The scale of student absenteeism today is large and worrisome, exacerbated by pandemic disruption to the routine of school. But the problem itself is not new, and numerous efforts have been undertaken over the years to address it. One of those is the Early Warning System Network (EWS), which has been in place in various forms since the 1980s. EWS is intended to predict those students who are at risk of multiple absences and to flag them to school administrators before the problem becomes chronic. New research indicates that it’s positive step toward a solution but not enough by itself.
Indiana University researcher Yusuf Canbolat looks at data from an unnamed school district in the southeastern United States. The district is large and socioeconomically diverse, with approximately 80,000 students. Just over half are low-income, as determined by eligibility for free or reduced-price lunch. Sixty percent of students are White, 24 percent are Hispanic, and 8 percent are Black. The study covers attendance data from the 2020–2021 and 2021–2022 school years, excluding kindergarten and twelfth grade students who do not have two consecutive years of data, for a total sample of 66,223 students. In both years, in-person school was the default although virtual schooling was available to families who wanted it. Absenteeism was counted the same in both in-person and virtual modes, and days in quarantine were not counted toward absence for any student.
The district requires schools to use EWS to monitor student absenteeism and to trigger intervention. EWS categorizes the risk levels based on the percentage of missed instructional time, which is equivalent to the percentage of absent days, either excused or unexcused, in a school year. District rules count students present if they attend at least one period in a day. EWS identifies students who miss less than 4 percent of instructional time during a school year as “on track”; those who miss more than 4 percent but less than 10 percent as “at risk”; and those who miss more than 10 percent as “off track.” Based on those cut-offs, 56 percent of the students were identified as on track; 28 percent as at risk; and 16 percent as off track in the 2020–2021 school year.
Canbolat’s models use students who are just above the “at-risk” and “off-track” thresholds as the two treatment groups—with the treatment being the simple act of applying those EWS labels to those students—while the control groups are their peers who are just below each cutoff.
He finds that EWS identification has no significant effect in reducing either moderate (at-risk) or chronic (off-track) absenteeism among low-income students, as well as moderate absenteeism among their higher-income peers. EWS identification did, however, reduce chronic absenteeism for higher-income students by 22 percent (or 1.3 percentage points).
We don’t know what which actions districts took no action in response to EWS triggers, but something positive is likely occurring with regard to higher-income students who reach the off-track threshold. For everyone else, perhaps the interventions are misaligned with the various causes of absenteeism—especially across socioeconomic strata—and are thus insufficient to change students’ trajectories. For example, if a student is absent because he must work a job late into the weekday evenings, no amount of pep talks about the importance of school—nor any amount of suspensions even—will change his patterns. Likewise, a typical attendance intervention will likely not motivate a bullied student to attend school every day, no matter how important they deem education.
In the end, an early warning system for student absenteeism is only as good as the interventions that follow. In this district, and probably many others, the causes of absenteeism are so resistant to existing remediation efforts that the warning might as well be sounding in a vacuum.
SOURCE: Yusuf Canbolat, “Early Warning for Whom? Regression Discontinuity Evidence from the Effect of Early Warning System on Student Absence,” Educational Evaluation and Policy Analysis (January 2024).
