Latino Students in Suburban High Schools: An Examination of School Malleable Predictors of High School Graduation

Despite recent gains, Latino students drop out of high schools in the United States at a higher rate than their peers and at a significant cost. As more Latino students move to the suburbs, it is important for suburban high schools to meet their needs and solidify their path to graduation. Practices that appear to be promising in keeping students on track to graduate have been developed and studied in urban education settings and need to be validated with other populations and settings. The purpose of this study was to investigate the indicators that are predictive of on time graduation for Latino students in a suburban school district and to determine if the Consortium for Chicago School Research’s on-track indicator was a useful predictor of on time graduation for a Latino population in a suburban setting. Demographic, attendance, discipline, and academic data were examined for 317 Latino first-time 9th graders during the 2009-2010 school year to determine the strongest predictors of graduation within four years from a suburban school district in the Midwest. Logistic regression was used to determine the strongest predictors of on time graduation. Overall, alterable variables (academics, behavior, attendance) were better predictors than static demographic variables. When utilizing data that is available during the first semester of freshman year and controlling for the effects of other variables and each other, English grade and unexcused absences were the strongest predictors of which Latino students would graduate on time within this suburban high school context. The on-track indicator was also a strong predictor. Whether or not a student was in an ESL program their freshman year provided additional predictive power to the English grade and unexcused absences model as well as to the on-track indictor, suggesting that using these models with students in ESL may underestimate the number of students at risk. Limitations of the current study and implications are discussed