Students’ third-grade reading proficiency is a common early indicator of later success. To ensure all students meet this critical milestone, schools need to identify who is off-track and tailor effective interventions as soon as possible.
SDP fellows in Rochester, Tulsa and Washington, D.C. gathered and analyzed student characteristics and achievement data in preschool and grades 1 and 2 to find common predictors of third-grade literacy across geographic differences.
The fellows found that testing in the early years was predictive of third-grade reading, and that attendance and exposure to preschool also had positive effects. They suggested funding and policy priorities for districts tackling this issue and informed pilot interventions at the school level.
Keeping Pace with Early Expectations
Early-warning systems based on students’ academic performance are common in the upper grades. But for the youngest students, it’s more complicated. Not only is most work untested in preschool and grades 1 and 2, but children are also following individual developmental trajectories during those early years, with wide variation in relative age, maturity, and social and emotional skills.
The unique growth that happens in early elementary school has raised questions about the appropriateness of standard tests and benchmarks. However, because third-grade literacy skills are a critical indicator of student outcomes, districts need to measure progress among students—and intervene as needed—before they reach that milestone.
Field-Testing Earlier Early Warning Systems
SDP fellows led efforts to identify struggling young readers in three districts: Rochester, NY; Tulsa, OK; and Washington, D.C.
In Rochester, boosting students’ early elementary literacy levels is a top priority. SDP Fellows Jing Che, Patty Malgieri and Vicky Ramos created data profiles of individual third-grade students at two elementary schools, including their demographics, attendance, behavior, and academic and developmental records in preschool and grades 1 and 2. They then compared each factor and its relationship to student characteristics like gender and class against students’ overall reading success in third grade, in order to determine which in-school factors were most predictive of on-time literacy by third grade.
The fellows found that students’ performance on benchmark tests in reading and math in second grade, as well as second-grade attendance, were predictive of third-grade reading proficiency. They also found that preschool participation was associated with better kindergarten readiness, attendance, and performance on benchmark tests in first and second grades. Importantly, they also discovered that students who had repeated a school year had worse attendance and lower scores on benchmark tests.
In Tulsa, state law requires third-grade students who fail a statewide reading test to repeat the grade. SDP Fellow Anna Holt investigated the validity of the third-grade Oklahoma Common Core Curriculum Test (OCCT) and analyzed a rich array of student data to determine the performance trajectories of students in grades 2 and 3 and discover the key identifiers that predicted who was most at risk of being held back.
She found that student performance on an interim computer-adaptive test that measures academic growth was a significant and reliable predictor—but that its administration in the fall of third grade was too late, because students were nearly two years behind by then. She also found that absenteeism predicted low performance, but low grades in the early years did not.
In Washington, D.C., which has the highest public preschool participation rate in the country, SPF Fellow Hannah Page studied results from a kindergarten skills assessment to determine the impact of attending one or two years of quality preschool. She also analyzed student-characteristic data to determine which students were most likely to fall behind in their early elementary years. She found that students who attended two years of preschool did better, overall.
The analyses by the SDP fellows revealed important trends in their home districts, and shed new light on a common issue when reviewed alongside one another.
Fellows in both Rochester and Tulsa found that benchmark tests in the younger years were predictive of third-grade performance. In Rochester, this created a new school-level opportunity for educators to act. In subsequent years, they were planned to develop new interventions and continue to review and refine the student-level data in order to determine how early in a student’s career such interventions should occur. The Rochester fellows also recommended that the district take a close look at its retention policy—while repeating a school year ostensibly gives students time to catch up, their analysis found that retained students continued to do poorly.
The Tulsa analysis informed a new, real-time data dashboard provided to elementary teachers to track student performance on ongoing assessments. In a pilot year, four out of the 10 schools that used these data decreased the number of students at risk for retention by about 40 percent. The fellows noted that this information was also used in efforts to improve instruction. And in Washington, D.C., the SDP fellows’ analysis uncovered the need for a standards-aligned assessment with the predictive power that fellows found in Rochester and Tulsa, so officials can develop a districtwide on-track indicator for young students.
Overall, the studies showed how powerful early testing can be in terms of identifying children likely to struggle in third grade. And they showed the importance of attendance in the early years, suggesting that schools looking to improve can first focus on stressing the importance of consistent attendance to families.
Read SDP Fellows' capstone report here.