A literature review is not a summary of sources. It is a structured synthesis that maps what is known, identifies what is contested or missing, and positions your own research within that landscape. This guide walks through the process step by step — from building your search strategy to writing the final draft.
Step 1 — Define the Scope
Before searching for sources, answer three questions: What is the exact topic or research question the review will address? What time range is relevant (last 5 years, last 20 years, all available literature)? What types of sources count — peer-reviewed journals only, or also conference papers, books, and reports?
A narrow, well-defined scope produces a focused review. A literature review on «artificial intelligence in education» is unmanageable; «machine learning for automated essay scoring in higher education (2015–2025)» is workable.
Step 2 — Build Your Search Strategy
| Element | How to handle it |
|---|---|
| Databases | Google Scholar, PubMed (health/bio), IEEE Xplore (engineering), JSTOR (humanities), Scopus, Web of Science |
| Search terms | Use Boolean operators: «automated essay scoring» AND «machine learning» NOT «K-12» |
| Synonyms | Map alternative terms before searching: «NLP», «natural language processing», «text mining» may all be relevant |
| Filters | Date range, language, publication type, peer-reviewed only |
| Citation tracking | Check who cites the key papers you find (forward citation search in Google Scholar) |
Step 3 — Screen and Select Sources
Apply a consistent inclusion/exclusion protocol. Read titles and abstracts first; only download full texts for sources that pass the initial screen. A simple screening table helps:
| Inclusion criteria | Exclusion criteria |
|---|---|
| Peer-reviewed journal articles or conference papers | Opinion pieces, editorials, or non-reviewed sources |
| Published 2015–2025 | Studies published before 2015 (unless seminal) |
| Directly addresses the research question | Tangentially related or background-only |
| Available in full text | Abstract-only sources |
Step 4 — Read and Take Structured Notes
For each source, record: the main argument or finding, the methodology used, the sample or dataset, the key limitations acknowledged, and how it relates to your research question. Tools like Zotero, Mendeley, or a simple spreadsheet work well for this. The goal is not to summarize each paper independently — it is to identify patterns across papers.
Step 5 — Identify Themes and Gaps
Group your sources by theme, finding, or methodology — not by author or date. Common organizational approaches:
| Organization approach | When to use it |
|---|---|
| Thematic | Most common; groups studies by topic or finding |
| Methodological | When comparing how different methods approach the same question |
| Chronological | When the field has evolved significantly over time and that history matters |
| Theoretical | When different theoretical frameworks lead to different conclusions |
Once themes are mapped, gaps become visible: topics that are underresearched, populations that are understudied, contradictions in findings that have not been resolved.
Step 6 — Write the Review
A literature review has three structural parts:
- Introduction — states the purpose, scope, and search strategy; explains why this review is needed
- Body — organized by themes; synthesizes rather than summarizes; uses transitions to show relationships between studies («In contrast to Smith (2021), Jones et al. (2023) found…»)
- Conclusion — summarizes main findings, identifies gaps, and explains how the reviewed literature justifies the current study or research question
Synthesis vs. Summary: The Core Distinction
| Summary (what to avoid) | Synthesis (what to do) |
|---|---|
| «Smith (2021) found that… Jones (2022) found that… Lee (2023) found that…» | «Three studies consistently found X (Smith, 2021; Jones, 2022; Lee, 2023), though they diverge on the role of Y.» |
| Each source gets its own paragraph | Sources are grouped by what they show, not who wrote them |
| No connections between studies | Agreements, contradictions, and gaps are made explicit |
Citation Format in Literature Reviews
Use whichever citation style your field requires — APA for social sciences, IEEE for engineering, MLA for humanities. In all cases, every claim drawn from a source must be cited. When synthesizing multiple sources for one point, list them all in a single parenthetical or bracket group: (Smith, 2021; Jones, 2022; Lee, 2023) in APA, or [1]–[3] in IEEE.
Common Mistakes
- Treating the review as an annotated bibliography — each source is not a separate entry; sources are woven together around themes
- Citing only sources that support your argument — include contradictory findings and explain why your approach addresses their limitations
- Missing seminal works — foundational papers in a field should appear even if they are older than your date filter
- No explicit connection to your own research — the review must end by explaining the gap that your study fills
For citation formatting and reference list rules in APA, see the APA references guide.