Reading and Reviewing Papers › Module 5: Writing a literature review

Finding the gap

Lesson 3 of 3 · 2 min

Previous

A research gap is something important that is not yet known. Your project fills it, or part of it. A clear gap is what makes reviewers say "this work is needed".

Kinds of gap

Not tested in this setting
Example: Methods work on English speech, but not tested on code-switched Hindi-English
Method gap
Example: Everyone uses the same evaluation, which misses an important failure
Conflicting results
Example: Two studies disagree, and nobody has tested why
Old result, new conditions
Example: A finding from small models has not been checked on today's large models
Missing resource
Example: No public dataset exists for a task
Not reproduced
Example: An important result has never been independently repeated

Where to look

  • The limitations and future work sections of recent papers. Authors often tell you what is still open.
  • The gaps in your literature matrix: empty cells, or columns where every paper uses the same thing.
  • Surveys, which often end with a list of open problems.

Stating your gap honestly

Write it as a short chain: what is known, what is not, and why it matters.

Example"Data augmentation reduces speech-recognition errors for low-resource languages (A, 2022; C, 2024). However, these studies used read speech only. Conversational speech, which is harder and more common in real use, has not been tested. We test whether the same gains hold for conversational Tamil."

Be careful with "first"

Avoid saying "no one has ever studied this" or "we are the first" unless you have searched very thoroughly. Reviewers often know a paper you missed. Safer wording: "to our knowledge, no study has tested...".

RememberA good gap is specific, real and worth filling. "Nobody has done exactly this" is not enough: it must also matter.