Schools serving similar students from similar neighborhoods had substantially different absence rates, with over half of school variation unexplained by student backgrounds or prior absences¶

Claim · All claims
Evidence · 1 study (2 entries) · associational r3 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2031: baselines and conditions change

Subclaims¶

q3 i? Adjusting for neighborhood barely changed school differences (high schools: 9.2 to 8.7 points post-pandemic), suggesting school attended mattered more than neighborhood. → Allensworth 2026 q3 i? After controlling for student backgrounds and prior attendance, high schools at the 75th percentile had absence rates 5.9 points higher than schools at the 25th percentile post-pandemic; middle grades differed by 1.8 points. → Allensworth 2026 (2)

Evidence¶

Allensworth 2026¶

Allensworth, E.M., Antony, M., Delgado, W., & de la Torre, M. (2026). Connection, trust, and learning: Student attendance in the middle and high school grades following the COVID-19 pandemic. Chicago, IL: University of Chicago Consortium on School Research. https://consortium.uchicago.edu/publications/connection-trust-and-learning

q2 · i? · associational · r3

Cross-nested HLMs with students nested in schools and census tracts, using CPS data for grades 6-11. High school 75th vs. 25th percentile differences were 8.7 points post-pandemic with neighborhood adjustment versus 9.2 without.

"School variation did not change much when we took into account the neighborhood where students resided, either before or after the pandemic. This suggests the school students attended affected their attendance much more than the neighborhood in which they lived."

Allensworth 2026 (2)¶

Allensworth, E.M., Antony, M., Delgado, W., & de la Torre, M. (2026). Connection, trust, and learning: Student attendance in the middle and high school grades following the COVID-19 pandemic. Chicago, IL: University of Chicago Consortium on School Research. https://consortium.uchicago.edu/publications/connection-trust-and-learning

q2 · i? · associational · r3

Cross-nested HLMs controlling demographics, prior attendance and test scores (grades 3-5 or 6-8). The brief notes this likely underestimates school effects in middle grades because K-8 schools' earlier-grade strategies are controlled out.

"At the high school level, schools at the 75th percentile had absence rates that were 5.9 points higher than schools at the 25th percentile, after controlling for their students’ absences in prior years and backgrounds. That’s about 10 days of school."

Discussion¶

Related Claims¶

• Students with the same absence rates had higher GPAs (and in grades 7-8, higher test scores) in post-pandemic years than pre-pandemic years — related
• Ninth-grade GPAs varied substantially by gender, race/ethnicity, neighborhood SES, incoming test scores, course level, and school — related
• Outcome differences between CHSRI new-start and redesigned schools were not statistically significant after accounting for incoming student differences — related
• Graduation rates differ largely across Chicago schools, with many new charter and magnet schools and several neighborhood high schools showing exceptionally high rates beyond what incoming freshman characteristics would predict — related
• Multiple dimensions of school climate were associated with lower school absences, with stronger relationships after the pandemic than before — related
• Schools varied considerably in how much their absence rates increased after the pandemic, with some showing little or no change — related
• Students' perceptions of their classes and schools matter more for attendance in post-pandemic years than before the pandemic — related
• Ninth-grade course failure rates varied substantially across CPS high schools even after controlling for demographics and prior achievement, and student-teacher relationship measures were associated with better-than-expected Freshman OnTrack rates — related
• No-credit rates during the pandemic varied considerably across elementary schools, including among schools serving similar student populations — related