Consider Elicit for a documented research workflow, then verify papers against their originals.
Start with the finished task
Define the question, inclusion criteria and search date before collecting papers. Keep discovery, screening and evidence extraction as separate stages.
- 1Record paper identifier, year, inclusion
Record paper identifier, year, inclusion decision, reason and the source passage.
- 2Check whether a claim comes
Check whether a claim comes from methods, results or discussion.
- 3Keep uncertain eligibility in a
Keep uncertain eligibility in a review queue rather than forcing an automatic decision.
- 4A second reviewer can audit
A second reviewer can audit disputed cases without repeating the entire search.
Choose a workflow that fits
Keep your own screening log even if a tool supplies summaries. Avoid treating an extracted claim as a substitute for the paper or allowing invented citations into a reference list.
| Tool | Source evidence | Confirm before buying |
|---|---|---|
| Elicit ↗ | Elicit describes research assistance and literature-review workflows. | Coverage, selected-plan exports and access to full text. |
| Manual or conventional route | Use a reference manager and spreadsheet when the paper set is small and already identified. | Editorial workflow advice; this is not a measured comparison or vendor performance claim. |
The correction in our worked example
Topic overlap is not eligibility. The corrected decision applies the criteria established before screening.
Fictional review includes adult studies with a comparison group. Three supplied abstracts include one pediatric case report.
Include all three because each mentions the topic.
Include the two abstracts meeting the stated criteria. Exclude the pediatric case report with the reason recorded; verify the original papers before extraction.
Build and inspect the handoff
Record paper identifier, year, inclusion decision, reason and the source passage. Check whether a claim comes from methods, results or discussion. Keep uncertain eligibility in a review queue rather than forcing an automatic decision. A second reviewer can audit disputed cases without repeating the entire search.
Keep a screening record
Record one inclusion or exclusion reason for each paper. Distinguish population, study design and measured outcome; similar titles do not establish that studies answer the same question. Use a reference manager and spreadsheet when the paper set is small and already identified.
Compare the full cost
Database and full-text access can be separate costs from the AI subscription. Record export restrictions and source coverage before choosing a paid tier. This example is an editorial screening exercise, not a completed scholarly review.
Research questions
Can AI citations go directly into my bibliography?
Verify title, authors, identifier and claim against the original source first. Discovery output alone is insufficient.
When is a simpler workflow a better choice?
Use a reference manager and spreadsheet when the paper set is small and already identified.
THE RESEARCH DESK
Sources & methodology
Official documentation was consulted on 2026-10-07. The worked example uses fictional inputs and was edited for this guide. No vendor application benchmark or outcome test was performed. Prices and allowances should be confirmed before purchase.
elicit.com · ↗Put this guide to work
Your decision criteria: Coverage, selected-plan exports and access to full text.
- Fictional review includes adult studies with a comparison group. Three supplied abstracts include one pediatric case report.
- Record one inclusion or exclusion reason for each paper. Distinguish population, study design and measured outcome; similar titles do not establish that studies answer the same question.
- Coverage, selected-plan exports and access to full text.
