This study of high school outcomes in the Baltimore City Public Schools builds on substantial prior research on the early warning indicators of dropping out. It sought to investigate whether the same variables that predicted a non-graduation outcome in other urban districts— attendance, behavior problems, and course failure – were also significant predictors of non-graduation in Baltimore.

The study specifically probed the relationship between eighth- and ninth-grade early warning indicators as predictors of graduation outcomes, as well as the relationship between ninth-grade indicators and college enrollment outcomes. In particular, it sought to address the following questions:

1. To what extent did students in two ninth-grade cohorts exhibit early warning indicators of nongraduation in eighth grade and ninth grade? To what extent were eighth- and ninth-grade early warning indicators correlated?
2. To what extent do eighth-grade early warning indicators (attendance, behavior problems, and course failure) explain the variation in graduation outcomes? How do they compare with ninth-grade indicators in their explanatory power?
3. To what extent do ninth-grade school-level factors influence non-graduation outcomes?
4. To what extent do eighth- and ninth-grade student outcomes influence college enrollment outcomes?
5. What do findings about the relationships between early warning indicators and graduation and college enrollment outcomes suggest about the kinds of intervention strategies needed to improve student outcomes?

The analysis was based on two cohorts of all ninth graders in Baltimore City Public Schools in 2004-05 and 2005-06,1 and drew on yearly data on school enrollment and withdrawal, grade level, attendance, test scores, suspensions, and course grades. In addition, data from the National Student Clearinghouse on college enrollment were merged into these cohort files.

The results are divided into three parts. Parts I and II present descriptive analyses of the data, including frequencies, cross-tabulations, means, and other descriptive summaries that show the relationship between various student behaviors/early warning indicators (such as absenteeism, GPA, or course failures) and high school graduation and college enrollment outcomes. Part III then reports the results of multi-level modeling analyses of the data, wherein the relative impact of eighth- and ninth-grade early warning indicators on high school graduation and college enrollment outcomes are presented.

• As expected, ninth-grade indicators proved to be more powerful predictors of high school outcomes than eighth-grade indicators, suggesting that interventions designed to prevent students from slipping into chronic absence and course failure in ninth grade are crucial for increasing the graduation rate in Baltimore and similar districts.
• At the same time, the strength of eighth-grade variables (particularly chronic absence) in predicting outcomes was striking. These findings provide evidence of the importance of interventions mounted prior to the beginning of ninth grade to help reverse chronic absenteeism and increase the probability of graduation for struggling students.
• Analyses also indicate the importance of explicitly addressing the needs of male students, since they are still significantly less likely to graduate, even when controlling for their higher levels of behavioral early warning indicators.
• In addition, the findings emphasize how being overage for grade reduces the probability of graduation, even controlling for the associated behavioral indicators. Finding ways to increase learning time during the school year and summer, rather than retaining students in grade in the elementary grades, may be a crucial step in reducing the number of students who fail to graduate from high school.
• The findings also indicate that eighth-grade proficiency in math and reading and both cumulative GPA and ninth-grade GPA increase the probability of college enrollment, suggesting that the “gatekeeper issues” of low ninth-grade GPAs and the lack of proficiency upon entrance to high school need to be addressed as crucial steps for reaching the goal of raising college readiness rates and the proportion of students who enroll in college after high school.

The discussion section of this report examines how the district has responded to research on early warning indicators. While important steps have been taken, particularly in the area of increasing attendance and reducing the number of suspensions, challenges remain for the district in preventing students from falling off-track to graduation through course failure.

The discussion explores structural factors that may be influencing the district’s orientation towards early warning indicators, including the district’s decentralized, “portfolio” approach to school governance. Several topics and areas for future research are identified, and the report concludes with an examination of the role the Baltimore Education Research Center could play in future district efforts to address and integrate research data into school reform efforts.

Download the full report available here in pdf.