One-to-One Tutoring by Teachers (TT)
(Such as Reading Recovery)
The review showed that one-to-one tutoring by teachers works, and that teachers are more effective as tutors than paraprofessionals or volunteers.
Effect size = +0.38 in 19 studies
Reading Recovery: Effect size = +0.23 in 8 studies
Other programs: Effect size = +0.60 in 11 studies
One-to-One Tutoring by Paraprofessionals and volunteers (T-PV)
(Such as Sound Partners, SMART, and Book Buddies )
However, the review also showed that the effects seen for paraprofessional and volunteer tutors using structured and intensive programmes were positive, and pose a real challenge to the idea that only qualified teachers can be effective tutors.
Effect size = +0.24 in 18 studies
Paraprofessionals: Effect size = +0.38 in 11 studies
Volunteers: Effect size = +0.16 in 7 studies
Small Group Tutorials (SGT)
(Such as Corrective Reading, Quick Reads, Voyager Passport, and Empower Reading)
Small Group Tutorials are the most common form of supplementary instruction for struggling readers.However, the studies of SGT that were included in this review (because they met the inclusion criteria) were not ‘run-of-the-mill’, but named programmes with extensive training, materials, and a strong emphasis on phonics.
Effect size = +0.38 in 11 studies.
Classroom Teaching Process Approaches (CTP)
The results for classroom teaching process programmes were very positive for pupils the lowest performance levels in their classes. The effect size was similar to the findings for one-to-one phonetic tutoring.
Effect size= +0.56 in 16 studies
Cooperative Learning: Effect size =+0.58 in 8 studies
Classroom Teaching Process Programmes with Tutoring (CTP + TT)
The effect size for the lowest achievers in Success for All similar to the effect size for phonetic tutoring programmes. Nb. Success for All was the only programme in this section.
Effect size = +0.52 in 9 studies.
Teaching Technology (IT)
IT had minimal impacts on the achievement of struggling readers.
Effect size = +0.09 in 14 studies.
*Overall, 96 experimental-control comparisons met the inclusion criteria, of which 39 used random assignment to treatments, and five used randomised quasi-experiments. Effect sizes ( experimental-control differences as a proportion of a standard deviation) ) were averaged across studies, weighting by study sample size.