Research Guidance

FEV Tutor further evolved last year when it announced a new AI-enhanced platform, Tutor CoPilot. The tool makes tutors more effective by giving them guiding questions to ask students. In a randomized trial, the National Student Support Accelerator at Stanford University, which studies tutoring models, found that when less-experienced tutors used the AI support, student math scores increased an average of 9 percentage points. 

Over the past couple of years, scaling well-researched solutions has been shown to also counter the negative effects of the pandemic, Loeb said, from widening achievement gaps and missed school time, to poorer social and emotional development. Her team recently launched the National Student Support Accelerator (NSSA) to address educational inequities resulting from the pandemic. NSSA conducts research on the most promising tutoring practices and works with district leaders and others to provide research-backed guidance on implementing high-impact tutoring. 

“Our students deserve this work,” Loeb said. “From our research, we learn so much about how to engage students and accelerate their learning. The practical, easy-to-use learnings from research need to reach decision makers so that our students can benefit.”

Tips by Text and NSSA are part of SCALE, Loeb’s new initiative at the Stanford Accelerator for Learning, a university-wide effort addressing some of the most challenging issues in education through research, partnerships, and technological innovation. 

In the past five years, the Texas district’s investments in staffing and high dosage tutoring are paying off.

That’s why the district piloted a virtual tutoring program in the 2020-21 school year. Middle school students were the first to participate. In spring 2021, the live virtual tutoring program expanded to serve 6,000 students in K-12, he said.

That same spring, Ector County ISD designated $10 million of its $93 million in federal pandemic relief funds to tutoring over the next three years, according to a report by university-based research nonprofits FutureEd and the National Student Support Accelerator.

Generative AI, including Language Models (LMs), holds the promise to reshape key sectors like education, healthcare, and law, which rely heavily on skilled professionals to navigate complex responsibilities. In education, for instance, effective teacher training with expert feedback is crucial yet costly, limiting opportunities to enhance educational quality on a larger scale.

High-impact tutoring has the strongest evidence base of any approach for improving student learning, and contributes to increased engagement and attendance. As far as proven education solutions go, it’s a pretty darn good one, and has rightfully been a bipartisan priority since the pandemic. 

But federal pandemic relief money that helped fuel the expansion of such programs dried up in September, and recent research has sparked debates about the high-impact tutoring’s effectiveness when implemented at scale. This includes an evaluation of Metro Nashville Public Schools’ tutoring program that reported small gains for students and a meta-analysis of large high-impact tutoring programs that showed challenges in maintaining evidence-based practices. 

Research reveals the most effective ways to help young struggling readers through tutoring.

Tutoring has gained popularity as a strategy to improve the academic achievement of struggling students. Intensive, relationship-based tutoring is a highly effective academic support for many students, particularly in the early elementary years when school schedules and classroom routines are flexible (Groom-Thomas et al., 2023). For schools considering how to begin tutoring or where to prioritize resources, early literacy tutoring — which is both effective and feasible — is a good place to start.