Launch bonus Yearly plans 20% 40% off

Gap Finding

A gap is a question your sources raise and do not answer. Finding one is how a review turns into a contribution, but a claimed gap is only worth something if you can show where the literature stops. This guide asks your library what it leaves open, maps each gap next to the papers that come closest to answering it, and helps you tell a gap in the field from a gap in your library.

The screenshots come from a real run on a library of ten open-access papers about social media use and adolescent mental health, using the default model.

Step 1: Ask your library what it leaves open

Start in the project's Chat page, from the project sidebar. It answers from every library attached to the project, with citations, and changes no mindmap, so it suits a first look.

Pick the ready-made prompt Find research gaps and replace [topic] with your question:

What does my library leave open about social media use and adolescent mental health?

Identify questions the sources raise but do not answer, conflicting findings that remain unresolved, material or methods that are missing, and claims that rest on thin evidence. For each gap, explain why it matters and cite the passages that reveal it.

The agent searches the library more than once, then answers gap by gap. Each gap cites the passages that reveal it, often the authors' own statements about what their study could not settle. This answer took under a minute and cost 0.6 credits.

Step 2: Map the gaps next to the closest papers

A chat answer is easy to lose. To keep the gaps and work on them, put them in a mindmap. Create a new mindmap in the same project and send:

Map what my library leaves open about social media use and adolescent mental health. Make one branch per kind of gap: questions the sources raise but do not answer, conflicting findings that remain unresolved, missing populations, measures, or methods, and claims that rest on thin evidence. Under each gap, add the claims from my library that come closest to answering it, with citations, and say what each one still leaves open.

This run took about a minute and cost 1.4 credits.

Under each gap you now see the papers that come closest, with the pages that show where they stop, and a closing claim on what the library does not yet establish.

Step 3: Check whether the gap is real

The agent can only see your library. A gap in the map can mean the field has not answered the question, or that your library is missing the paper that did. Before you build on a gap:

  1. Search the wider literature for the closest papers' titles and the gap in your own words, in your usual databases.
  2. Add any paper you find to the library and run the ready-made prompt Fill the gaps in the gap map. It adds what the new paper covers, so you can see whether it answers the question.
  3. Click the citations under the gap and read the passages. Authors often say exactly what their design could not test, and that sentence is the best evidence for your gap.

Step 4: Turn a gap into a question

Once a gap holds up, test the question you want to ask against the library before you commit to it. Two ready-made prompts on the Chat page help:

  • Test a hypothesis breaks your idea into the conditions it depends on and finds the evidence for and against each one.
  • Check a claim gives a verdict on a single statement: supported, partly supported, contested, or not addressed.

More prompts to try

For the gap gap, list the study designs that could close it, and say which papers in my library come closest to each design.

Which findings in my library come from one country, one age group, or one platform only? Add a branch for each limit, with citations.

Find the claims in this map that rest on a single study, and add a sibling claim saying what a second study would need to show.

Which questions do the authors in my library explicitly name as future research? Quote the passage for each, with its page.

Tips

  • Prefer the authors' own words. A limitation an author states about their own study is stronger evidence for a gap than an inference across papers.
  • Keep gaps and claims in separate maps. A gap map changes often while you read. Keep it apart from the map you write your chapter from.
  • Revisit the gap map when the library grows. Run Fill the gaps after each batch of new papers and see which gaps close.

What's next