Evidence4Impact.org
Structured, AI-curated summaries of research papers, published only after administrator review. Human verification is recorded separately so readers can see how each summary has been checked.
A short, guided digital programme reduced anxiety symptoms among secondary students waiting for specialist care. Weekly practitioner messages improved completion compared with an unguided app. The study did not…
This review of 31 studies finds that cash transfers tied to school attendance usually increase enrolment and attendance in secondary school, with smaller and less consistent effects on learning. Programmes that…
This multi-city study found that heat-warning systems were associated with fewer deaths during extreme heat, particularly among people aged 75 and over. The apparent benefit was larger in cities that combined w…
Adding trained community health workers to primary-care teams increased the share of adults with controlled blood pressure. Home visits, medication support and blood-pressure monitoring were the core activities…
A two-year programme of structured literacy coaching for Year 1 and Year 2 teachers improved children’s decoding and reading fluency compared with usual professional development. The largest gains were among pu…
Preschools that used a play-based curriculum with planned adult scaffolding saw modest improvements in children’s self-regulation by the end of the year. Academic school-readiness scores did not differ from a m…
Teaching, learning and school improvement evidence.
Public Health 2 publishedPopulation health, prevention and health services.
Climate & Environment 1 publishedClimate action, conservation and environmental policy.
Social Policy 1 publishedPoverty, housing, employment and social protection.
Mental Health 1 publishedPsychological interventions and mental health systems.
Early Childhood 2 publishedDevelopment, care and education in the early years.
Contributors add bibliographic details and AI-curated summaries covering aim, methods, participants, findings, limitations and implications.
Administrators review every submission. Drafts, returns and rejected records stay private. Nothing is public without explicit publication.
Human verification is independent of publication. A live record may be labelled AI curated, partially verified, or fully human verified.