research-methodology

Overview

This skill provides comprehensive research methodologies for conducting high-quality, systematic investigations across web sources, documentation, and code. It encompasses planning multi-step research, evaluating source credibility, cross-referencing information, and synthesizing findings into actionable insights.

Core Research Principles

1. Systematic Investigation

Multi-Step Research Process:

1. Define Goal: Clearly articulate what you need to learn
2. Map Gaps: Identify what you don't know
3. Plan Steps: Break into specific, answerable questions
4. Execute Searches: Systematically gather information
5. Verify Claims: Cross-reference and validate
6. Synthesize: Combine findings into coherent narrative
7. Document: Provide citations and enable verification

Key Insight from Research Best Practices:

"Claude operates agentically, conducting multiple searches that build on each other while determining exactly what to investigate next. It explores different angles of your question automatically and works through open questions systematically."

Apply this by:

• Conducting sequential searches that build on each finding
• Adapting search strategy based on initial results
• Exploring multiple perspectives on each question
• Following information threads to deeper understanding

2. Search Strategy Design

Query Construction:

Bad Query:  "I2C"
Good Query: "I2C vs SPI protocol comparison embedded systems"
Better Query: "I2C vs SPI trade-offs Raspberry Pi hardware communication"

Query Patterns:

• Overview Queries: "[Topic] overview | introduction | basics"
• Comparison Queries: "[Option A] vs [Option B] pros cons comparison"
• Specification Queries: "[Technology] datasheet | specification | documentation"
• Problem-Solution Queries: "[Problem] solution | fix | troubleshooting | best practices"
• Authority Queries: "[Topic] site:ieee.org OR site:.edu OR datasheet"

Progressive Depth:

1. Start broad: "Authentication methods web applications"
2. Narrow down: "OAuth 2.0 vs JWT authentication Node.js"
3. Get specific: "JWT refresh token rotation best practices security"
4. Verify: "JWT security vulnerabilities OWASP recommendations"

3. Source Evaluation Framework

Source Credibility Hierarchy:

Tier 1: Authoritative (90-100 credibility)

• Technical: Official datasheets, IEEE/ACM/W3C/IETF standards, official framework docs
• Design/UX: Nielsen Norman Group, Material Design, Apple HIG, WCAG standards
• General: Wikipedia (for overview), Britannica, .edu/.gov with peer review
• Examples: NXP datasheets, React docs, nngroup.com, Wikipedia

Tier 2: Reliable (70-89 credibility)

• Technical: High-reputation Stack Overflow (200+ votes), major tech blogs, well-maintained GitHub repos
• Design/UX: Smashing Magazine, A11Y Project, CSS-Tricks, established design blogs
• Business/Product: Product Hunt, IndieHackers, established tech publications
• General: Quality blogs with demonstrated expertise, educational content
• Examples: StackOverflow accepted answers, smashingmagazine.com, css-tricks.com

Tier 3: Supplementary (50-69 credibility)

• Technical: Medium articles from recognized authors, specialized community forums
• Design/UX: UX Planet, Design Modo, designer portfolios on Dribbble/Behance
• Business/Product: TechCrunch, Hacker News discussions, startup blogs
• General: Tutorial sites with recent content, dev.to articles
• Examples: Quality Medium articles, recent tutorials, community discussions

Tier 4: Questionable (0-49 credibility)

• Content farms with generic information (avoid)
• Outdated tutorials (> 5 years old for fast-moving tech)
• Sites with no author attribution or expertise
• Yahoo Answers, low-quality Quora responses
• AI-generated content without verification
• Note: W3Schools (known for outdated info - use MDN instead)

Evaluation Criteria:

• Authority: Is the author/publisher an expert in this domain?
• Recency: Is the information current? (Check publish date)
• Peer Review: Has this been reviewed or validated by others?
• Citations: Does the source cite its own sources?
• Bias: Are there conflicts of interest or commercial bias?
• Consistency: Does this align with other authoritative sources?

4. Citation Management

Proper Citation Format:

**Claim**: "I2C supports multi-master configuration with clock stretching."

**Source**: NXP I2C-bus specification and user manual (UM10204)
**URL**: https://www.nxp.com/docs/en/user-guide/UM10204.pdf
**Section**: Section 3.1.9, Page 10
**Accessed**: 2024-01-15
**Credibility**: Authoritative (manufacturer specification)

Citation Best Practices:

• Always include URL for web sources
• Note access date (important for changing content)
• Include section/page numbers for long documents
• Provide direct quotes for critical technical claims
• Link claims to specific sources (not generic attribution)
• Verify URLs are accessible before citing

Red Flags:

• Claims without any source attribution
• Vague citations like "according to research"
• Broken or inaccessible URLs
• Sources that don't actually support the claim
• Over-reliance on single source

5. Cross-Reference Verification

Triangulation Method:

For critical technical claims, verify across 3 independent sources:

Example: "I2C maximum speed is 3.4 Mbps (High-Speed mode)"

Source 1: NXP I2C specification (authoritative) ✓ Source 2: Wikipedia I2C article (reliable, cites spec) ✓ Source 3: Embedded.com tutorial (supplementary) ✓

Confidence: High (authoritative + 2 reliable sources confirm)

Handling Contradictions:

If sources contradict:

1. Identify the most authoritative source
2. Check recency (newer may supersede older)
3. Look for context differences (e.g., different versions)
4. Explicitly note the contradiction in findings
5. Recommend seeking expert clarification if critical

Example:

Finding: "I2C pull-up resistor values"
Source A (datasheet): 2.2kΩ - 4.7kΩ recommended
Source B (forum): "Always use 4.7kΩ"

Resolution: Datasheet is authoritative. Resistor value depends on
bus capacitance and speed. 4.7kΩ is common default, but 2.2kΩ may
be needed for high-speed or long traces. Trust datasheet formula.

Research Workflow Patterns

Pattern 1: Technology Comparison Research

Use When: Comparing options (libraries, protocols, architectures)

Steps:

1. Define comparison criteria (speed, complexity, compatibility, cost)
2. Research each option independently (avoid comparison articles initially)
3. Build trade-off matrix with sources for each cell
4. Look for authoritative comparisons to validate your matrix
5. Identify use-case recommendations
6. Synthesize with clear "when to use" guidance

Deliverable: Trade-off matrix + recommendations with conditions

Pattern 2: Problem-Solution Research

Use When: Debugging, troubleshooting, finding best practices

Steps:

1. Precisely define the problem (symptoms, context, constraints)
2. Search for exact error messages or symptoms
3. Find multiple solutions and their trade-offs
4. Verify solutions work in similar contexts
5. Identify root cause if possible
6. Document solution with rationale and caveats

Deliverable: Root cause + prioritized solutions with trade-offs

Pattern 3: Learning Research

Use When: Learning a new technology, framework, or concept

Steps:

1. Start with official documentation (overview/getting started)
2. Find authoritative tutorials or courses
3. Identify common patterns and best practices
4. Learn anti-patterns and pitfalls to avoid
5. Find real-world examples and case studies
6. Summarize mental model and key principles

Deliverable: Mental model + key concepts + resources for deep dive

Pattern 4: Specification Research

Use When: Need exact technical specifications (hardware, protocols, APIs)

Steps:

1. Find official datasheet or specification document
2. Extract exact values, constraints, capabilities
3. Cross-reference with manufacturer application notes
4. Look for errata or known issues
5. Find practical implementation examples
6. Document gotchas and edge cases

Deliverable: Specification summary + practical considerations

Pattern 5: Security Research

Use When: Evaluating security implications or best practices

Steps:

1. Identify security concerns and threat model
2. Consult OWASP guidelines or security frameworks
3. Find CVE reports or known vulnerabilities
4. Research mitigation techniques
5. Cross-reference with current best practices
6. Verify recommendations are current (security evolves fast)

Deliverable: Threat assessment + mitigation strategies with sources

Pattern 6: Design & UX Research

Use When: Improving visual design, user experience, or interface patterns

Steps:

1. Research current design trends and best practices
2. Gather visual examples and inspiration from leading implementations
3. Identify UX principles and patterns for the domain
4. Analyze competitor designs and user feedback
5. Consult accessibility guidelines (WCAG, a11y best practices)
6. Synthesize design recommendations with rationale

Deliverable: Design trends analysis + UX recommendations + visual examples

Authoritative Sources:

• Nielsen Norman Group (nngroup.com)
• Smashing Magazine (smashingmagazine.com)
• Material Design (material.io)
• Apple Human Interface Guidelines
• A11Y Project (a11yproject.com)

Pattern 7: Idea Generation Research

Use When: Seeking innovative features, new capabilities, or project direction

Steps:

1. Survey similar projects, tools, and competitors
2. Identify unique features and innovative capabilities
3. Research emerging technologies and industry trends
4. Explore adjacent problem spaces and use cases
5. Analyze user pain points and unmet needs
6. Brainstorm novel combinations and applications

Deliverable: Innovation opportunities + feasibility assessment + implementation paths

Useful Sources:

• Product Hunt (producthunt.com)
• GitHub Trending (github.com/trending)
• Hacker News (news.ycombinator.com)
• IndieHackers (indiehackers.com)
• Industry blogs and newsletters

Pattern 8: Competitive Analysis Research

Use When: Understanding market landscape, positioning, or differentiation strategy

Steps:

1. Identify direct and indirect competitors
2. Analyze feature sets, pricing, and positioning
3. Assess strengths, weaknesses, and unique selling points
4. Identify market gaps and opportunities
5. Research user reviews and pain points
6. Synthesize positioning and differentiation strategy

Deliverable: Competitor matrix + market gap analysis + positioning recommendations

Useful Sources:

• G2 (g2.com)
• Capterra (capterra.com)
• AlternativeTo (alternativeto.net)
• User review sites and forums

Pattern 9: General Knowledge & Learning Research

Use When: Building understanding, learning concepts, or improving project knowledge

Steps:

1. Start with authoritative foundational sources (Wikipedia, educational content)
2. Identify fundamental principles and mental models
3. Gather diverse perspectives from multiple sources
4. Find practical applications and real-world examples
5. Identify common pitfalls and anti-patterns
6. Synthesize actionable insights for project improvement

Deliverable: Mental model + key concepts + practical applications + actionable insights

Authoritative Sources:

• Wikipedia (wikipedia.org)
• Educational institutions (.edu domains)
• Official documentation and guides
• Established technical publications
• Expert blog posts with demonstrated expertise

Research Quality Metrics

Completeness Checklist

• All research questions answered
• High-confidence answers for critical questions
• Alternatives explored (not just first option found)
• Trade-offs identified and documented
• Edge cases and limitations noted
• Recommendations are actionable
• Sources are authoritative and cited
• Contradictions resolved or flagged
• Next steps or gaps clearly identified

Quality Indicators

High-Quality Research (80-100):

• 3+ authoritative sources for critical claims
• All technical specifications verified
• Trade-offs clearly articulated
• Recommendations have clear conditions
• Citations are precise and accessible
• No unresolved contradictions

Moderate-Quality Research (60-79):

• Mix of authoritative and reliable sources
• Most claims verified
• Basic trade-offs identified
• General recommendations provided
• Citations present but may lack detail
• Minor contradictions noted

Low-Quality Research (0-59):

• Reliance on questionable sources
• Claims not verified
• Missing trade-offs
• Vague or generic recommendations
• Poor citation practices
• Unresolved contradictions

Common Research Pitfalls

Pitfall 1: Confirmation Bias

Problem: Only searching for information that confirms initial assumptions Solution: Explicitly search for counter-arguments and alternative perspectives

Pitfall 2: Recency Bias

Problem: Assuming newest information is always best Solution: Check if core principles have changed or just implementation details

Pitfall 3: Authority Bias

Problem: Trusting everything from a high-reputation source Solution: Even authoritative sources can have errors; cross-reference critical claims

Pitfall 4: Depth-Breadth Imbalance

Problem: Either too shallow (surface-level) or too deep (rabbit hole) Solution: Define success criteria upfront; know when you have "enough" information

Pitfall 5: Outdated Information

Problem: Using information from 5+ years ago for fast-evolving tech Solution: Always check publication date; prefer last 1-2 years for best practices

Pitfall 6: Generic Tutorials

Problem: Following tutorials that don't match your specific context Solution: Verify tutorial assumptions match your environment (versions, platform, etc.)

Pitfall 7: Missing Context

Problem: Applying advice without understanding when it applies Solution: Always extract the "when to use" and "when NOT to use" conditions

Research Report Structure

Executive Summary

• 2-3 sentences: What was researched, key findings, main recommendation

Research Questions

• List of specific questions addressed

Key Findings

For each finding:

• Question: [Original question]
• Answer: [Synthesized answer]
• Confidence: High/Medium/Low
• Sources: [Cited sources with URLs]
• Notes: [Caveats, context, additional details]

Trade-Off Matrix (if applicable)

Criterion	Option A	Option B	Winner
[criterion]	[details] [[source]]	[details] [[source]]	[Option X]

Recommendations

For each recommendation:

• Recommendation: [What to do]
• Rationale: [Why this is recommended]
• Conditions: [When this applies]
• Alternatives: [What to do if conditions don't match]
• Sources: [Supporting citations]

Remaining Gaps

• List of unanswered questions or low-confidence findings
• Suggested next steps for additional research

Sources Consulted

Organized by credibility tier:

• Authoritative sources: [list with URLs]
• Reliable sources: [list with URLs]
• Supplementary sources: [list with URLs]

Integration with Pattern Learning

After completing research:

1. Store research pattern with topic, sources, and outcomes
2. Track which sources were most valuable
3. Record time spent vs. value gained
4. Note effective search queries for similar future research
5. Learn which research workflow pattern worked best

This enables:

• Faster research on similar topics in future
• Better source selection based on past success
• Improved search query formulation
• Recognition of similar research needs

When to Apply

Use this skill when:

• Technical Research: Exploring technical topics, specifications, protocols, or implementations
• Design & UX Research: Improving visual design, user experience, or interface patterns
• Idea Generation: Seeking innovative features, new capabilities, or project direction
• Competitive Analysis: Understanding market landscape, positioning, or differentiation
• General Knowledge: Building understanding, learning concepts, or improving project knowledge
• Problem Solving: Troubleshooting complex issues or finding solutions
• Decision Making: Evaluating multiple options or approaches with trade-off analysis
• Verification: Verifying claims, cross-referencing information, or assessing credibility
• Learning: Learning new technologies, frameworks, methodologies, or best practices
• Synthesis: Combining information from multiple sources into actionable insights

This methodology ensures high-quality, systematic research across all domains—technical, creative, strategic, and general—producing reliable, actionable insights with proper attribution and verification.