent substance use, violent behavior, and suicidality have been targeted as priority health risk behaviors. It should be noted that most research on the prevalence of these behaviors is conducted using self-report surveys.;The first manuscript of this dissertation examined the influence of school size on student perceptions of bullying, fighting, and threatening behavior at school. Although there is a widespread perception that larger schools have a greater incidence of student misbehavior and violence than smaller schools, this study presents reasons to be wary of this view. Student victimization was assessed in a statewide sample of approximately 7,431 9th grade students and 2,353 teachers in 290 Virginia high schools. School size was distinguished from the proportion of students receiving free or reduced price meals, percentage of minority students, ethnic diversity (heterogeneity), and urbanicity. In larger schools, teachers and students reported that they perceived more bullying and teasing taking place, but student self-reports of being a victim of bullying were not associated with school size. Additionally, school discipline records showed that, although the total number of incidents was higher, the rate of bullying offenses was lower in larger schools. Similar results were found for measures of student threats and physical assaults. These findings raise the possibility that the link between larger schools and higher student victimization is an illusion based on perceived frequency rather than rates of victimization.;The second manuscript examined the use of validity screening items in adolescent survey data. In three different studies, adolescent respondents were asked whether they were telling the truth and paying attention in answering survey questions. In Study 1 (N = 7,801), the prevalence rates of student risk behaviors were significantly lower after inappropriate responders were screened out of the sample. In Study 2 (N = 7,732), confirmatory and multi-group factor analyses demonstrated significant differences between the factor structures of valid versus invalid responders. In Study 3, student perceptions of school climate were correlated with teacher perceptions in 291 schools. A bootstrap resampling procedure compared the correlations obtained using valid versus invalid responding students in each school and found that valid responders produced higher correlations with teacher perceptions. Together, these studies support the value of validity screening items to improve the quality of adolescent survey data.;The third manuscript used exploratory and confirmatory factor analyses with a sample of 3,687 Virginia 9--12 grade students to establish reasonable fit for 20 items with three hypothesized school climate scales that measure student willingness to seek help for bullying and threats of violence, attitudes supportive of aggressive behavior, and the perceived prevalence of bullying and teasing at school. Structural equation modeling established the relationship of these three measures with individual student risk behavior scales, including externalizing, internalizing, and bullying risk behavior. Multi-group confirmatory factor analyses identified differential effects across gender and race. Results extend previous research and support the view that a positive school climate can be an important protective influence on student risk behavior.