Schwartz, R. M., Schmitt, M. C., & Lose, M. K. (2012). The Elementary School Journal, 112(4), 547-567.

Background
This study used a randomized experimental design to examine the relationship between teacher-student ratio and literacy learning outcomes for experienced intervention teachers working with the most at-risk first-grade students. Eighty-five Reading Recovery teachers, working with 170 students, each taught in a 1:1 and a small group instructional format with teacher-student ratios of 1:2, 1:3, or 1:5. The at-risk students were assessed at pretest and posttest with the six subtests of An Observation Survey of Early Literacy Achievement (Clay, 2002, 2005), the Slosson Oral Reading Test-Revised (SORT-R), and two spelling measures.

Findings
The 1:1 instruction yielded significantly higher outcomes than the combined small group conditions on eight of the nine measures. The small group conditions did not differ significantly from one another, but a trend analysis indicated a reduction of literacy performance as group size increased.

To evaluate acceleration across the ratio conditions, the percentages of children reading at Level 10 or above at the end of the intervention period were identified. For the 1:1 teacher-student ratio instructional context, 61% of the students were reading at Level 10 or above. For the other treatments, the percentages are 38%, 26%, and 19%, respectively, for the 1:2, 1:3, and 1:5 groups. Based on this analysis, it appears that the 1:1 treatment is the only condition that reduces the percentage of students who are at risk by reducing the gap between these initially low-performing students and their average peers.

The authors conclude that a mix of individual and small group services should be sufficient to reduce the achievement gap across first grade for 70– 80% of the students who would struggle to make progress in the classroom context alone. This requires careful monitoring of outcomes at the end of first grade and, if necessary, action to implement more-intensive and effective systems of early intervention services. Using local data in this way, schools and districts may find it necessary and efficacious to invest in additional professional development and staffing to support early intervention services. The question is not whether individual, small group, or classroom instruction is most effective; it is clear that all are essential. Rather, an RTI approach should focus on how best to achieve optimum literacy outcomes for all learners in a timely manner and based on their individual needs.