How to find research gaps?

#research

What is research gap?

A gap is a question that the literature doesn't answer well. It is not a topic no papers. A topic with no papers is often a topic with no value. Mile (2017) names seven gap types. The main ones are: evidence gaps (results disagree), knowledge gaps (no study exists), methodological gaps (the method limits the result), empirical gaps (a claim has no test), theoretical gaps (no theory explains the data), and population gaps (a group or setting has no study). Use these names when you write your gap. They make the gap easy to check.

Procedure to find a gap

  1. Read the last five years of survey papers in your area. Note the "open problems" and "future work" sections. These are gaps that experts already accept.
  2. Read the limitations section of each strong paper. Make a table with one row per paper. Record the method, the dataset, the assumptions, and the stated limits.
  3. Look for conflicts in the table. Two papers that report opposite results are an evidence gap. A common assumption that no paper tests is an empirical gap.
  4. Look for borrowed methods. Ask: does a method from a nearby field solve a problem in my field? This is a frequent source of PhD topics.
  5. Look at what people build but do nnot measure. Many system papers show a design, but no paper measures the design under realistic conditions.
  6. Reproduce one key result. Reproduction fails often. Each failure is a gap, and you learn the area in depth.
  7. Write each candidate gap in one sentence. Use this form: "No study shows X under condition Y, and this matters because Z." Delete every gap where you cannot fill in Z.
  8. Test the gap with people. Ask you advisor and senior students. Ask them: "Is this gap real, and why has no one filled it?" A gap that no one filled for a bad reason is a good gap. A gap that no one filled because it is impossible or useless is a trap.

Common errors

  • You claim a gap after a weak search. Reviewers find the paper you missed. Search in at least three datases before you claim novelty.
  • You choose a gap that no one cares about. Always state the "why it matters" sentence.
  • You choose a gap that is too large for one PhD. Cut the gap until one paper can close one part of it.
  • You wait for a perfect gap. Start with a small gap. Good gaps often appear while you do the first project.

References