Research Methods Skill

Literature search, source screening, annotated bibliography creation.

🎯 Skill Purpose

Find, evaluate, and organize academic sources for thesis research.

Literature Search Strategy

Phase 1: Query Development

Start with research question → extract keywords → create search queries.

Example:

RQ: How should Equinor govern AI deployment?

Keywords:
- Primary: AI governance, artificial intelligence, energy sector
- Secondary: ethics, workforce, sustainability
- Specific: Equinor, Norway, oil and gas

Queries:
- "AI governance" AND "energy sector"
- "artificial intelligence" AND "oil and gas" AND ethics
- Equinor AND (AI OR "artificial intelligence")

Phase 2: Database Selection

Phase 3: Search Execution

Best Practices:

• •Start broad, then narrow
• •Use Boolean operators (AND, OR, NOT)
• •Apply date filters (last 5 years for empirical)
• •Check cited-by lists for seminal works
• •Check reference lists for related work

Source Evaluation (CRAAP Test)

Scoring Guide

• •20-25: Excellent source
• •15-19: Good source
• •10-14: Use with caution
• •<10: Avoid or heavily qualify

Source Screening Process

Tier 1: Title & Abstract (30 seconds)

• •Does it relate to research question?
• •Is it the right scope/level?
• •Quick Include/Maybe/Exclude

Tier 2: Skim (3 minutes)

• •Read intro and conclusion
• •Check methodology section
• •Review key figures/tables
• •Confirm relevance

Tier 3: Deep Read (15-30 minutes)

• •Full read for included sources
• •Take notes using annotation template
• •Extract quotes with page numbers

Annotated Bibliography Template

For each source:

## Author (Year)
**Full Citation (APA 7):**
[Complete reference]

**Summary (2-3 sentences):**
What the source argues and how.

**Methodology:**
How the research was conducted.

**Key Findings:**
Main results or arguments.

**Relevance to Thesis:**
How this connects to your RQ.

**Quotes to Use:**
"Direct quote" (p. X)

**Limitations:**
Any weaknesses or gaps.

**Tags:** #theme1 #theme2

Example Entry

## Floridi (2019)
**Full Citation:**
Floridi, L., & Cowls, J. (2019). A unified framework of five principles for AI in society. *Harvard Data Science Review*, 1(1). https://doi.org/10.1162/99608f92.8cd550d1

**Summary:**
Synthesizes existing AI ethics frameworks into five principles: beneficence, non-maleficence, autonomy, justice, and explicability.

**Methodology:**
Comparative analysis of existing frameworks (EU, IEEE, etc.).

**Key Findings:**
All major AI ethics frameworks reduce to these five principles; tensions between them require contextual navigation.

**Relevance:**
Provides analytical framework for evaluating Equinor's AI governance approach.

**Quotes:**
"The ethical use of AI requires not just compliance but active stewardship" (p. 7)

**Limitations:**
Framework is abstract; limited guidance on implementation.

**Tags:** #theoretical-framework #AI-ethics #principles

Source Organization

Recommended Tools

• •Zotero (free, open source)
• •Mendeley (free, owned by Elsevier)
• •EndNote (paid, institutional)
• •Notion/Obsidian (manual)

Folder Structure

Sources/
├── 1_Theoretical/
│   ├── AI_Ethics/
│   └── Governance/
├── 2_Empirical/
│   ├── Case_Studies/
│   └── Industry_Reports/
├── 3_Methodology/
└── 4_Background/

Tagging System

• •By theme: #AI-ethics, #governance, #sustainability
• •By use: #framework, #evidence, #background
• •By status: #read, #to-read, #cited

Search Log Template

Track your searches:

## Search Log

### Search 1
- **Date:** 2026-01-22
- **Database:** OpenAlex
- **Query:** "AI governance" AND "energy"
- **Filters:** 2020-2025, English
- **Results:** 147
- **Screened:** 30
- **Included:** 8

### Search 2
...

Finding Seminal Works

Strategies:

1. •Check most-cited papers in search results
2. •Look at reference lists of recent papers
3. •Search for "review" or "meta-analysis"
4. •Check textbooks in the field
5. •Ask experts (advisor, professors)

Gray Literature

For non-peer-reviewed but credible sources:

Quality Checklist

• • Searched 3+ databases
• • Used Boolean operators
• • Applied date filters appropriately
• • Screened systematically (not cherry-picked)
• • Evaluated with CRAAP criteria
• • Documented search strategy
• • Annotated key sources
• • Organized in reference manager
• • Identified 3-5 seminal works
• • Balanced primary and secondary sources