Streamline Screening with Data Sampling

Create randomized subsets, divide the workload, and coordinate your review team — all from within Rayyan's collaborative screening platform.

The Coordination Challenge

• Large datasets require multiple reviewers working in parallel
• Manual work division risks overlap, gaps, and selection bias
• Teams need structured coordination without slowing down

Organized, Bias-Free Team Screening

• Create randomized samples to distribute work fairly
• Multiple assignment methods to fit any review protocol
• Full audit trail ensures transparent, reproducible reviews

Powerful Sampling & Assignment Tools

From randomized subsets to structured team coordination, Rayyan gives you every tool to manage collaborative screening at scale.

1 Randomized Data Sampling

Eliminate bias with truly randomized subsets of your dataset.

Create randomized samples of any size from your dataset to distribute among your review team. Set a percentage or exact article count, and Rayyan generates unbiased subsets — ensuring every collaborator works on a fair, representative portion of the literature.

• Generate randomized subsets by percentage or article count
• Reduce selection bias through true randomization
• Create unlimited samples with a premium subscription
• Samples appear as filters in your Workbench for easy access

2 Team Work Division

Organize your review team and assign references efficiently.

Divide the workload across your review team using Rayyan's flexible assignment methods. Whether you use sampling, decision filters, or search-based assignments, every team member knows exactly which references to screen — keeping your review on track and on schedule.

• Assign specific sample datasets to individual collaborators
• Use Min/Max decision filters so reviewers only see unscreened articles
• Divide references by uploaded search files for structured assignment
• Supports both blinded and unblinded review workflows

3 Smart Decision Filters

Control reviewer workload with precision filtering.

Rayyan's decision-based filtering lets each team member focus on articles that still need attention. Filter by 'At Most 1' to see references awaiting a first or second decision, or combine filters to target exactly the right subset — preventing duplicate effort and ensuring complete coverage.

• Filter by number of collaborator decisions per reference
• 'At Most 1' filter shows articles needing first or second review
• Combine 'At Most' and 'At Least' for precise targeting
• Prevents reviewers from duplicating each other's work

4 Search-Based Assignment

Divide work by source files for structured team coordination.

Upload your references as separate search files and assign each file to specific team members. Rayyan lists all uploaded searches under the Search Methods filter, letting collaborators select their assigned file and screen only those references — ideal for large, multi-database reviews.

• Upload multiple search files to a single review
• Each file appears as a selectable filter in the Workbench
• Assign team members to specific search files
• Collaborators see only their assigned references when filtered

5 Prediction Classifier Training

Improve Rayyan's AI predictions with strategic sampling.

Use Data Sampling to create training sets that improve Rayyan's prediction classifier. By screening a representative, randomized sample first, you help the AI learn your inclusion criteria faster — resulting in more accurate Compute Ratings across your entire dataset.

• Train Rayyan's prediction engine with randomized screening samples
• Improve Compute Ratings accuracy across your full dataset
• Strategic sampling accelerates the AI learning curve
• Combine with collaborative screening for faster calibration

Key Benefits

• Accelerate Reviews: Divide and conquer — distribute work efficiently so your team screens in parallel.
• Reduce Bias: Randomized sampling ensures fair, unbiased distribution of references.
• Scale Your Team: Coordinate any number of reviewers with structured assignment methods.
• Flexible Workflows: Choose the division method that fits your review protocol and team size.
• Train AI Faster: Strategic sampling improves prediction accuracy across your dataset.

Ready to Organize Your Review Team?

Data Sampling helps your team screen faster, reduce bias, and stay coordinated — no matter how large the dataset.