In the almost 40 years since Ronald Reagan’s A Nation at Risk report and subsequent education reforms, many programs have been developed to help students improve their academic achievement. Some of these programs included forms of tutoring or academic coaching as a way to increase academic achievement.
These programs by and large had mixed adoption, and mixed success, for a number of reasons, the foremost being No Child Left Behind’s Supplementary Education Services which saw many schools and districts trying to work with outside agencies and partners for tutoring that became logistical and legal nightmares when trying to work with minors. And while the many prior programs overall may not have achieved their goals, tutoring still remains an effective support strategy when implemented correctly, which will be discussed later in this blog.
Since the late 1970s, research has demonstrated that 1:1 human tutoring (online or in person) can have a wide variety of benefits for students, including gains in academic achievement, improved problem solving and critical thinking skills, better conceptual understanding, and growth in socioemotional learning components (Chi, Siler, Jeong, Yamauchi, & Hausmann, 2001; Cohen, Kulik, & Kulik, 1982; Kraft & Falken, 2021; VanLehn, 2011; Wood, Bruner, & Ross, 1976). Gains are especially evident for low achieving students in the areas of math and reading (Chappell, 2009).
As technology has grown more advanced, researchers and developers have explored using it for a variety of educational applications, including tutoring. The most promising of these types of technologies are the intelligent tutoring systems or ITS.
ITSs are self-paced, learner-led, highly personalized, and adaptive learning environments typically operated through a computer. These types of tutoring systems make use of two loops that select topics and tasks for learners based on their inputs (Steenbergen-Hu & Cooper, 2013). These tutoring systems are typically most effective in settings that pose questions with definite solutions such as math, or development of a specific skill in a corporate or industrial environment.
The short answer is, it depends on the context.
In recent years, because of advances in ITS, in some cases it has been shown to be almost as effective as a human tutor (VanLehn, 2011). Most experts, however, still recommend ITS as a scaffold to assist teaching and learning and not as a tool to use low achieving students to ‘close the gap.’
This is because even in an intelligent and adaptive program, it is difficult to assess true conceptual understanding, critical thinking, and problem solving skills that are essential for learning (Mark & Greer, 1993). Again, most research regarding the effectiveness of ITS is largely within the areas of mathematics or other areas that lend themselves to a program that can easily learn from a student’s performance on defined tasks and is not yet suitable for more constructivist approaches like project-based learning (Jonassen,1996; Kulik & Fletcher, 2016).
Over time, human tutoring has been consistently shown to produce larger effect sizes than other forms of tutoring, including ITS, regarding achievement and regardless of subject area, and is associated with other positive learning outcomes outside of achievement as well (Chappell, 2009; Chappell, Arnold, Nunnery,& Grant, 2015; Cohen, Kulik, & Kulik, 1982; Kulik & Fletcher, 2016). Reading and problem solving tutoring have been found to be especially effective. This is because human tutors are incredibly sensitive to other human patterns and emotions and can give students valuable feedback in a variety of contexts to help individualize the learning process.
So what makes a human tutor or tutoring program effective at what they do?
Here are the best practices from the literature:
The most effective tutors are highly qualified (have 4 year degrees) and are trained in a specific technique or curriculum for helping students.
In the most successful programs, tutoring looks more akin to what we would call academic coaching. Tutors meet with students in very small ratios (1:4 or less) on a consistent basis for an extended period of time (Kraft & Falken, 2021). This allows the tutors and coaches to develop relationships with students and better connect with their students’ lives and prior knowledge to create deeper learning experiences.
Programs that have outsourced tutors and coaches rather than embed them in the student’s regular cadence have had much less success than programs that use tutors that students are familiar with, and with whom students have a personal connection. (Chappell, 2009; Kraft & Falken 2021).
So, knowing what we know about tutoring and coaching, what could we do in our schools and classrooms given the challenges that we all face in the wake of one of the largest disruptions to traditional education in the 21st century?
Make tutors or coaches a regular part of the schedule with students similar to or actually as our RTI (Response to Intervention) or MTSS (Multi-tiered Systems of Support) time. Because most of us are online, we can take advantage of small windows of availability to meet with students at a regular cadence (daily, semi daily, or weekly intervals).
Ensure tutors are highly qualified and use a type of technique or curriculum that is research-based, possibly partnering with teachers in training from local colleges and universities.
Don’t push students off to tutoring websites that charge by the minute or use chatbots as solutions-- make it part of their regular routine for maximum effectiveness.
Research shows that tutoring, when implemented correctly, can have a plethora of positive effects, especially--but not limited to--low achieving students. It can be something that we can use to support students from slipping down the slope of learning loss that we all see approaching due to the challenges we are currently facing as educators. Not only that, but recent surveys amongst over 1,000 teachers, principals, and district administrators regarding learning loss among students due to the COVID-19 Pandemic present human tutoring and coaching as a highly desirable method for implementation due to its demonstrated effectiveness (Cavanaugh, 2021).
Coming Soon
Cuccolo & Green’s (2025) report highlighted the relationship between students’ assignment submission patterns and final course scores. Given that pacing has important implications for student performance, knowing what assignment submission patterns look like across schools with varying demographics could help prompt early identification and intervention. As such, this blog explores students’ assignment submission patterns based on school-level demographic information.
Explore how immersive VR simulations helped students step into real-world roles: from EMTs to chefs, all without leaving the classroom.
In this blog, MVLRI researchers synthesize the key findings from two research studies about student assignment submission patterns in Michigan Virtual online courses.
In this interview, MVLRI researchers discuss key findings from a report highlighting how personalized, consistent, and timely communication in online courses can help students feel more connected to their online teachers and may also impact their success in the course. This blog also explores practical strategies for communicating effectively and building relationships with online students.
This blog digs into the key findings from two MVLRI research studies exploring educator engagement with professional learning (PL), their beliefs about implementing what they’ve learned, and insights into continuing to tailor PL to meet educators’ needs.