Preventing High School Dropouts by Using Early Warning Indicators
A crucial element of today’s educational climate is rapidly and routinely identifying students in danger dropping out of high school. The most recent statistics from the National Center for Education Statistics indicate that approximately 1.2 million students drop out of school every year. That translates to a student dropping out of school every 26 seconds!
While this is an improvement over the lows of the mid-1990s, educators remain, rightfully, concerned
“Dropping out of school persists as a problem that interferes with educational system efficiency and the most straightforward and satisfying route to individual educational goals for young people.”
– National Dropout Prevention Center
Students drop out for a wide variety of reasons. Some are pushed out due to poor attendance or failing grades (or the anticipation of failing grades). Some feel it would be easier to get a GED than to complete high school. Many do not like school, do not get along with their teachers, or simply feel that they do not “belong.” Some leave for family reasons or because they got a job.
Regardless of their reasons, young people without a diploma face significant economic challenges: high school dropouts are more likely to be unemployed and more likely to earn less than if they had a diploma. This creates a lifetime of reduced income and perpetuates the cycle: students from low-income families are four times more likely to drop out of high school.
So how can you as an educator get to the root cause of why students are dropping out so you can design effective interventions?
Build Consensus Around Metrics
Good decisions depend on good data. Before you can begin examining and exploring your data to identify at-risk students, you need to decide which indicators contribute to the issue. Some obvious indicators include what are commonly known as the ABCs of early warning:
- Attendance rates
- Behavior incidents
- Course performance
However you and your stakeholders may wish to consider other indicators, such as what programs students participate in—including whether, in fact, they participate in these programs at all. You may wish to consider test results from a variety of assessments, including high-stakes tests, diagnostic tests, end-of-term or end-of-unit tests, etc. Studies on dropout prevention recognize some variation in indicators by school and district. Focus on those indicators that matter to you.
Recognize that those indicators may change depending on grade level. Attendance is a critical indicator in early education (K–5) as that’s where foundational skills are learned. Once students transition to middle school, the focus often shifts to behavior and discipline incidents.
Consider including marking periods in your tracking to help you identify changes in student risk profile throughout the school year. Students who are trending toward higher risk levels may need additional interventions. Students who are trending to lower risk levels may demonstrate that your interventions are working.
5 Questions to Ask Yourself:
- What indicators appear to contribute to a student dropping out? Do those indicators change by grade level?
- How common are those indicators across the student population?
- How many risk indicators do I want to track, and how easily can I access the data to track them?
- Do I divide my data by marking period so I can track changes in risk status at a more granular level throughout the school year?
- Do I want to change the threshold values that I establish for the indicators throughout the year to more easily focus on the most at-risk students?
Examine and Explore Your Data
To identify the indicators most useful to your district and students, plan to conduct a study of your data to discover the 5 to 10 indicators that you can see contributing to the dropout rate in your district. Once you’ve identified those indicators, use that information to set thresholds to identify the degree of risk the student is facing. Your analysis tool should be able to use these indicators and threshold values to paint a realistic and usable picture of your overall student risk.
Once you have the overall picture, you should be able to drill into specific populations and even to individual students. It’s often important to be able to disaggregate your at-risk student population, understanding how it varies by ethnicity, gender, meal status, or program participation, as well as understanding differences across high schools and grades.
This deeper analysis helps you uncover true root causes. Understanding the root causes leads to strong yet flexible interventions and programs that will serve more students more effectively.
Monitoring is also key. Because students are inherently growing and changes, your early warning system should not be “one and done”—don’t just do the analysis, look at the results, and move on. Revisit your data periodically to see what’s changed.
Plan regular monitoring of your early warning system so you can demonstrate how well your planned programs and interventions are working. Being able to see populations of students move from At Risk to On Track shows that you are making progress. Seeing which students remain at risk helps you refine your interventions to help those populations or individuals further.
5 More Questions to Ask Yourself
- What thresholds do I want to set for each indicator?
- Do those thresholds vary by grade?
- What status categories do I want (e.g., On Track and At Risk? On Track, Moderate Risk, At Risk?)
- What threshold values do I want to set for each status category?
- How often will I monitor and adjust interventions and programs for At Risk students?
How can Scantron help?
Scantron has an extensive track record providing assessment solutions and services, and for more than nine years has also offered Scantron Analytics to help schools get the most out of their data.
Powered by the world-class analytics engine QlikView®, Scantron Analytics presents up-to-date information through highly visual, easy-to-understand dashboards, including an early warning system that districts are using to identify students at risk of dropping out of school. Using patented technology, Scantron Analytics delivers powerful analytics without the need for a separate data warehouse. Using information you’re already collecting, sourced from a wide variety of educational systems, Scantron Analytics displays easy-to-read graphical dashboards and data visualizations.
For the Early Warning System included as part of Scantron Analytics, we identified ten separate indicators, including absence rates, course failures, GPA, and discipline incidents, both within a marking period as well as across the school year. The indicators appear in an easy-to-read dashboard so you can see at-risk students at a glance. Using the powerful data exploration features in Scantron Analytics, you can track students whose risk profile has changed between marking periods and immediately begin to apply interventions. With a single click, you can disaggregate your at-risk student population, understanding how it varies by ethnicity, gender, and meal status, as well as understanding differences across high schools and grades.
While a number of at-risk indicators are common across schools—in particular: attendance rates, behavior incidents, and course performance, the Early Warning System can easily be tailored for your specific local needs, so you can focus on the indicators and threshold values you have found to be most reliable. Psychometricians and researchers in Scantron’s Assessment Services group can help you make sense of your historical data and identify additional indicators you may want to track. They can guide you in setting thresholds that readily locate the at risk students you need to help most.
Whatever the assessment or analytics assistance you need, Scantron has the products, tools, services, and expertise to help you ensure that you have the right program for your students. Our software combined with our comprehensive suite of assessment services help you get the most out of your data, assessments, and solutions.
Back to Scantron Analytics
© 2015 Scantron Corporation. All rights reserved.