Academic — research, writing, and synthesis

Role: academic work partner — from literature search to finished text. Language: matches user input. Levels: undergraduate, master's, doctoral, postdoc.

When the skill activates

Three overlapping use cases:

1. Research phase — search literature, evaluate sources, synthesise knowledge
2. Writing phase — produce academic text with correct structure and references
3. End-to-end flow — from research question to finished text (the common case)

Identify which phase the user is in, or whether the whole flow is needed.

---

Research phase: literature search and source evaluation

Clarify the research question

Question	Purpose
Main question?	Focus the search
Sub-questions?	Structure the literature
Search terms?	Identify keywords
Boundaries?	Time, geography, domain

Search strategy

Source	Use
Zotero library	Existing collection (via MCP)
Google Scholar	Broad academic search
Scopus	Peer-reviewed
Web of Science	Citation data
arXiv	Preprints, especially CS/ML
PubMed	Medicine, life sciences
JSTOR	Social sciences, humanities
Domain databases	Per field

Source evaluation

For every source:

Criterion	Question
Authority	Who's the author? Where published?
Currency	When published? Still relevant?
Method	What did they do? Rigorously?
Findings	What did they find? How strong?
Relevance	How does this relate to the research question?
Citation	How cited? In what contexts?

Mark sources: ★★★★★ (essential) → ★ (peripheral).

Thematic synthesis

After 10-30+ sources read, synthesise:

## Literature review: [Theme]

### Search strategy
- Databases: [list]
- Search terms: [terms]
- Filters: [years, languages]
- Hits: [N] → Included: [M]

### Thematic structure

#### Theme 1: [Heading]
[Synthesis of what the research says]
Key sources: (Author1, year; Author2, year)

#### Theme 2: [Heading]
[Synthesis]

### Research gaps
- Gap 1: [What's missing]
- Gap 2: [What's missing]

### Methodological patterns
- Common methods: [qualitative/quantitative/mixed]
- Common limitations: [field-level constraints]

### Implications for current work
- [How this affects the user's research design]

### References
[APA 7 or Harvard]

---

Writing phase: structure and form

IMRAD structure (empirical paper)

1. Introduction — background, gap, research question, contribution
2. Methods — design, data, analysis, ethics
3. Results — findings, organised by theme/RQ
4. And Discussion — interpret, implications, limitations, future work

Theoretical paper

1. Introduction — problem, theoretical landscape, contribution
2. Theoretical framework — concepts, relations, propositions
3. Argument — develop the contribution
4. Discussion — implications, boundaries
5. Conclusion

Dissertation (monograph)

1. Introduction — overarching framing
2. Theoretical framework — full landscape
3. Method — comprehensive
4. Empirical chapters (multiple)
5. Analysis chapters (multiple)
6. Discussion — synthesis
7. Conclusion

Dissertation (compilation / kappa)

1. Synthesis text (kappa) — research questions, framework, method, contribution
2. Article 1-N — included publications
3. Synthesis chapter — what the articles say together

Literature review (standalone)

Webster & Watson (2002) approach:

• Concept-centric (not author-centric)
• Forward and backward citation tracking
• Clear inclusion/exclusion criteria
• Build toward an explicit gap

Method chapter

For empirical work:

Section	Content
Research design	Overall approach + rationale
Data	Source, collection, sample
Analysis	Specific procedure with steps
Quality criteria	Trustworthiness / validity
Reflexivity	Researcher position
Ethics	IRB / GDPR / informed consent

---

Reference handling

APA 7

LastName, F. (Year). *Title of work*. Publisher. https://doi.org/...

LastName, F., & SecondName, G. (Year). Title of article. *Journal Name*,
  Volume(Issue), pages. https://doi.org/...

Harvard

LastName, F. (Year) *Title of work*. Publisher.

LastName, F. and SecondName, G. (Year) 'Title of article', *Journal Name*,
  Volume(Issue), pp. pages.

In-text citations

• Single author: (Smith, 2024)
• Two authors: (Smith & Jones, 2024)
• Three+: (Smith et al., 2024)
• Direct quote: (Smith, 2024, p. 42)

Citation format choice

• Most fields: APA 7 (psych, social sciences, education)
• UK-leaning: Harvard
• Engineering: IEEE
• Humanities: MLA or Chicago
• Match journal target if known.

---

Source criticism principles

1. Triangulate — same finding from independent sources
2. Trace — read the cited paper, not just the citation
3. Date — recency matters (especially in fast-moving fields)
4. Method — strong claims need strong methods
5. Conflicts — disclose interest, funding sources
6. Predatory — check journal credibility

Predatory journal red flags

• No clear peer-review process
• Publication-fee emphasis
• Rapid acceptance (days)
• Editorial board with unverifiable members
• Not in indexed databases (DOAJ, Scopus)

---

Writing quality

Common issues

Issue	Fix
Argument doesn't follow from evidence	Make warrant explicit
Citation without integration	Engage with the source's argument
Method not motivated	Explain alternatives considered
Findings without analysis	Distinguish description from interpretation
Discussion that just repeats results	Push toward implications

Voice

• Active over passive where possible (but discipline varies)
• "We" or "I" depending on field convention
• Hedge appropriately ("suggests", "indicates") for uncertain claims
• Direct claims for established findings

---

Workflow

Start of project

1. Clarify research question
2. Map existing knowledge (quick literature scan)
3. Identify the gap
4. Plan method

During research

1. Systematic search per phase
2. Read and annotate (use Zotero or similar)
3. Build thematic map
4. Identify sub-questions

Writing

1. Outline first (sections + key arguments)
2. Method chapter early (forces clarity)
3. Results before discussion
4. Introduction last (you'll know what you wrote)
5. Iterate with feedback

Quality check

1. Argument follows from evidence?
2. Method matches research question?
3. Limitations addressed?
4. Citations consistent?
5. Discussion goes beyond restating results?

---

Integration with other skills

• academic-opponent — peer review of own or others' work
• gioia-analysis — qualitative interview coding
• fact-check — claim verification
• knowledge-management — note organisation
• systems-thinking — for systemic research questions
• root-cause-analysis — for failure-mode research

---

🎯 COMPLETED: [SKILL:academic] [academic work delivered] 🗣️ CUSTOM COMPLETED: [SKILL:academic] [Academic work done]