Evaluating the Efficacy of In-School Tutoring Interventions to Promote Math and Reading Achievement

• In Event: Supporting Students Through Educational Disruptions: State and District Strategies for Post-Pandemic Education Renewal

Abstract

Objectives and Perspective
Individualized and small group tutoring has been widely recommended to accelerate achievement growth among K—12 students in the wake of the COVID-19 pandemic. However, tutoring interventions vary widely by their modality, student-tutor ratio, tutor background, targeted student population, and dosage. This paper evaluates the efficacy of two tutoring interventions with different characteristics that targeted students in Grades K—9 in a large school district in the Atlanta metro area.
One intervention used high-dosage, small-group tutoring with a student-tutor ratio of 4:1. The program offered tutoring in math and reading to students in Grades K—9. Tutoring was conducted by district staff or external vendors. A second intervention used AmeriCorps tutors and targeted students in Grades 4—8 for math and Grades K—3 for reading, where sessions involved two to three students per tutor.

Data and Methodology
This paper considers the following research question: How effective were two tutoring interventions in promoting student achievement in a metro Atlanta district? We use detailed student-level administrative data supplied by the school district. The data include students’ demographic information, enrollment, and test scores. The district assigned course numbers to the various types of tutoring offered, providing a binary measure of tutoring participation at the program level. To determine the intensity of tutoring participation, we use tutoring roster dates, instructional days between the relevant exams, and session-level tutoring information. We use fall and winter formative assessment scores from school year 2023-24 to measure student achievement growth.
The methods used to estimate the effectiveness of the tutoring interventions include multivariate regression models, school and grade fixed effects, and a regression discontinuity design. The RDD is feasible because, for the first intervention, formative and summative assessment scores partially predicted tutoring participation. Likewise, for the AmeriCorps intervention, students were targeted for the program if they fell below a particular achievement level.

Results
The primary outcome of interest is achievement gain per instructional day between the fall and winter formative assessments. Participating in math tutoring is not associated with a statistically significant gain in achievement. Participating in reading tutoring for an entire academic year is associated with a positive and statistically significant gain of 0.2 standard deviations in reading achievement, holding baseline achievement levels and student demographic characteristics constant. The results hold when making within-school and within-grade comparisons using school and grade fixed effects, respectively.

Scientific or Scholarly Significance
The results of this study can aid school administrators in designing and implementing future tutoring interventions. The information on the relative efficacy of differing tutoring interventions is particularly relevant as many students continue to make up for pandemic-era achievement losses and federal support for acceleration initiatives dwindles. The tutoring programs in this study were not limited to students at the lowest end of the achievement distribution; middle achievers could participate as well, thereby expanding the generalizability of the results to a broader student population.

Authors

• Salma Mohammad Ali, Georgia State University

• Thomas Goldring, Georgia State University

• Tim Sass, Georgia State University