boundary
Should This Paper Enter Your Evidence Table or Stay Out?
Judge the paper against the inclusion criteria you wrote before reading it, record the decision with its reason, and treat a documented exclusion as a result in its own right. This guide is for researchers stuck on a borderline source who suspect their criteria might be bending. ClaimWeft keeps such decisions traceable. You can open these pages and sign in to the workspace today. That does not mean ClaimWeft has connected accounts on research databases or library services, and it does not mean the product has formally launched. The workspace organizes the decision trail; the judgment stays with you.
Why this decision carries weight
Inclusion decisions compound. One paper admitted because its abstract sounded relevant quietly resets the standard for the next fifty, and a review that admits on charm ends up arguing with itself. The reverse failure is just as real: excluding everything imperfect until only friendly sources remain.
Deciding here means applying fixed criteria and recording the reasoning, not sensing whether a paper feels important. ClaimWeft may describe candidate workflows and sample outputs from the materials you supply. Coverage and interpretation still need expert confirmation, and no live database connection exists to fetch papers you have not supplied.
What a defensible decision needs
Your written inclusion criteria, the paper itself, and enough access to check whether the full text matches the abstract's promise. If you only have the abstract, the honest decision is 'pending full text', not a guess.
A second reader helps most at this exact step. Borderline calls reviewed alone drift; borderline calls reviewed in pairs leave a record both people signed.
Step 1: Re-read the criteria before the paper
Open your inclusion criteria first and read them as written. The order matters: criteria read after the abstract already prime a decision, and you will mistake familiarity for fit.
Step 2: Test the paper against each criterion in turn
Walk the criteria one by one: population, methods, publication type, date range, whatever your review requires. A paper that fails one hard criterion is out no matter how well it scores on the rest; criteria are gates, not a scoring rubric.
Step 3: Record the decision with its specific reason
Write which criterion decided the case and what in the paper triggered it. 'Not relevant' is not a reason; 'no control condition, fails methods criterion 3' is. The workspace keeps these records beside the claim they would have fed, so exclusions stay auditable.
Step 4: Review the borderline pile as a set
After several decisions, read the borderline records together. If every exclusion favors one conclusion, the criteria or their application need a second reader. Consistency across the pile is what a methods reviewer actually checks.
Checking the decisions hold up
Pick three included and three excluded papers at random and re-run their records against the criteria. Each decision should reproduce from the written reason alone, without your memory of the paper. If a colleague cannot reconstruct why a paper entered or stayed out, the trail has a hole.
Boundaries of this routine
Researchers still verify passage locations, citation links, and inclusion criteria. Coverage, evidence interpretation, and research conclusions need expert confirmation. ClaimWeft may describe candidate workflows and sample outputs from the materials you supply.
A recorded exclusion is not a verdict on the paper's quality, only on its fit for this review's question. Do not present the decision trail as complete coverage, and do not treat stored records as a substitute for reading the sources.
Where ClaimWeft sits in the decision
ClaimWeft holds the screening queue, the criteria text, and each decision record together, so a borderline call produces a traceable entry rather than a mental note. Ask it to keep exclusions beside the claims they would have touched, and to surface the borderline pile when it grows.
It does not fetch papers, rank quality, or vote. The routine above is yours to run; the workspace just makes sure the run leaves evidence.
FAQ
Questions this guide is for
Can I include a paper and mark it low quality instead?
Sometimes that fits your protocol, but mixing inclusion with quality judgment muddies both. Decide entry by criteria first; handle quality inside the evidence record afterward.
What if the full text is unavailable to me?
Record the decision as pending with the reason. An honest gap in the trail is defensible; a guess either way is not.
Should a paper that conflicts with my hypothesis be included?
Yes, and its conflict recorded openly. Excluding disagreement is the fastest way to make a review useless to anyone who knows the field.
Start in the workspace
Make the decision traceable
Sign in or create an account. Bring the criteria and the borderline paper, and record the inclusion call in the ClaimWeft workspace where the reasoning stays auditable.