In terms of the demographic make-up of the schools, one study has suggested that the total student enrollment at at-risk adolescents' schools has implications for their school engagement. In a sample of at-risk 8th graders from low-SES or rural communities, Finn and Voelkl (1993) found that adolescents from schools with greater enrollment had a greater likelihood of being absent or tardy, lower teacher reports of their engagement at school, lower self-reported attendance, and lower self-reported preparation for class than students from smaller schools. It is important to note that this study is cross-sectional and therefore causality cannot be determined. Yet, the authors include in their analysis an array of variables that might be related to the types of schools students attend to attempt to minimize any pre-existing differences in students who attend larger or smaller schools, such as family SES, the adolescent's race, and the percentage of students and faculty who are African American.