Write a literature review. Use when the user says "literature review", "lit review", "review the literature", "summarize prior research", "what does the research say about", "synthesize research on", "background section", "related work section", "what has been studied on this topic", or needs to systematically identify, synthesize, and critically evaluate existing research on a topic - even if they don't explicitly say "literature review".
Based on Research Design (Creswell & Creswell) and Research Methods in Education (Cohen, Manion & Morrison). A literature review is not a list of summaries - it is a critical synthesis that builds the case for your research question. Creswell's standard: the literature review should justify why your study is needed by showing what is known, what is contested, and what gap remains.
The test: after reading the literature review, does the reader understand why this specific research question matters and has not been adequately answered before?
Before writing, establish the boundaries:
Research question or topic: [the specific question being investigated]
Time range: [e.g. 2010–2024, or "foundational works from 1980s onward"]
Disciplines included: [e.g. cognitive psychology, education research, HCI]
Disciplines excluded: [e.g. clinical medicine, economics]
Source types: [peer-reviewed journals / conference papers / books / grey literature]
Databases searched: [Google Scholar / PubMed / IEEE Xplore / JSTOR / etc.]
Search terms: [list primary and secondary keywords used]
Inclusion criteria: [what makes a study relevant]
Exclusion criteria: [what disqualifies a study]
Number of sources reviewed: [total found vs. total included]
Do not write "Smith (2019) found X. Jones (2020) found Y. Lee (2021) found Z." Organize by what the research says, then attribute:
Structure options:
Set the context and state what the review covers:
[Introduce the topic and why it matters - 1 paragraph]
[State the scope of the review - what is and isn't covered]
[Preview the structure - "This review first examines X, then Y, before identifying the gap that motivates the present study"]
For each theme:
Template per theme:
[Theme heading]
The predominant view is that [consensus finding] (Author A, year; Author B, year; Author C, year).
[1-2 sentences elaborating the finding with specific detail]
However, this consensus is challenged by [contrasting finding] (Author D, year; Author E, year).
[Explain the tension: different populations? different methods? different operationalizations?]
Methodologically, most studies in this area [limitation, e.g. rely on self-report measures /
use small convenience samples / are conducted in Western WEIRD populations], which limits
generalizability to [context of your study].
The specific question of [your research angle] remains underexplored. [Cite the one or two
closest prior attempts, and explain why they don't fully answer your question.]
This is the most important paragraph in the review. It must:
While substantial research has established [what is known], the literature has not
adequately addressed [specific gap]. Existing studies have either [limitation A] or
[limitation B]. To date, no study has examined [your specific angle] in [your context /
population / method]. This gap is consequential because [why it matters practically or
theoretically]. The present study addresses this gap by [brief description of your approach].
Close the review with:
1. Annotated bibliography disguised as literature review Bad: Para 1 = summary of Smith. Para 2 = summary of Jones. Para 3 = summary of Lee. Good: Para 1 = consensus finding with Smith, Jones, Lee as evidence. Para 2 = contradictions with Davis and Patel as evidence.
2. Gap statement that's too vague Bad: "More research is needed in this area." Good: "No study has examined the relationship between X and Y in [specific population], despite theoretical predictions that [mechanism] should produce [effect]."
3. Only citing supportive literature Bad: Every source confirms the same point. Good: Literature review actively seeks and presents contradictory findings. Addressing them strengthens, not weakens, your argument.
4. Methodology ignored Bad: Treat all findings as equal regardless of study design. Good: Note when findings come from RCTs vs. correlational studies, large samples vs. small, and what that means for confidence in the conclusion.