Does Tutoring Work?
Is tutoring effective? And if so, how?
Conventional wisdom—and collective market forces—have answered the first question in the affirmative. The second is not so easily answered, though a recent meta-analysis sheds some light.
In the early 2000s, tutoring was on the rise
Flash back to 2007, when families/school systems collectively spent between $8 and $12 billion per year on tutoring.[1] Taking stock of the beginning of this shift in education was the now out-of-print The Tutoring Revolution: Applying Research for Best Practices, Policy Implications, and Student Achievement. Four academics summarized wide-ranging studies in what they correctly noted was the rise of tutoring—a “revolution,” as they claimed. They wanted to understand what educational policy decisions were affecting growth in the tutoring sector and what scientific research and education theory might reveal about how best to proceed.
“The literature clearly supports the view that there is no single best method of teaching,” they wrote. They then presented several different movements in education theory, including “process-product,” Constructivism, Functionalism, the Reconceptualized Curriculum Movement, among many other jargoned terms.
The authors also highlighted the need to differentiate between a tutor and a teacher: “A tutoring curriculum needs to be used that is significantly different than the teaching practices used on a daily basis. Teachers need to be trained in tutoring best practices [my italics] that are more individualized than their regular classroom teaching methods.” In what proved to be prescient, they theorized that tutoring might become “recognized as an education discipline and defined as a professional career [my italics].”
But Tutoring Revolution didn’t say with any certainty that tutoring is actually effective, nor did it propose any metrics to quantify its success: “…there are many hypotheses but very limited evidence-based findings. The details related to why certain tutoring structures yield positive results require additional investigation.” That was 2007.
By 2020, the tutoring market had matured
Enter the National Bureau of Economic Research’s (NBER) 2020 working paper, The Impressive Effects of Tutoring on PreK-12 Learning: A Systematic Review and Meta-Analysis of the Experimental Evidence.[2] Three researchers determined that tutoring was “one of the most versatile and potentially transformative educational tools in use today.” This conclusion was buoyed by the overall US tutoring market capitalization of $21.8 billion, almost triple what it had been thirteen years earlier.[3]
The NBER researchers confidently affirmed tutoring’s effectiveness: “We find that tutoring programs yield consistent and substantial positive impacts on learning outcomes, with an overall pooled effect size estimate of 0.37 SD.” (A standard deviation of 0.37 is a somewhat modest effect, but it’s statistically significant.)
Other key findings included:
1. Tutoring is more effective when conducted by a teacher or paraprofessional tutor, and not by a parent or friend.
2. Tutoring achieves better results in earlier grades.
3. Tutoring in reading leads to better outcomes in earlier grades, while math tutoring leads to better outcomes in later grades.
4. Tutoring during school has a larger impact than tutoring after school.
Note that these findings were published in 2020, before the Coronavirus pandemic severely disrupted the education system. Since 2020, there have been growing calls for schools to facilitate “high-intensity” tutoring as the most effective way to make up for learning loss.[4] The NBER paper also predates the Zoom Revolution, which normalized online delivery as an accepted instructional medium; remote tutoring has not yet been subjected to rigorous research.
The NBER paper lays out some basic ideas that may make tutoring effective. It calls them “Mechanisms of Impact”:
1. More instruction time.
2. Customized learning at the student’s ability level can help balance disparities observed in the classroom.
3. More engagement and quick feedback lead to increased focus and effort.
4. The relationship with a mentor motivates and inspires students to do their best.
5. Distinguishing between students who have fallen behind from those students with learning differences.
Many of these “mechanisms” could be considered obvious, but it is important to establish these common-sense foundations for any future research in the field. In particular, I want to highlight the importance of the fourth mechanism. The quality of the relationship between the tutor and the student can—in and of itself and irrespective of the specific subject matter—make a most significant difference—provided that the tutor is not the student’s parent; in such cases, I have found that tutoring can become a radioactive experience.
So does tutoring work? And, if so, how?
The clear answers to the two questions are: yes and we’re still figuring it out.
What seems evident is that the effectiveness of tutoring has little to do with education theory (as Tutoring Revolution had analyzed in depth). What it does reflect is the ability of one-on-one personal instruction to complement the classroom’s group-learning dynamic. (But Revolution was somewhat right: “Tutoring best practices” must be different from classroom teaching methods.)
One key difference in findings between 2007 and 2020 is that Revolution concluded that it was “desirable to conduct the tutoring in the student’s own home outside of school hours,” whereas NBER concluded that “tutoring programs conducted during school tend to have larger impacts than those conducted after school.” Some of the $122 billion of funding in the 2021 American Rescue Plan is now being spent on one-on-one or small-group interventions during the course of the school day, which can provide “high-impact tutoring” using evidence-based enrichment programs.[5]
A major caveat in this tutoring research is that the scope of both studies includes many different kinds of tutoring. The NBER paper, for example, only analyzes data from preschool through secondary level programs. As I have pointed out in other posts, the term “tutoring” encompasses many concepts, courses, and topics. Much of elementary school tutoring is aimed at helping young minds develop skills of literacy and numeracy, whereas middle and high school tutoring specializes in different academic domains (algebra, foreign language, biology, essay writing, etc.). Tutoring children in the “science of reading,” for example, is very different from tutoring them in trigonometry.
NBER’s conclusion that tutoring achieves better results in early grades highlights the importance of foundational knowledge and skills as the building blocks of more complex academic tasks. Because math and English instruction build on previous lessons, a student who has not achieved mastery of a particular topic (subtraction, for example) but who continues to meet grade-level requirements may have trouble with related topics (i.e., subtracting fractions) in later grades.
But neither study illuminates differences in outcomes between disadvantaged and privileged school populations. It may be that tutoring can improve results in wealthy classrooms where there is more funding and support, but not in underfunded classrooms lacking resources and adequate instruction. (The NBER paper explicitly says that research focusing on tutoring in lower grades promotes equity throughout the educational system.) Finally, the researchers do not indicate whether the improvements in tutoring only bring students up to grade level, or whether they help students surpass grade-level expectations.
But even without precise answers, the research is significant, if only as a way of making sure the vast investments in education will be spent effectively. As most studies do, the NBER paper points to the need for more specialized study, particularly of the impact of tutoring on older students and in more specialized subject matter, but it should bolster the efforts of policymakers to further integrate one-on-one tutoring into education.
It took a while for the science to catch up to the money, but now it has.
[1] Gordon, Edward E., et al, The Tutoring Revolution: Applying Research for Best Practices, Policy Implications, and Student Achievement (Lanham: Rowman & Littlefield Education, 2007), 27.
[2] https://www.nber.org/papers/w27476. Accessed November 29, 2023.
[3] https://blog.gitnux.com/us-tutoring-industry-statistics/. Accessed November 5, 2023.
[4] https://www.npr.org/transcripts/950814131. Accessed December 1, 2023.
[5] https://www.nytimes.com/2022/10/06/education/learning/tutoring-learning-loss.html. Accessed December 1, 2023.