Research Gap Identification

How do I identify a research gap?

Start from the decision, not from the literature. Write the decision as a choice between options, list every proposition that has to be true for each option to be right, and attach the evidence status to each. The gaps that matter are the decisive dependencies with no evidence behind them, and they are usually a much shorter list than the unanswered questions in the literature.

What are the different types of research gap?

Seven: no evidence at all; weak evidence; conflicting evidence that survived adjudication; evidence about a different population or market; evidence that is out of date; evidence about a different construct; and evidence that exists and is adequate but was never applied to this question. The last is the most commonly missed and the cheapest to close.

How do I know whether a gap is real or just something I did not find?

Ask which source types would carry the evidence if it existed and whether you searched them, what vocabulary the community that studies this uses, whether the evidence would be published at all or would sit in unpublished or commercial form, and whether the organisation has already answered it in work nobody wrote up. Then record a verdict: confirmed absent, probably absent, or not established.

Which research gaps are worth filling?

Only those where a different answer would change what the organisation does, where the uncertainty is wide enough to matter, where the cost is proportionate to what the decision has at risk, and where research could actually reduce the uncertainty enough to settle it. Most gaps fail at least one of these, and saying so is the professional contribution.

When should we decide without filling the gap?

When the gap cannot be filled before the decision, cannot be filled at all, would cost more than the decision is worth, or would return an answer too weak to settle it. That recommendation is incomplete unless it names the assumption being made in place of the evidence, the exposure if the assumption is wrong, and a mitigation such as a reversible first step or a monitoring tripwire.

Can primary research fill every gap?

No. Some questions are about what happens after a change that has not happened yet, or about a duration that has not elapsed, and no design compresses time. Some cannot be studied for ethical, legal or access reasons. Some are about counterfactuals nobody can observe. And some are not empirical at all: risk appetite, values and what an organisation is willing to be seen doing are judgements, not gaps, and framing them as research questions produces a study that cannot succeed.

How should a research agenda be prioritised?

On decision relevance, likelihood that the answer changes the action, and feasibility within the budget and calendar, applied in that order rather than combined into a weighted score. State the rule in the output so a reader can argue with the ranking rather than with the register. Weighted composites in prioritisation tend to be retuned until they produce the answer someone already wanted.

What should a gap register contain?

Every gap with its type, coverage verdict, materiality, fillability, priority and disposition; a specification for each gap to be researched including the precision required; a list of gaps answerable from evidence already held; a decide-now list with assumptions and mitigations; a monitoring list; the gaps that were discarded and why; and the decision the register was built for, with a shelf life.