AI Screening Practice Tool
Upload or paste study abstracts, apply your inclusion and exclusion criteria, and see how your screening decisions compare with AI classifications — then refine your criteria based on disagreements.
Define Your Review Criteria
Clear inclusion and exclusion criteria are the foundation of reproducible screening. Define your criteria once here — they apply to every abstract you screen and guide the AI's classifications throughout the session.
📚 What Makes Good Screening Criteria?
Screening criteria translate your PICO question into operational decision rules. Good criteria are specific enough to be applied consistently by different reviewers, yet broad enough not to exclude relevant studies. Common criteria cover: study design, population characteristics, intervention type, geographic scope, language, and publication date.
- Too vague: "Studies on youth programs" — any reviewer could interpret this differently.
- Specific enough: "RCTs, quasi-experiments, or cohort studies targeting youth aged 15–24 in LMICs with a vocational training intervention and employment outcome."
Briefly describe the topic and purpose of your systematic review. This helps the AI apply your criteria accurately.
Enter your criteria as numbered or bulleted lists. Be as specific as possible — these are applied verbatim by the AI screener.
Import Your Abstracts
Paste abstracts directly, import a CSV export from a reference manager, or load a built-in practice set. The tool supports up to 50 abstracts per session.
TITLE: [title] on the first line, then ABSTRACT: [text] on the next. Or paste plain paragraphs — the tool will auto-number them.Screen & Compare Decisions
Rate each abstract yourself first — Include, Maybe, or Exclude — then reveal the AI's verdict and reasoning. Work through all records before reviewing the comparison analysis.
Refine Criteria & Export Results
Analyse where your decisions diverged from the AI's, identify ambiguities in your criteria, and generate a refined criteria set. Then export your screening results and a transparency statement.
A full record of your decisions vs the AI's. Sort by the Match column to quickly find systematic disagreements.
| # | Title | Your Decision | AI Decision | Match | Your Notes |
|---|
Based on the pattern of disagreements, the AI will identify which criteria are ambiguous and suggest more precise wording.
Copy your results as a structured text table or generate a PRISMA-ready screening log suitable for inclusion in your systematic review protocol.