All research initiatives

Research Initiative 03

AI for Research Methods & Evidence Synthesis

We study how reliably AI can help researchers conduct systematic reviews and meta-analyses. We compare AI-assisted decisions with independent expert assessments to see where AI performs well, where it misses evidence, and which tasks need closer human review.

For example, when screening titles and abstracts, we examine which eligible studies an AI tool would miss. At full-text review, we check whether its reasons for including or excluding a study match the review criteria.

Projects

Each entry is labelled with its current stage. Select an area or a study to read more.

1 entry

Methodological direction

Scope
Evaluates AI-assisted literature identification, title-and-abstract screening, full-text eligibility assessment, structured study-characteristic and effect-size extraction, and risk-of-bias assessment support across integrative medicine, environmental science, mental health, clinical medicine, and consumer research against expert-verified reference datasets.