School structural and policy variables associated with student smoking

Abstract

School of Psychology, Deakin University, Geelong, Victoria, Australia 3217 V Clarke Abstract Objective -To assess the relationships between school structural and smoking policy variables and students' self-re ported smoking. Design Questionnaire data relating to schools were collected from teachers and data relating to self-reported smoking were collected from students. Setting Australian secondary schools. Participants One teacher from each of 347 schools and 26429 students from these schools. Main outcome measure Self-reports of smoking. Results For each school the percentages of students who had smoked in the last week, month, and year were obtained for three groups of students: (a) years 7 and 8, (6) years 9 and 10, and (c) years 11 and 12. For consistency across time periods, reporting was limited to smoking in the last week. Schools varied in the pro portions of students who reported smok ing. School-based variables related to smoking more consistently for the younger students, for whom smoking was more prevalent in schools which were (a) co-educational rather than single-sex, (6) government rather than non-govern ment, (c) had a student representative on the school council, (d) did not have a house system, and (e) did not have a past students' association. For older students smoking prevalence was positively re lated to non-English speaking back ground. For all groups, smoking preva lence was unrelated to (a) school location (rural or urban), (6) school size, (c) means of selecting prefects, (d) student smoking policy, (e) staff smoking policy, (/) staff smoking prevalence, (g) visitor smoking policy, and (h) school no smoking signs. Specialised health edu cation programmes related positively to smoking, probably as these programmes were introduced into the schools with greater smoking prevalences. When the variables which were significant indi vidually were entered into regression analyses to predict smoking there were few significant relationships. ConclusionSchool-based structural and smoking policy variables have minimal association with students' reported smoking behaviour.