Literature Review Writing

When to Use

• Writing the Introduction or Background section of a manuscript or grant
• Creating a standalone literature review paper
• Synthesizing findings from an opencite search session
• Identifying gaps in the literature to justify new research
• Preparing a systematic or scoping review

Review Types

Review Pipeline

Step 1: Gather Papers

opencite

# Foundational papers
uvx opencite canonical "topic" --max 15

# Recent work
uvx opencite search "specific query" --max 30 --sort date

# Citation network
uvx opencite cite "key-paper-DOI" --direction both

# Download and convert for reading
uvx opencite batch-fetch --from-json results.json --convert -o ./papers --summary report.json

Step 2: Read and Annotate

• Key finding (1-2 sentences)
• Method (approach/technique used)
• Population/context (who/what was studied)
• Limitations noted by authors
• How it connects to other papers in the set

Step 3: Identify Themes

• By methodology: group studies using similar approaches
• By finding: group studies with converging/diverging results
• By population: group by subject type or context
• By theoretical framework: group by underlying model
• Chronological evolution: trace how understanding changed over time (use sparingly)

Step 4: Write the Review

Structure Per Section/Paragraph

1. Topic sentence: State the theme or claim this paragraph addresses
2. Evidence: Cite 2-5 papers supporting or nuancing this claim
3. Synthesis: Compare, contrast, and connect the cited works
4. Transition: Lead to the next theme or the gap

Citation Weaving Patterns

Avoid Common Pitfalls

• Annotated bibliography: Don't just summarize each paper sequentially. Synthesize.
• String-of-pearls: Don't write "Smith found X. Jones found Y. Lee found Z." Connect them.
• Overclaiming: Don't say "proves" or "definitively shows." Use "suggests," "indicates," "provides evidence for."
• Recency bias: Include foundational/seminal work, not just the last 2 years.
• Confirmation bias: Include contradictory findings and explain discrepancies.

Step 5: Identify the Gap

• "Despite these advances, no study has examined {specific question}."
• "While X has been established for {context A}, it remains unclear whether this extends to {context B}."
• "Existing studies are limited by {methodological issue}, and {new approach} could address this."
• "The relationship between X and Y has been studied in isolation, but their interaction under {condition} is unexplored."

Step 6: Generate Bibliography

# Export BibTeX for all cited papers
for doi in $(cat cited_dois.txt); do
  uvx opencite lookup "$doi" -f bibtex --append-bib review_refs.bib
done

Output Format

# Literature Review: {Topic}

## {Theme 1 Title}

{Synthesized paragraphs with citations}

## {Theme 2 Title}

{Synthesized paragraphs with citations}

## {Theme 3 Title}

{Synthesized paragraphs with citations}

## Current Gaps and Future Directions

{Gap analysis and research opportunities}

## References

{BibTeX file or formatted reference list}

Quality Checklist

• Organized thematically (not paper-by-paper)
• Each paragraph synthesizes multiple sources
• Citations woven into prose (not parenthetical dumps)
• Contradictory findings acknowledged and explained
• Seminal/foundational work included
• Recent work (last 2-3 years) included
• Gap statement clearly articulated
• Transition sentences connect paragraphs and sections
• No overclaiming (appropriate hedging)
• All cited papers in reference list

Additional Resources

• Reference: references/review-frameworks.md - PRISMA, PICO, and other review frameworks