research-methodology（研究方法）

Research Methodology Skill

This skill provides systematic procedures for gathering, evaluating, and organizing research for book writing.

When to Use This Skill

• Beginning research on a new book topic
• Evaluating source credibility and relevance
• Organizing research notes and citations
• Fact-checking claims during writing or editing
• Managing bibliography and references
• Planning research timelines for rapid book generation (target: 1 week)
• Managing digital research assets

Research Workflow

Phase 1: Planning (10% of research time)

1. Define Research Questions

- What are the core questions this book answers? - What subsidiary questions emerge from the core? - What knowledge gaps need filling?

1. Set Source Targets

- Quality over quantity: Use as many sources as needed for conceptual reliability - Guideline: 5-15 sources per major section (adjust based on topic depth) - Niche topics may have fewer authoritative sources — that's acceptable - Balance: Prioritize Tier 1, supplement with Tier 2, avoid Tier 3 - Aim for diverse perspectives (avoid echo chambers)

1. Establish Timeline

- Research sprint: 3-4 hours per chapter section (rapid generation mode) - Full book research: 5-7 days maximum - Source evaluation: real-time (as you discover sources) - Synthesis: continuous (don't wait until end)

Phase 2: Discovery (40% of research time)

1. Initial Exploration

- START ONLY WITH AUTHORITATIVE SOURCES: Academic databases (Google Scholar, JSTOR, PubMed), official documentation, peer-reviewed journals - NEVER use Wikipedia or user-editable platforms as research sources (content can be edited by anyone, unreliable) - Identify key terms, concepts, seminal works from authenticated sources - Map the intellectual landscape using Tier 1 sources

1. Deep Dive

- Follow citations backward (what influenced this?) - Follow citations forward (who built on this?) - Use academic databases: Google Scholar, JSTOR, PubMed, IEEE Xplore - Access official documentation and technical standards

1. Authority Identification

- Who are the recognized authorities in this field? - What institutions lead this research area? - Which papers/books are most cited by other Tier 1 sources?

Phase 3: Evaluation (20% of research time)

Apply source evaluation criteria (see below) to all discovered sources.

Phase 4: Synthesis (30% of research time)

1. Pattern Recognition

- What themes emerge across sources? - Where do sources agree/disagree? - What narratives compete?

1. Knowledge Integration

- Connect findings to research questions - Identify supporting evidence for key claims - Document gaps and uncertainties

Source Evaluation Criteria

Tier 1: Highly Authoritative (Prioritize)

• Academic Journals: Peer-reviewed papers in reputable journals
• Academic Books: Published by university presses or major academic publishers
• Official Documentation: Government reports, technical standards, official statistics
• Expert Sources: Published works by recognized domain experts

Verification Checklist:

• [ ] Author has relevant PhD or equivalent expertise
• [ ] Published by recognized institution/press
• [ ] Peer-reviewed or editorially reviewed
• [ ] Cited by other Tier 1 sources
• [ ] Methodology clearly documented

Tier 2: Reliable (Use with verification)

• Reputable News: Major newspapers, established news organizations
• Trade Publications: Industry-specific magazines and journals
• Professional Blogs: Recognized experts in their field
• Technical Documentation: Official software/product documentation

Verification Checklist:

• [ ] Cross-referenced with at least one Tier 1 source
• [ ] Author expertise verified through credentials or body of work
• [ ] No obvious bias or conflicts of interest
• [ ] Recent publication (within 5 years for technical topics)

Tier 3: Supplementary (Avoid)

• General Blogs: Personal opinion pieces (only if from recognized experts)
• Social Media: Trends and public opinion data (only for cultural context)
• Opinion Pieces: Clearly labeled as commentary (only from credentialed authors)

Usage Guidelines:

• Never cite as primary source
• Never use as factual reference
• Always trace to primary Tier 1 source before including in manuscript

Sources to Avoid

• Wikipedia and user-editable platforms (anyone can edit, no authentication, unreliable)
• Content farms (sites generating low-quality content for SEO)
• Outdated information (>5 years unless historical context)
• Sources with clear undisclosed bias
• Anonymous or unverifiable authors
• Predatory journals (check DOAJ, Beall's List)
• Press releases without independent verification
• User forums and Q&A sites (Reddit, Quora, Stack Overflow for facts)
• Crowdsourced content without editorial oversight

Citation Format Standards

APA 7th Edition

• In-text: (Author, Year) or Author (Year)
• Single author: (Smith, 2020)
• Two authors: (Smith & Jones, 2020)
• Three or more: (Smith et al., 2020)
• Direct quote: (Smith, 2020, p. 42)

MLA 9th Edition

• In-text: (Author Page) or Author (Page)
• Single author: (Smith 42)
• Two authors: (Smith and Jones 42)
• Three or more: (Smith et al. 42)

Chicago 17th Edition

• In-text: Superscript numbers with corresponding footnotes/endnotes

Fact-Checking Procedures

Verification Workflow

1. Identify Claims Requiring Verification

- Mark all factual statements in manuscript - Prioritize: statistics, dates, quotes, technical facts - Tag with confidence level: [VERIFY-HIGH], [VERIFY-MEDIUM], [VERIFY-LOW]

1. Cross-Reference

- Check claim against minimum 2 independent sources - For critical claims: require 3+ sources - Document which sources confirm/contradict

1. Document Confidence

- High: 3+ Tier 1 sources agree, recent data - Medium: 2 Tier 1 or 3+ Tier 2 sources agree - Low: Single source or conflicting sources - Flag: Unverifiable or conflicting

Verification Standards

Research Organization

Directory Structure

research/
├── [topic-1]/
│   ├── primary-sources.md
│   ├── synthesis.md
│   ├── bibliography.md
│   ├── fact-checks.md
│   └── assets/
├── [topic-2]/
│   └── ...
├── cross-references.md
└── research-log.md

Quality Assurance Checklist

Before Moving to Writing Phase

Source Quality:

• [ ] Sufficient sources for conceptual reliability (5-15 per section, topic-dependent)
• [ ] Tier 1 sources prioritized (majority when available)
• [ ] No sources from "avoid" category (NO Wikipedia, user-editable platforms)
• [ ] All sources authenticated and verified
• [ ] Diverse perspectives represented (avoid echo chamber)
• [ ] Source gaps documented if topic has limited authoritative coverage

Citation Completeness:

• [ ] Citations formatted correctly and consistently
• [ ] Access dates recorded for all web sources
• [ ] DOIs included for all academic papers (where available)
• [ ] Page numbers noted for all direct quotes

Time Efficiency:

• [ ] Research time target met (3-4 hours per chapter section)
• [ ] No time wasted on Wikipedia or unverified sources
• [ ] Citation metadata captured immediately (no backtracking)

Common Pitfalls to Avoid

1. Over-reliance on Secondary Sources: Always trace to primary source
2. Confirmation Bias: Actively seek sources that challenge assumptions
3. Using Wikipedia: Start ONLY with academic databases, peer-reviewed journals
4. Outdated Information: Check publication dates for technical topics
5. Missing Citations: Record source immediately
6. Incomplete Metadata: Capture all citation elements on first pass
7. Assuming AI Accuracy: Verify all AI-provided facts with primary sources

Time Budget Quick Reference

Chapter section (3,000-5,000 words): 3-4 hours
Major chapter (10,000-15,000 words): 8-12 hours
Full book research varies by depth:
  - Light research (established topics): 30-40 hours (5-7 days)
  - Standard research (mixed sources): 50-70 hours (1-2 weeks)
  - Deep research (novel/technical): 80-120 hours (2-4 weeks)

Efficiency keys:
- Academic databases only (no Wikipedia browsing)
- Parallel research (multiple topics simultaneously)
- Immediate citation capture (no backtracking)

Timeline Realism

⚠️ Note: Research timelines depend heavily on: - Topic familiarity (established vs. cutting-edge) - Source availability (abundant vs. niche) - Depth required (overview vs. comprehensive)

Adjust expectations based on actual source landscape, not arbitrary deadlines.

Skill Version: 1.2.0 Last Updated: 2025-11-27 Maintained By: Universal Pedagogical Engine Team