Evidence for Learning: Phonics

The criteria used to judge the inclusion of studies in the Toolkit are:

• The population sampled involved early years and school age learners from 3-18 learning in their first language.
• The intervention or approach being tested was educational in nature, including named or clearly defined programs and recognisable approaches classifiable according to the Toolkit strand definitions (e.g. peer tutoring or small group teaching). The intervention or approach is undertaken in a normal educational setting or environment for the learners involved, such as an early years centre or school or a typical setting (e.g. an outdoor field centre or museum).
• A valid comparison was made between those receiving the educational intervention or approach and those not receiving it.
• Outcomes include the assessment of educational or cognitive achievement which reports quantitative results from testing of attainment or learning outcomes, such as by standardised tests or other appropriate curriculum assessments or school examinations or appropriate cognitive measures.
• The study design provided a quantitative estimate of the impact of the intervention or approach on the educational attainment of the sample, calculated or estimated in the form of an effect size (standardised mean difference) based on a counterfactual comparison.

Standardised mean differences and confidence intervals for the most appropriate estimates of the impact of the intervention or approach for the Toolkit were extracted from each included study, along with other study variables. These effect sizes were further synthesised into a single pooled effect using a random effects meta-analysis adopting a restricted maximum likelihood (REML) estimation methods. For the full details of the methodology see the EEF's Protocol and Analysis Plan.