ResearchRabbit and Connected Papers help you explore relationships among academic papers. They are useful when a keyword search has produced a promising study and you want to understand its surrounding literature. The decision is less about which map looks more impressive and more about how you will use the discoveries.
This guide compares documented approaches and offers a practical evaluation workflow. It is not a benchmark of database coverage, and the example below does not report a completed search.
What ResearchRabbit is useful for
ResearchRabbit presents discovery around seed papers and an evolving collection. You can explore related work and connect a reference library to the discovery process. It is a candidate when you want to return to a collection and follow several promising paths over time.
For example, a teacher-researcher investigating feedback could begin with one well-matched study, inspect related papers, and organise candidates for later screening. The collection should remain a working list, not an automatic list of studies to include in a review.
What Connected Papers is useful for
Connected Papers builds a graph around an origin paper. Its published graph legend explains that papers are arranged by similarity, rather than as a citation tree. Node size represents citation count and colour represents publication year. Those encodings help you read the display; they do not establish study quality.
It is a candidate when you want to inspect the neighbourhood around a particular paper. Similarity can surface useful connections, but a nearby paper may still address a different population, method, or question.
Which workflow fits your project?
- An evolving collection: explore ResearchRabbit if you want to build on selected papers across several discovery sessions.
- A close look around one origin paper: explore Connected Papers if that graph is the immediate question you want to investigate.
- A formal evidence synthesis: use either as a supplementary discovery route and keep the wider search method documented.
- A reading list: screen candidates before recommending them to students or colleagues.
Use the same seed paper to evaluate both
Choose a paper you have read and that closely matches your research question. Avoid using a broad review solely because it has many citations. Record its title, DOI or other identifier, and why it is a suitable starting point.
- Run a discovery session from the seed in each tool.
- Record the date, starting paper, and any filters.
- Screen a small number of suggestions using the same inclusion rules.
- Open the original records and check publication details.
- Save useful candidates and write a one-sentence reason for each.
- Repeat from a different relevant seed to see whether you keep finding the same narrow group.
For an education topic, a useful screening rule might be: “Include studies of feedback on undergraduate writing; exclude workplace training and tools used only for automatic scoring.” This is an example scope, not a finding about the available literature.
Keep a discovery log
Copyable log: Date: ___ | Tool: ___ | Seed identifier: ___ | Candidate identifier: ___ | Route found: ___ | Include / exclude / unsure: ___ | Reason: ___ | Next action: ___
The route matters. A paper found through a map and a paper found through a database query can both be useful, but your methods record should say how you located them. Keep discovery separate from full-text screening and quality appraisal.
Do not read a map as a ranking of evidence
- A frequently cited paper is not automatically the strongest study for your question.
- A recent paper may have fewer citations because less time has passed.
- A similarity connection does not mean two authors agree.
- A cluster can reflect a shared topic without a shared method or outcome.
- A missing paper is not proof that the study is unimportant or does not exist.
Return to titles, abstracts, methods, and references to interpret the relationship. Use maps to generate reading decisions, then justify those decisions with the papers themselves.
Move from discovery to your reference manager
Once you have useful candidates, save accurate citation records and attach your screening notes. ResearchRabbit documents a Zotero importer and an export route; see its current integration guide. Check what is transferred instead of assuming every note, attachment, or later change will synchronise automatically.
A practical choice
Try the narrow task you need today. If one tool gives you clearer routes to relevant papers and a manageable record of your decisions, keep it in that role. Neither tool removes the need for database searching, source reading, or an explicit inclusion method when those are required by your project.
For the account connection and transfer steps, follow the ResearchRabbit and Zotero import tutorial.
Related guides and further reading
Explore more Research Tools and AI in Education.
