A research gap is a specific question the existing literature cannot answer, where answering it would change what someone does or believes. Both halves matter.
Where gaps reliably hide
- Limitations sections. Authors state plainly what they could not do. Read twenty and the pattern emerges.
- Future work paragraphs. Explicit invitations, often unclaimed years later.
- Systematic review conclusions. A good review states where the evidence is insufficient. That is a gap already validated by someone else's exhaustive search.
- Contradictory findings. Two studies disagreeing is a gap: nobody has explained why.
- Untested populations and settings. A finding established in one country, age group or income setting and never checked elsewhere.
- Methodological weakness. A question answered only by small, unblinded or cross-sectional studies is a question still open.
Use reviews as a shortcut
A recent systematic review in your field has already done months of searching. Its discussion section will name the evidence gaps directly. Start there rather than searching from scratch.
Verify the gap actually exists
Before committing, run a proper search — multiple databases, controlled vocabulary, no date restriction. Many apparent gaps are search failures. Our PubMed search guide covers doing this properly.
Check preprint servers and registered trials too. A gap being closed right now by a study already registered is not a gap you should enter.
Then test whether it is worth filling
Ask three questions:
- Who changes behaviour if this is answered? Clinicians, policymakers, other researchers, industry — name them. If you cannot, the gap may be uninteresting rather than unexplored.
- Is it unanswered for a reason? Sometimes the data cannot be obtained ethically or the measurement does not exist. That is a wall, not a gap.
- Is it the right size? Too narrow and nobody cares; too broad and you cannot answer it in one study.
The best gaps are ones where you can explain in a sentence why the answer matters and in a sentence why nobody has produced it yet.
Record it while it is fresh
Keep a running document: the gap in one sentence, the sources establishing it, why it matters, and what answering it would require. By the time you have twenty entries, the good ones stand out on their own.