AgentSkillsCN

Literature Review

系统性地开展文献检索、综合分析、缺口识别与叙事构建,为学术研究注入新活力。

SKILL.md
--- frontmatter
name: "Literature Review"
description: "Systematic literature search, synthesis, gap identification, and narrative construction for academic research"

Skill: Literature Review

Systematic literature search, synthesis, gap identification, and narrative construction for academic research.

Metadata

FieldValue
Skill IDliterature-review
Version1.0.0
CategoryResearch
DifficultyAdvanced
PrerequisitesNone
Related Skillsacademic-paper-drafting, citation-management, dissertation-defense

Overview

Literature review is foundational to academic research. This skill provides structured workflows for systematic searching, critical synthesis, and gap identification—transforming scattered sources into coherent scholarly narratives.

Core Competencies

  1. Systematic Search - Reproducible, comprehensive search strategies
  2. Critical Appraisal - Quality assessment of sources
  3. Synthesis - Thematic integration across sources
  4. Gap Identification - Locating research opportunities
  5. Narrative Construction - Weaving sources into argument

Module 1: Systematic Search Strategy

The PICO(S) Framework

For empirical research questions:

ElementDescriptionExample (AIRS)
PopulationWho is studiedKnowledge workers, AI users
InterventionWhat is appliedAI-powered tools, copilots
ComparisonAlternativeTraditional tools, no AI
OutcomeWhat is measuredAdoption intention, behavioral intention
Study typeResearch designQuantitative survey, SEM

Database Selection

DatabaseBest ForCoverage
Google ScholarBroad discovery, citation trackingAll disciplines
Web of ScienceHigh-impact journals, citation analysisSTEM, social sciences
ScopusComprehensive coverage, author metricsMultidisciplinary
ACM Digital LibraryComputer science, HCICS, CHI proceedings
PsycINFOPsychology, behavioral researchPsychology, social sciences
SSRNWorking papers, preprintsBusiness, economics, law

Search String Construction

Boolean operators:

code
("artificial intelligence" OR "AI" OR "machine learning")
AND
("adoption" OR "acceptance" OR "intention")
AND
("UTAUT" OR "TAM" OR "technology acceptance")

Truncation and wildcards:

  • adopt* → adoption, adopting, adopter
  • behavio?r → behavior, behaviour

Search Protocol Template

markdown
## Search Protocol

**Research Question**: [Your RQ]
**Date Conducted**: [Date]
**Databases Searched**: [List]

### Search Terms
- Primary: [terms]
- Secondary: [terms]
- Exclusions: [terms with NOT]

### Inclusion Criteria
- Published: [date range]
- Language: [languages]
- Study type: [empirical, theoretical, etc.]
- Peer-reviewed: [yes/no]

### Exclusion Criteria
- [List exclusions]

### Results
| Database | Initial Hits | After Dedup | After Title Screen | After Abstract Screen | Final |
|----------|--------------|-------------|--------------------|-----------------------|-------|

Module 2: Source Management

The Funnel Process

code
Initial Search Results (hundreds)
        ↓ Title screening
Potentially Relevant (~50-100)
        ↓ Abstract screening
Candidates for Full-Text (~30-50)
        ↓ Full-text review
Included Sources (~15-30)
        ↓ Backward/forward citation
Final Corpus (~20-40)

Screening Criteria

Quick reject (title/abstract):

  • Wrong population
  • Wrong intervention
  • Wrong outcome
  • Non-empirical (if seeking empirical)
  • Wrong language
  • Duplicate

Full-text assessment:

  • Methodological rigor
  • Relevance to research question
  • Contribution to synthesis

Citation Tracking

Backward citation (reference mining):

  • Review reference lists of key papers
  • Find foundational works
  • Identify theoretical ancestry

Forward citation (cited by):

  • Use Google Scholar "Cited by"
  • Find recent developments
  • Identify methodological improvements

Module 3: Critical Appraisal

Quality Assessment Dimensions

DimensionQuestions
ValidityAre findings believable? Methods appropriate?
ReliabilityWould study replicate? Measures consistent?
GeneralizabilityTo what populations/contexts?
RigorAdequate sample? Controls? Analysis?
TransparencyCan methods be reproduced?

Study Type Checklists

Quantitative survey studies:

  • Sample size adequate (power analysis?)
  • Response rate reported
  • Validated instruments used
  • Reliability reported (Cronbach's α)
  • Common method bias addressed
  • Appropriate statistical analysis

Qualitative studies:

  • Sampling strategy justified
  • Data saturation discussed
  • Researcher positionality stated
  • Coding process transparent
  • Member checking or triangulation

SEM/CFA studies:

  • Model fit indices reported (CFI, RMSEA, SRMR)
  • Convergent validity (AVE ≥ .50)
  • Discriminant validity (HTMT < .85)
  • Sample size adequate (N ≥ 200 preferred)

Module 4: Synthesis Strategies

Thematic Synthesis

  1. Code individual sources - Key concepts, findings, methods
  2. Group codes into themes - Patterns across sources
  3. Develop analytical narrative - What themes mean together

Example themes for AIRS research:

  • Trust and reliance
  • Value perception
  • Effort-benefit trade-offs
  • Social and organizational factors

Comparative Matrix

SourceTheoryMethodSampleKey FindingLimitation
Author1 (Year)TAMSurveyN=312PE→BI significantCross-sectional
Author2 (Year)UTAUT2SEMN=523PV strongestSelf-report

Theoretical Integration

Identifying theoretical tensions:

  • Where do theories contradict?
  • What boundary conditions apply?
  • How can conflicts be resolved?

Building theoretical contribution:

  • What's missing from current theories?
  • How does your work extend understanding?
  • What new constructs are needed?

Module 5: Gap Identification

Types of Research Gaps

Gap TypeDescriptionSignal Phrases
EmpiricalUntested relationships"No study has examined..."
TheoreticalMissing constructs or mechanisms"Theory does not account for..."
MethodologicalBetter methods needed"Prior studies relied on..."
ContextualUntested populations/settings"Research in [context] is lacking..."
TemporalOutdated findings"Since [year], technology has changed..."

Gap Articulation Template

markdown
## Research Gap

**What we know**: [Summary of existing knowledge]

**What we don't know**: [The gap]

**Why it matters**: [Significance of filling gap]

**How this study addresses it**: [Your contribution]

Module 6: Writing the Literature Review

Structure Options

Chronological:

  • Best for: Tracing evolution of a concept
  • Risk: Can become mere historical summary

Thematic:

  • Best for: Synthesizing diverse perspectives
  • Risk: May obscure temporal development

Methodological:

  • Best for: Evaluating research approaches
  • Risk: Can be dry, less engaging

Theoretical:

  • Best for: Comparing frameworks
  • Risk: Requires deep theory knowledge

The Funnel Structure

code
Broad context (the problem space)
    ↓
Theoretical foundations (key theories)
    ↓
Prior empirical work (what's been studied)
    ↓
Gap and contribution (what's missing → your work)

Synthesis vs. Summary

Summary (weak):

"Smith (2020) found X. Jones (2021) found Y. Brown (2022) found Z."

Synthesis (strong):

"While early studies emphasized X (Smith, 2020), subsequent research revealed the moderating role of Y (Jones, 2021), leading to more nuanced frameworks that integrate both factors (Brown, 2022)."

Transition Phrases for Synthesis

PurposePhrases
Agreement"Consistent with...", "Similarly...", "Extending this..."
Contrast"However...", "In contrast...", "Alternatively..."
Extension"Building on...", "Taking this further...", "Integrating..."
Gap signal"Yet...", "Nevertheless...", "What remains unclear..."

Quick Reference

Literature Review Checklist

  • Search strategy documented and reproducible
  • Multiple databases searched
  • Inclusion/exclusion criteria explicit
  • Backward and forward citation completed
  • Sources critically appraised
  • Themes identified across sources
  • Gaps clearly articulated
  • Synthesis (not just summary)
  • Clear link to research questions

Common Pitfalls

PitfallSolution
Too narrow searchUse multiple databases, synonyms
Uncritical acceptanceApply quality assessment
Descriptive onlyFocus on synthesis and analysis
No clear gapExplicitly state what's missing
Outdated sourcesInclude recent (last 5 years) work
Missing seminal worksTrack backward citations

Activation Patterns

TriggerResponse
"literature review", "lit review", "sources"Full skill activation
"search strategy", "database search"Module 1: Systematic Search
"synthesize", "themes", "patterns"Module 4: Synthesis
"research gap", "what's missing"Module 5: Gap Identification
"how to write lit review"Module 6: Writing

Skill created: 2026-02-10 | Category: Research | Status: Active


Synapses

  • [.github/skills/citation-management/SKILL.md] (High, Uses, Bidirectional) - "Lit review requires proper citations"
  • [.github/skills/academic-paper-drafting/SKILL.md] (High, Feeds, Forward) - "Lit review becomes paper section"
  • [.github/instructions/empirical-validation.instructions.md] (Medium, Applies, Forward) - "Evidence-based source evaluation"