While most schools deem students chronically absent after they miss at least 10% of school days, poor attendance habits may affect academic achievement well before they reach that threshold, a new study finds.
Absence rates of around 3% to 7% signal a risk of scoring below grade-level expectations on state tests down the road, finds an analysis of 9,000 Boston public school students. Researchers studied data from a cohort of students who enrolled in district prekindergarten programs between the 2007-2008 and 2009-2010 school years, tracking year-over-year attendance and achievement information through 8th grade.
“We rarely stop to determine if we are measuring [chronic absenteeism] in a useful way,” said Tiffany Wu, a doctoral student in education and psychology at the University of Michigan and the report’s lead author. “There has been relatively little empirical evidence to justify that cutoff.”
The findings, published in a January working paper, come as schools work to improve attendance patterns that cratered during the pandemic and as states hold districts accountable for reducing rates of chronic absenteeism.
Nationwide, 15% of students were chronically absent in 2018, according to the Return to Learn tracker maintained by the American Enterprise Institute. That number hit its peak in the 2021-22 school year, when 28% of students were chronically absent. Of the 41 states that have reported data from 2024-25, 34 have shown declining rates of chronic absenteeism, but none have returned to pre-pandemic levels, according to the tracker.
The new study’s authors relied on a statistical concept called signal detection theory, which is used to predict whether a measure can significantly predict a binary outcome. In this case, the measure was each student’s rate of absenteeism and the binary outcome was whether they performed at grade level in math and reading on the next scheduled state test, and in tests administered in subsequent grades.
Wu likened the statistical model to programming a smoke detector so that it’s sensitive enough to predict a fire early enough to intervene but not so overly sensitive that it signals frequent false alarms.
“We wanted to know how well a given attendance measure correctly flags students who are truly at academic risk without flagging students who are not,” she said.
Some key findings:
- Unexcused absences are a stronger predictor of academic achievement in 3rd grade and beyond than they are in early elementary school. That may be in part because younger students’ unexcused absences are more frequently caused by adult factors, like a family emergency, but students in older grades are more likely to play hookey because they are disengaged, researchers speculated.
- The correlation between absences and below-grade-level performance grows stronger as students get older.
- Unexcused absences were more predictive of failure in English language arts than overall absenteeism rates, perhaps because unexcused absences are often linked to disengagement or disinterest.
- In math, total absences, rather than just unexcused ones, were the most predictive of later performance on state tests. Because math lessons build on each other sequentially, even an engaged student who missed school for an excused reason like illness may struggle to catch up upon return, researchers theorized.
Taken together, the findings show that “no single absence cutoff is likely to serve as a strong standalone predictor of academic risk, underscoring the importance of integrating attendance with other measures” in early warning systems, the study says. Early warning systems are strategies through which districts monitor student data and flag indicators that students may drop out of school or struggle academically so that educators can intervene early.
The report’s authors offered cautions, noting that the findings aren’t conclusive. The findings relied on data from a large, urban district, and analyses of rural or suburban districts may show different patterns. Additionally, the predictive model used in the study—which is based on yes-or-no results—could make some inaccurate predictions.
Wu hopes other researchers will validate the report’s findings by replicating them across additional districts.
It may not be practical for districts to adopt measures of chronic absenteeism that are more strict or complex than the current 10% attendance threshold, Wu said. But understanding the nuances in smaller numbers of absences affect academic performance could help strengthen early warning systems, she said. And more precise understanding of those relationships could also help inform attendance messaging for parents, a key strategy for preventing problematic rates of absences, Wu said.
“Our findings imply that substantially lower thresholds would better identify students at risk of poor academic performance in a more timely way while balancing sensitivity and specificity,” the study concluded.