Research has identified factors that are highly predictive of students who drop out of school, such as failing middle school courses or having poor attendance.

School staff sometimes know which students have factors that put them at risk. Few schools, however, have a process that looks at all students across multiple risk factors and that shows the combined effect of those factors. With a reliable predictor of risk, the principal-headmaster and teachers know which students to start helping right away, allowing for prevention rather than recovery.

In response to requests from schools involved in BPE’s initiatives to target struggling students, the organization’s research team created the Excel-based Composite Learning Index (CLI), which incorporates 15 indicators of risk: six social/behavioral and nine academic. Each indicator is assigned a weight: the greater the correlation to risk, the higher the weight. Each student’s weighted scores are then used to create a single CLI score and to place him/her on a continuum of risk, from “Off Track” to “On Track.” The result is a much more complex, holistic profile of each student as he/she starts school, extending beyond the narrow lens of past test performance.

Using the CLI color-coded spreadsheet, school data teams can easily sort and filter the data to answer questions for each student:
Q. Is he/she at risk of failing or dropping out? If so, how great is the risk?
Q. What specific academic, social, or behavioral indicators put him/her at risk?
Q. What interventions are most appropriate?

Piloting the CLI this year, school staff sometimes found that their assumptions about students were wrong. In most schools, for example, the majority of students “Off Track” were not in special education classes, as expected. In one school, many students at high risk of dropping out had already passed MCAS. The CLI pointed to other needs and different interventions.
As now used by nine schools working with the Boston Plan, the CLI has demonstrated many benefits:
• It tracks risk factors that are recognized by national and local research as highly predictive of failing or dropping out
• It incorporates data from multiple sources and reports data on any indicator or combination of indicators
• Unlike most tools that display data by groups of students, it focuses on the individual student — both assets and needs
• It enables school staff to hone in on helping students with the greatest immediate risk
• It allows school staff to manipulate data based on their contexts and circumstances, look for patterns, and answer their own critical questions
• It informs planning, professional development, social services, and resource allocation
• It can be refreshed regularly, by adding students' grades and attendance after each marking period, for example
• It is available in August for students who are entering the school and is updated through the school year

Why must a district (or district partner, such as a local education foundation) create and manage the CLI? Why can’t schools do this on their own? Although the data comes from the central office, it comes from different departments and in different formats, not easily integrated. Other data is not available in the central office and has to be compiled by the school. Finally, few school staff have both the time and the skills to pull together such a complex set of data without errors.

The CLI can also be customized to other districts.

Click here to see a sample CLI. For more information, e-mail cli@bpe.org or call 617. 227.8055 x346.

* In Boston, two major independent studies by The Parthenon Group and by Dr. Robert Balfanz of Johns Hopkins University have analyzed BPS dropout data to identify BPS-specific predictors, most of which parallel national research.

SPECIAL THANKS ...

We would like to extend our thanks to those who have influenced our thinking in our development of the CLI: Mary Skipper at TechBoston Academy who helped us at a conceptual level; and researchers at The Parthenon Group, Johns Hopkins University,and the Chicago Consortium on School Reform whose recent workon early indicators of dropping out helped us with determining both which indicators should be used and how much they should be weighted.