This research report presents the results from the second year of a randomized controlled trial of an early elementary reading tutoring program that has been designed to be affordable at scale. During the 2021-22 school year, over eight hundred kindergarten students in a large Southeastern school district were randomly assigned to receive supplementary tutoring with the Chapter One program. The program continued during the 2022-23 school year, while the children attended first grade.

Contracting relationships between public school districts and vendors are a common feature of education provision in the United States. Contracted services in schools can range from broad, essential functions such as school meals, bussing, and janitorial services to more specialized services such as the analysis of student data, curriculum mapping, and professional development for staff members. The strength of these contracting relationships depends on vendors providing consistent services and on payment between vendors and districts.

During the 2022-23 school year, Try Once, Inc. (“Once”) partnered with a large, urban school district on the East Coast to provide high-impact early literacy tutoring to 105 kindergarten and first grade students in 13 schools. The district identified students as eligible for tutoring services if they scored below grade-level benchmarks in their early literacy skills. The Stanford research team randomly assigned eligible students into a tutoring program group (n=105) and a comparison group (n=199).

Research shows that tutoring, especially when delivered intensively and with personalized support, is one of the most effective interventions for improving math and reading outcomes, particularly for students from low-income backgrounds. Successful tutoring programs are characterized by strong tutor-student relationships, frequent sessions, small group sizes (no more than three students per tutor), high-quality materials aligned with the curriculum, and data-informed practices.

Artificial intelligence (AI) applications to support human tutoring have potential to significantly improve learning outcomes, but engagement issues persist, especially among students from low-income backgrounds. We introduce an AI-assisted tutoring model that combines human and AI tutoring and hypothesize this synergy will have positive impacts on learning processes.

Although learners are being connected 1:1 with instructors at an increasing scale, most of these instructors do not receive effective, consistent feedback to help them improved. We deployed M-Powering Teachers, an automated tool based on natural language processing to give instructors feedback on dialogic instructional practices —including their uptake of student contributions, talk time and questioning practices — in a 1:1 online learning context.