The detrimental effects of dropping out of school have been well documented, ranging from an increased rate of unemployment and a greater likelihood of being incarcerated to an increased propensity for health problems. Despite widespread and long-standing attention to the issue, school dropout persists with some estimates as high as 1 of every 5 students not graduating. Given the adverse impact on an individual and persistent presence of the issue, it is imperative that school leadership can identify students who are “at risk” or “off-track” early enough for interventions to effectively re-engage students. Therefore, the purpose of this session – “Using Early Warning System Data to Ensure Students Don’t Fall Through the Cracks” – is to examine which student-level indicators can be used to predict on-time cohort graduation and how those indicators can be systematically monitored through Performance Matters’ Early Warning System.
A systematic examination of the relevant literature on dropout and early warning systems has identified several indicators that can be used to predict on-time graduation including attendance, behavior, academic performance, retention, student attitudes, and socioeconomic status. While individually these factors are predictive of dropout rates, combining these factors have greatly enhanced the predictive power for identifying students at risk of dropping out. This points to the fact that the root cause of dropping out does not lie in the individual indicators. Rather, these are but symptoms of the antecedents of disengagement from the educational process. Therefore, it is vital that schools possess the ability to identify students exhibiting multiple indicators, as each individual student’s disengagement will manifest itself through a variety of behaviors. It is through the identification of the multiple indicators that interventions can effectively be matched to individual student needs. However, merging this data from several different data systems can be cumbersome, if not impractical for schools.
Understanding the potential dividends and inherent difficulty in systematically tracking multiple engagement data sources, Performance Matters developed a series of filters through an Early Warning System. This system utilizes attendance, behavior, demographic, and academic data and allows school districts to define thresholds to monitor student level indicators. These filters can then be easily used to cluster students with characteristics of disengagement behavior, allowing school systems to systematically monitor and intervene. The School District of Indian River County has been using Performance Matters’ Early Warning System for the past several years and has seen their district graduation rate climb from 79.1% in 2013-14 to an estimated 86% in 2015-16.
Education remains an integral mainstay to any successful society. Unfortunately, despite an enormous collective effort by researchers, educators, and community members, students drop out of school at alarming rates. Even with a growing body of research exploring this very issue, many questions remain. However, the use of an Early Warning System has shown to be a huge advantage for school systems seeking to address this persistent issue.
Brian McMahon is a Performance Data Analyst with the School District of Indian River County and is working on his dissertation on Early Warning Systems through Florida Atlantic University. Bringing 14 years of experience in education, Brian is passionate about using data in problem-solving. However, he is most proud of his wife, Dr. Christina Jacobs, and their four children. You can follow Brian on twitter at @SDIRCMcData.