While the existing evidence suggests high-impact tutoring interventions consistently produce positive learning gains for students, there is ample opportunity for future research to deepen our understanding of what works and why. The Research Agenda Moving Forward section in our Research Agenda strives to advance knowledge by identifying worthwhile research questions — those that will lead to answers that could help policymakers and practitioners effectively implement tutoring initiatives.
Moving forward, we hope scholars, practitioners, and policymakers will push forward on two, complementary lines of research, both of which have a goal of understanding how best to deliver high-impact tutoring to students in need across the U.S. First, we aim to understand what makes tutoring interventions effective. This research will uncover specific characteristics of effective tutoring programs. Second, we introduce research questions regarding how to implement tutoring effectively at scale. This research line articulates supply, demand, and implementation questions; new knowledge generated by this research will be integral to scaling high-impact tutoring.
Research Priority #1: Identifying the Characteristics of Effective Tutoring
The Accelerator has identified several promising areas for research regarding what characteristics comprise high-impact tutoring. Our priority research questions, outlined in this document, focus on:
- The cost-effectiveness of different tutoring models;
- Tutor skills;
- Tutor-student relationships;
- Adoption by stakeholders;
- How programs increase engagement and fidelity; and
- How best to extend the research focus.
Below, we provide a brief overview of our “Research Agenda Moving Forward” section, contemplating how future studies might evaluate these specific areas of interest.
First, evaluating the characteristics that drive tutoring program costs will be critical to scaling so that all students experiencing poverty can have access to high-impact tutoring. With a goal of determining how to deliver tutoring in the most cost-effective way, we propose conducting evaluations that intentionally vary: student-tutor ratio; time students spend with tutors each week; whether tutoring is delivered in-person, virtually, or through a combination of the two (i.e., blended); and the use of computer-assisted learning technologies. Additionally, we need to study the amount of training each tutor type requires to boost student outcomes.
Second, understanding what skills and supports tutors will need to be successful will contribute to evaluating both the costs and quality of the tutoring. In addition to assessing tutor type and training, we propose studying what recruitment policies are most effective, what forms of ongoing support are most beneficial for tutors and students, and how training and support can ensure quality and safety.
Third, high-impact tutoring involves human interaction between a tutor and a student. Many programs state that cultivating positive tutor-student relationships is a critical feature. Therefore, understanding how to build strong relationships and how the relational component of tutoring impacts student learning is important for evaluating and learning to create quality. Potential opportunities to explore the import of tutor-student relationships could involve: testing the importance of having the same tutor; intentionally embedding relationship-building content in tutoring programs; matching students with tutors based on their traits; and “looping” students and tutors such that students learn from the same tutor over the course of multiple years (not just one).
Fourth, successfully scaling high-impact tutoring requires buy-in and adoption from some combination of districts, schools, educators, parents, and students. Research efforts focused on understanding how programs facilitate or hinder program adoption can provide necessary information for scaling. These efforts might explore the impact of: need-driven vs. universal tutoring; opt-in vs. opt-out enrollment; different recruitment strategies; during-school vs. out-of-school time tutoring; and curricular alignment.
Fifth, once tutoring is offered, schools, teachers, parents, and students need to actively engage with the program. For tutoring to be successful, we need to consider the types of practices that lead to increased engagement among various stakeholders. We propose testing levers for promoting engagement, including: school-based coordinators; tutor-teacher interaction; and family involvement.
Finally, there is great opportunity to learn more about the multidimensional benefits of tutoring. While our agenda focuses on student learning outcomes, high-impact tutoring has the potential to affect a wide range of student outcomes. For instance, in addition to tested outcomes, tutoring could also result in improvements in student grades, motivation, relationships, advanced course-taking, attendance, behaviors, and school completion. By including additional measures when assessing tutoring program characteristics, we can learn how and why tutoring leads to student success. We further need to consider the long-term outcomes associated with tutoring, and if they vary by tutoring program characteristics. Also, tutoring may have benefits for the tutors. Tutoring may lead tutors to complete more school, enter teaching, engage more positively with their community, or improve on a range of other outcomes. Tutoring may also reduce unemployment in the short run and lead to more consistent employment in the longer run.
These focus areas cover just a handful of the promising questions for researchers to address regarding high-impact tutoring. By no means is this list comprehensive. We hope educational researchers across disciplines will unite to determine how high-impact tutoring can lead to more equitable outcomes for the students who need it most.
Research Priority #2: Implementing Tutoring at Scale
Our second research priority, Implementing Tutoring at Scale, broadly address issues such as: Where will the supply of tutors come from? What is the demand for high-impact tutoring? And how are tutoring programs most successfully implemented?
The goal of “Implementing Tutoring at Scale” section is to identify questions that guide research on how practitioners and policymakers will implement high-impact tutoring
For instance, to implement tutoring at scale, we need to understand (1) where the supply of tutors would come from and (2) the extent to which districts want and are capable of implementing tutoring initiatives. We need answers to questions such as:
- What skills do tutors need?
- When are tutors available?
- How can the tutoring be delivered?
- What types of supervision do tutors need?
- Who can be tutors?
- What is the best way to recruit tutors?
- How many students can a tutor handle?
- How much do tutors cost?
- Where is the money for tutoring coming from?
- What are the constituencies that might want high dosage tutoring?
- What stakeholders at different levels need to be brought in to make high-dosage tutoring happen?
- How does state or district governance shape demand for tutoring?
- Which students receive tutoring?
Simultaneously, we have developed a line of implementation-related questions that will guide researchers in building a deeper understanding of how district leaders, school leaders, teachers, parents, and students experience a tutoring initiative. Each question addresses either the motivations, opportunities, logistical needs, or barriers for delivering tutoring to students. Answering questions at these multiple stakeholder levels will ultimately help answer the overarching question: What are the necessary conditions required to effectively implement a high-dosage, school-driven tutoring program?