Repeating third grade helped, not hurt, the Indiana students who were held back for weak reading scores. They outscored comparable classmates who had been promoted, and their reading advantage was still there four years later, with no rise in suspensions or missed school.
The study, by NaYoung Hwang and Cory Koedel, was published online on October 11, 2023 in Educational Evaluation and Policy Analysis.
A test score cutoff as a natural experiment
From the 2011 to 2012 school year through 2016 to 2017, Indiana required third graders to pass a reading test called IREAD-3 before moving on to fourth grade. Students scoring below 446 points had to repeat the year unless they were exempt, as English learners and students with disabilities were.
Children who failed the spring test got intensive help and a second try in summer, and that summer score decided promotion. Over 96% of those who failed in spring took the retest. Across the cohorts studied, about 1.8% of Indiana third graders were held back, a much lower rate than in Florida or Chicago.
The researchers used state administrative data on five cohorts of third graders. They compared students who scored just below the cutoff with those who scored just above it, a regression discontinuity design. Because those children were nearly identical in ability, any later gap can be attributed to being held back. They then followed outcomes for up to five years.
Big gains that faded but did not vanish
Compared at the same grade level, retained students scored more than 0.50 standard deviations higher than their narrowly promoted peers on the fourth grade state tests in both English and math. The authors note these short-term effects closely match earlier estimates from Florida and are larger than those from Chicago.
The advantage shrank over time. By seventh grade, retained students still outperformed in English by just over 0.20 standard deviations. In math the seventh grade estimate was similar in size but not statistically distinguishable from zero.
The fourth grade gains appeared across groups defined by gender, race and ethnicity, and eligibility for free or reduced-price lunch. The results held when the team changed the statistical model, the width of the score window around the cutoff, or added school and cohort controls.
No sign of behavioural fallout
A common worry about holding children back is stigma, demotivation and trouble at school. The study found no evidence that retention changed suspensions, expulsions or unexcused absences in any grade through middle school.
The authors are frank that these behavioural estimates are imprecise and could only rule out large effects. Still, the estimates pointed in no consistent direction, which they read as most consistent with no impact.
The results apply to students near the cutoff, not to every child who is held back. They add to a pattern in the research: retention in early grades tends to look positive, while most of the negative findings come from students held back in sixth grade or later.
Study Details:
- Title: Helping or Hurting: The Effects of Retention in the Third Grade on Student Outcomes
- Authors: NaYoung Hwang, Cory Koedel
- Journal: Educational Evaluation and Policy Analysis
- Publication Date: October 11, 2023
- DOI: 10.3102/01623737231197639