Using Data to Stem Dropout Rates

Researchers from Edvance Research explore early warning systems and how longitudinal data can be used effectively

Historically, about 25 to 30 percent of a ninth-grade high school class will quit before graduating. While the current high school dropout rate is not the highest it has ever been, many believe the United States is indeed facing a dropout crisis. Why is that?

Two factors make it a crisis today. First, the dropout rate for minority students is as high as 45 to 50 percent in some states. Second, the skills necessary for many 21st century jobs are expected to be much higher-level than in the past. Unlike previous generations, for whom unskilled jobs were plentiful, young people who drop out today will be unable to find sustainable employment without gaining additional skills. 

As a nation, we will be unable to sustain our economy and society without an educated workforce. Staying in school and graduating is the primary student outcome on which we must focus. 
For all the debate surrounding the impact of No Child Left Behind, NCLB’s focus on accountability has brought sustained attention to student-level data over the last 10 years. The question is how to use data as part of early warning systems to intervene and help keep students in school. Answers are coming from the likes of: 

  • Robert Balfanz, codirector of the Everyone Graduates Center at Johns Hopkins University, and a team of researchers from the University of Chicago, who worked with the School District of Philadelphia to understand how early students begin the prolonged process of dropping out
  • Researchers at the Consortium on Chicago School Research (CCSR), who studied the relationship between ninth-grade performance and the likelihood of students graduating on time, identifying a ninth-grade on-track indicator of credits earned and failure in core subject areas
  • Researchers at REL Southwest at Edvance Research, who adapted the CCSR study in Texas to determine how well the ninth-grade on-track indicator identified on- and off-track students across five districts, and
  • Researchers at the RAND Corp., who described “actionable indicators” that can inform intervention on student progress, rather than merely report on student outcomes.

Converging efforts are now under way in several parts of the country to establish early warning systems with the help of the growing body of knowledge about how best to put data to work in education. A recent report by Civic Enterprises, funded by the AT&T Foundation, describes efforts in 12 locations to use early warning systems to reduce student dropout rates. Such efforts are varied, involving schools, districts, community organizations, business, state education agencies, and researchers. 

At REL Southwest at Edvance Research, we have worked with districts and states to synthesize these and other findings in an effort to determine how best to incorporate these research findings into educators’ daily practice. For example, we worked with districts in Texas to determine if the CCSR ninth-grade on-track indicator could be adapted to these districts’ needs.

All benefit from the common denominator of early warning indicators, described by Balfanz as “the ABCs of dropping out”: attendance, behavior, and course grades. His efforts to focus on these seemingly simple indicators—which educators ought to have easy access to—are yielding outstanding results for schools that use them.

The problem is that most educators, in fact, do not have ready access to these data. Although plenty of data is entered into student information systems (SIS) by school and central office staff, most data systems were not designed to provide reports or analyses to school-level staff. 

Until recently, reporting of student and school outcomes by states only occurred annually, based on data submitted by districts and aggregated at the district and school levels. While annual results are important for accountability purposes, such indicators are of little value to school site staff, other than as historical markers. Such information is sometimes referred to as autopsy data: It tells you why the patient died, but what you really needed was information to keep him alive.

Annual attendance, for example, is not as helpful in keeping students in class as knowing the percentage of chronically absent students in a given reporting period (i.e., week, month, or grading period). Similarly, teachers who monitor the percentage of students failing, or near failing, in math or reading each reporting period then can work together to intervene as needed to keep students on track for graduation.

What is needed is timely access to the ABC data captured by all SIS, in formats that enable school staff to see trends and patterns that negatively impact the district’s achievement goals. How many absences are too many? What constitutes “poor behavior”? Which course grades matter the most? Schools and districts must work together to make these data available to the educators who are responsible for making decisions that impact students. Your SIS vendor can help by adding reporting features to their systems that allow users to easily create and access reports.

In our experience at Edvance Research, just providing access to data, even to early warning indicators, does little to improve student outcomes. However, linking these indicators to interventions and instructional strategies that have demonstrated effective results can be a powerful lever to accelerate change.

With the help of a Michael and Susan Dell Foundation grant, in working with districts we found that educators spend a lot of time searching for effective interventions by sorting through research, talking to colleagues, attending conferences, and making site visits. But tools that could compile this information and make it easily accessible will shorten this process and enable quicker responses to student needs. That kind of database is now part of what Edvance Research will have available to districts later in 2012.

The work does not end there, however. Frequent monitoring of implementation results and adjustments as needed are the keys to closing the loop on continuous improvement.
How can your school or district build the right system to ensure that indicator data and interventions make it into the hands of teachers working with individual students? We offer these suggestions that are both practical and evidence-based.

  1. Understand early warning indicators. There is value in manually compiling early warning indicator data yourself so you see how they are constructed, and to get a glimpse of how useful this data would be. Make sure the indicators you use are research-based and effectively identify at-risk students in your school or district.
  2. Replicate early warning indicator studies. If your school or district has the time and resources for such an effort, or can work with a university or regional educational laboratory to do this, you can try to replicate studies done elsewhere to determine how well they would describe your students. The results may provide powerful justification for broader buy-in by your colleagues.
  3. Identify a data coach in each school. Great demands are placed on teachers’ time, so many are unable to make effective use of data, even when it is readily available to them. A campus data coach can work with teachers to identify individual student needs, and help teachers see patterns and trends in performance.
  4. Provide professional development for teachers. Make sure all teachers are trained to recognize the early signs of dropping out and that they review their student data on a frequent basis to identify students at risk. 

Implementing an early warning system is a tangible, doable, high-yield strategy that will allow you to take advantage of data available when you make important decisions in your school or district. From that experience, you will be able to identify early warning indicators for other aspects of student success. It’s never too late to get started on such an important task. You can have a real impact on the life of a child starting right now.

References

Allensworth, E. M., and Easton, J. Q. (2005). The on-track indicator as a predictor of high school graduation. Chicago: University of Chicago, Consortium on Chicago School Research.

Balfanz, R., Herzog, L., and MacIver, D. J. (2007). Preventing student disengagement and keeping students on the graduation path in urban middle-grades schools: early identification and effective interventions.

Balfanz, R., Bridgeland, J. M., Moore, L. A., Fox, J.H. (2010). Building a grad nation: progress and challenge in ending the high school dropout epidemic.

Bruce, M., Bridgeland, J. M., Fox, J. H., Balfanz, R. (2011). On track for success: the use of early warning indicator and intervention systems to build a grad nation.

Hartman, J., Wilkins, C., Gregory, L., Gould, L. F., and D’Souza, S. (2011). Applying an on-track indicator for high school graduation: adapting the Consortium on Chicago School Research indicator for five Texas districts.

Marsh, J. A., Pane, J. F., and Hamilton, L. S. (2006). Making sense of data-driven decision making in education: evidence from RAND research.

Pinkus, L. (2008). Using early warning data to improve graduation rates: closing cracks in the education system. Washington, DC: Alliance for Excellent Education.

Roderick, M. R. (1993). The path to dropping out: evidence for intervention. Westport, CT: Aubum House.

Schaffhauser, D. (2011). Swimming With Data. T.H.E. Journal.

Sparks, S.D. (n.d.). More States Flag Potential Dropouts With Warning Data. Education Week.

Featured