Reading recovery: a longitudinal meta-analysis

Abstract

Abstract Reading Recovery(RR) is a constructivist reading intervention used to provide tier 3 instruction to struggling readers in the first grade. The program has been previously evaluated and found effective by Evidence for ESSA (John Hopkins University), What Works Clearing House (intervention report institute for education sciences 2013), and in a meta-analysis by D’Agostino et al. (J Educ Stud Placed Risk 21:29–46, 2016) However, the National Reading Panel (United States Government, 2000), showed some conflicting findings. Moreover, May et al. (CPRE Research Reports, 2016), suggested that RR might be detrimental over the long term, for student reading outcomes. This meta-analysis examined 19 experimental and quasiexperimental studies to evaluate the efficacy of RR over the short and long term. Cohen’s d, effect sizes were calculated by subtracting the mean difference between the treatment groups and controls at post-test, then dividing by the pooled standard deviation. Effect sizes were then weighted by their inverse variance to account for sample size. For assessments taken within the assessment year, the meta-analysis showed a mean overall effect size of .19, a weighted mean effect size of .05, and 95% confidence intervals of = [-.16, 54.] For assessments taken more than 1 year after the intervention, the meta-analysis showed a mean negative effect size of -.14 and 95% confidence intervals of = [-.59, .31], with a weighted effect size of -.21. These results suggest that RR may not currently be the most effective approach, for literacy intervention.