Literature Review Guide: How to Write One That Actually Works
This guide provides a practitioner's approach to writing literature reviews that genuinely synthesize, rather than merely summarize. It covers structure, methods, and the crucial step of gap identification, essential for robust academic and professional work.
By OASIS Research Writing
- literature review guide
- how to write a literature review
- systematic literature review
- narrative literature review
- literature review structure
- identifying research gaps
- literature review synthesis
- literature review best practices
- AI in literature reviews
- difference between literature review and annotated bibliography
- academic research writing
- writing a thesis literature review
A literature review is often a core component of academic research, grant proposals, and strategic decision-making. However, many reviews fall into the trap of merely summarizing sources, rather than achieving true synthesis. This guide from OASIS Research & Writing clarifies the purpose and methodology of an effective literature review, moving beyond simple annotation to reveal the interconnectedness of existing knowledge.
We define what a literature review actually does, differentiate between various types, and outline a step-by-step process for crafting a review that maps what is known, identifies what is contested, and precisely articulates remaining knowledge gaps. Our insights focus on thematic organization, precise gap identification, and the judicious use of AI tools, ensuring your review is a foundational asset for new research or critical decisions.
Download the original PDF (24 Literature-Review-Guide-OASIS_1.pdf)
What a Literature Review Actually Does
A literature review maps the current state of knowledge on a defined question, evaluates the strength and consistency of existing evidence, and identifies what remains unresolved. It is a required component of theses, dissertations, and grant proposals, and a valuable stand-alone document whenever a decision or new research project needs to be positioned against what's already known. Its most common failure is degenerating into a source-by-source summary that never actually synthesises anything.
The most common failure in literature reviews isn't poor research — it's poor synthesis. A strong literature review shows how sources relate to each other: where they agree, where they conflict, and what none of them have yet answered.
Types of Literature Reviews
- Narrative review: A thematic synthesis of the literature, typically used within a thesis or paper's introduction.
- Systematic review: A rigorously documented, reproducible search and inclusion process, often used in medical and scientific research.
- Scoping review: Maps the breadth of research on a topic, often to identify gaps before a systematic review is undertaken.
- Meta-analysis: Statistically combines results from multiple studies to produce a pooled estimate of effect.
The Literature Review Process, Step by Step
- Define the review question precisely. A broad topic produces an unmanageable and unsynthesizable review.
- Set inclusion and exclusion criteria. Which sources count, and why, decided before searching begins.
- Search systematically. Across relevant databases, not just the first page of a general search engine.
- Evaluate each source's quality and relevance. Before including it in the synthesis.
- Organize findings by theme. Not by source or publication date.
- Identify agreement, disagreement, and gaps. Explicitly, across the body of literature as a whole.
- Write the synthesis. Connecting sources to each other, not summarizing them in sequence.
Organizing by Theme, Not by Source
The clearest sign of a weak literature review is a structure that mirrors the list of sources rather than the themes within them. A strong review is organized around sub-questions or themes — for example, "studies on X mechanism," "studies on Y population," "methodological critiques of prior work" — with each source appearing wherever it's relevant, sometimes in multiple sections.
Identifying the Gap
The gap is the specific thing existing literature has not yet established — a population not studied, a mechanism not tested, a contradiction not resolved. Identifying it precisely is what justifies new research (or a new decision); a vague gap statement ("more research is needed") signals the review hasn't actually synthesised the literature closely enough to find the real one.
Systematic vs. Narrative Reviews
| Dimension | Narrative Review | Systematic Review |
|---|---|---|
| Search process | Informal, researcher-directed | Formally documented and reproducible |
| Typical use | Thesis/paper introduction, general synthesis | Medical/scientific evidence synthesis, policy evidence bases |
| Reproducibility | Low — depends on the reviewer's judgment | High — search strategy and criteria are documented and repeatable |
| Time required | Days to weeks | Weeks to months |
AI in Literature Reviews
AI tools can accelerate the search and initial screening stages, surfacing candidate sources faster than manual database searching. However, they cannot reliably judge study quality or synthesise nuanced disagreement between sources without human oversight. Any AI-summarized source must be checked against the original before being relied upon, as AI summarization tools have a documented tendency to misrepresent a study's actual findings or overstate its confidence.
Common Mistakes
- Summarizing sources sequentially instead of synthesising them by theme.
- Including sources without evaluating their quality or relevance.
- Stating a vague gap ("more research is needed") instead of a precise one.
- Failing to document the search strategy, making the review impossible to reproduce or update.
- Relying on an AI-generated summary of a source without reading the source itself.
Best Practices
- Define the review question and inclusion criteria before searching.
- Organize the review by theme, not by source.
- State explicitly where sources agree, disagree, or leave a question unanswered.
- Identify the gap precisely, not generically.
- Verify any AI-assisted summary against the original source before relying on it.
Frequently Asked Questions
How many sources should a literature review include?
It depends on the scope of the question and the review type — a narrow, well-defined question can be reviewed thoroughly with fewer sources than a broad one, and quality of synthesis matters more than raw count.
What's the difference between a literature review and an annotated bibliography?
An annotated bibliography summarizes each source individually; a literature review synthesises sources together around themes and identifies the resulting gap.
Can AI write a literature review?
AI can assist with search and initial screening, but reliable synthesis and gap identification still require human judgment, and every AI-assisted summary needs verification against the original source.
Key Takeaways
- A literature review synthesises sources by theme; it does not summarize them one by one.
- Precisely identifying the gap is what justifies the new research or decision that follows.
- AI can accelerate search, but cannot replace human judgment in evaluating quality and synthesising disagreement.