The last study is the one that ties the others together. A randomized controlled trial written up in an EdWorkingPaper released in November 2025, "Beyond the Classroom: Impact of a High-Dosage Tutoring Program on Student Literacy Achievement" by Ayman Shakeel, followed 333 students in grades 2 to 5 through a high-dosage literacy program with low tutor-to-student ratios, delivered during the school day (EdWorkingPaper, 2025).
The overall average effect was not statistically significant, which, read quickly, looks like the program did not work. But the average hid the real story. Students who started further behind gained meaningfully, roughly 0.2 of a standard deviation, while students who were already reading well saw no benefit and, in some cases, slipped slightly. The tutoring helped exactly the children it was meant to help, and the average washed that out by mixing them together.