Team Screening and Dual Review
Invite reviewers, run blinded dual screening, resolve conflicts, track Cohen's kappa agreement, assign workloads, and coordinate in the project discussion.
Two ways to dual screen
Independent dual screening, two reviewers screening every record without seeing each other's calls, is the methodological standard. The tool supports it two ways. Free and offline: each reviewer screens their own copy of the project (exported with Save project), then one of you clicks Reconcile two reviewers to merge both files, score agreement with Cohen's kappa, and resolve conflicts. Blinding comes from working in separate files.
On the Team plan, dual screening moves in-app: everyone works in one shared cloud review, decisions are hidden between reviewers until reconciliation, and conflicts, agreement statistics, assignment, and discussion all live in the Team review panel.
Invite members and set roles
Save the review to your account, then click Team review in the sidebar. On the Members tab, use Add a reviewer: enter the person's account email and pick a role. Four roles exist: owner (full control), arbiter (resolves conflicts and consensus), reviewer (screens independently), and viewer (read-only). Each member also has a blinding toggle; while blinded, a reviewer cannot see anyone else's decisions.
Conflicts, consensus, and agreement
The Conflicts tab lists every record where two reviewers reached opposing keep and drop calls, with each reviewer's decision shown side by side. The owner or an arbiter records the final include or exclude call, optionally with a reason, and the resolution is logged.
The Agreement tab computes pairwise Cohen's kappa and raw percent agreement for every reviewer pair, with a plain-language reading of each kappa (from slight to almost perfect). Use it during pilot screening to catch criteria misunderstandings early, and cite the final kappa in your manuscript.
Pilot sampling and reviewer assignment
Before full screening, draw a calibration sample: click Draw random sample in the Pilot samples section, name it, and have each reviewer screen only that sample (a chip above the list filters the view to the sample). Then check kappa and refine your criteria. The free plan includes one named sample per review; Pro and above are unlimited.
The Assignment tab splits the screening workload. Every record is assigned to two screeners, and per-member workload weights control how much of the list each person gets, so a member with weight 2 receives twice the share of a member with weight 1. Each reviewer can filter their list to their share with the My assignment chip above the record list.
Discussion, activity, and notifications
The Discussion tab is a project chat thread for coordinating the review; recipients are emailed about new messages, throttled to at most one email every four hours per person so inboxes stay sane. The Activity tab shows a live feed of member changes, conflict resolutions, and extraction events, complementing the per-record audit trail.
Tips
- Aim for kappa of at least 0.6 on the pilot sample before starting full screening; below that, your criteria need another pass together.
Invite your co-reviewers and put dual screening into practice.
Set up your review team in the toolRelated guides
Title and Abstract Screening Workflow
Screen titles and abstracts with include, exclude, and maybe calls, criteria and PICO highlighting, relevance ranking, and remappable keyboard shortcuts.
Full-Text Screening and Exclusion Reasons
Move included and maybe records to full-text screening, upload PDFs, find open-access full texts, record exclusion reasons, and feed your PRISMA 2020 counts.
Data Extraction Forms and AI Extraction
Build data extraction forms from templates, run dual extraction with a consensus step, use AI extraction with evidence quotes, and export everything to CSV.