Qualitative Research Skill

Complete workflow for designing, conducting, analyzing, and reporting qualitative research.

Workflow Overview

Phase 1: Objective Clarification -> Define research questions
Phase 2: Participant Profiling   -> Identify and profile interviewees
Phase 3: Protocol Design         -> Create interview guide
Phase 4: Data Collection         -> Conduct interviews
Phase 5: Thematic Analysis       -> Code, theme, interpret
Phase 6: Report Writing          -> Produce final report

Execute phases sequentially. Each phase feeds into the next.

Phase 1: Objective Clarification

Goal: Transform a research topic into precise qualitative research questions.

Steps

1. Understand the research topic: What phenomenon needs exploration?
2. Identify the gap: What can quantitative methods NOT answer?
3. Draft RQs: Formulate 3-5 interpretive research questions starting with "How" or "Why"
4. Choose paradigm: Select interpretivism / constructivism / critical theory
5. Select method: Semi-structured interview (default), focus group, ethnography, case study

Quality Checklist

• [ ] RQs are interpretive, not confirmatory
• [ ] RQs specify whose perspective is sought
• [ ] Method chosen matches research aim
• [ ] Qualitative component is justified (not duplicating quant work)

Output

• Research objective statement
• 3-5 research questions
• Methodological justification (2-3 paragraphs)

Phase 2: Participant Profiling

Goal: Define who to interview and why they matter.

Steps

1. Define selection criteria: Who has experienced the phenomenon? Who holds key perspectives?
2. Determine sample strategy: Purposeful sampling (default), snowball, maximum variation
3. Set sample size: 6-12 for interviews (until saturation), 8-12 for focus groups
4. Create personas: Build 6-10 detailed participant personas covering diverse perspectives

Persona Template

[PX] "Name" — Role | Perspective Label
- Age, gender, education, years of experience
- Position, industry, organization type
- AI/phenomenon usage frequency and depth
- Typical stance: "quote capturing their worldview"
- Why selected: what unique insight they provide
- Interview focus: what topics to probe deeply

Coverage Checklist

• [ ] Range of experience levels (junior to senior)
• [ ] Multiple industries / contexts
• [ ] Both "insiders" and "outsiders" to the phenomenon
• [ ] Both supporters and skeptics
• [ ] Information richness over representativeness

Output

• 6-10 participant personas with rationale
• Selection criteria document
• Sampling strategy justification

Phase 3: Protocol Design

Goal: Create a semi-structured interview guide.

Structure (See references/templates.md for full template)

Question Design Principles

• Use open-ended questions ("Tell me about..." "How did you..." "What was that like?")
• Include projective techniques ("What advice would you give..." "Three words to describe...")
• Prepare follow-up probes for each question ("Can you give an example?" "What did you feel?")
• Sequence from concrete to abstract, from past to future
• End with an emotional closing question

Output

• Semi-structured interview protocol (12-15 questions + probes)
• Interview logistics plan (duration, format, recording, transcription)

Phase 4: Data Collection

Goal: Conduct rich, in-depth simulated interviews.

Steps

1. Simulate interviews: For each persona, generate a realistic verbatim transcript
2. Ensure richness: Each transcript should include:
   - Specific incidents and stories (not just opinions)
   - Emotional language and pauses ("...", "(deep breath)")
   - Contradictions and ambivalences
   - Metaphors and analogies used by participants
   - Follow-up responses to probes
3. Mark key quotes: Highlight statements that directly illuminate research questions
4. Document field notes: Record interviewer observations about tone, hesitations, body language

Interview Simulation Quality Standards

• Minimum 6 interviews, target 8-10
• Each interview 2000-4000 characters of transcript
• At least 3 "breakthrough moments" per interview (unexpected insights)
• At least 1 contradiction or ambivalence per interview
• Varying emotional registers across interviews

Output

• 6-10 verbatim interview transcripts
• Field notes for each interview
• Initial impression memos

Phase 5: Thematic Analysis

Goal: Systematically identify, analyze, and report patterns in qualitative data.

Method: Braun & Clarke (2006) six-phase thematic analysis.

The Six Phases

Phase 5a: Familiarization
- Read all transcripts multiple times
- Write initial notes and impressions
- Record patterns, contradictions, surprises

Phase 5b: Initial Coding
- Code EVERYTHING of potential interest (semantic + latent)
- Aim for 80-150 initial codes across all transcripts
- Use descriptive labels that capture the essence
- Include both explicit content and implicit meanings

Phase 5c: Searching for Themes
- Collate codes into potential themes
- Create a thematic map showing relationships
- Look for: patterns, contradictions, outliers, turning points

Phase 5d: Reviewing Themes
- Check themes against coded extracts
- Check themes against entire dataset
- Split, combine, or discard themes as needed
- Ensure themes tell a coherent story

Phase 5e: Defining and Naming Themes
- Write a 2-3 paragraph definition for each theme
- Create a compelling, concise name
- Identify the "story" each theme tells
- Order themes to create narrative flow

Phase 5f: Producing the Report
- Write theme narratives with vivid quotes
- Show analytic moves (not just describe)
- Link back to research questions
- Discuss theoretical and practical implications

Quality Standards for Thematic Analysis

• [ ] 80-150 initial codes extracted
• [ ] 5-8 overarching themes identified
• [ ] Each theme supported by data from 3+ participants
• [ ] Contradictions and negative cases acknowledged
• [ ] Quotes selected for vividness AND typicality
• [ ] Analytic narrative goes beyond description to interpretation

Output

• Complete codebook with definitions
• Thematic map
• 5-8 defined themes with narrative and quotes
• Analysis memo connecting themes to RQs

See references/methodology.md for detailed coding procedures and references/templates.md for codebook and thematic map templates.

Phase 6: Report Writing

Goal: Produce a rigorous, compelling qualitative research report.

Report Structure

1. Abstract (200-300 words)
   - Purpose, method, key findings, implications

2. Introduction (1-2 pages)
   - Research background and gap
   - Research questions
   - Methodological justification

3. Method (1-2 pages)
   - Research design and paradigm
   - Participant selection and profiles
   - Data collection procedures
   - Analytic approach
   - Ethics and rigor criteria

4. Findings (6-10 pages)
   - One section per theme (5-8 themes)
   - Each theme: definition, narrative, 3-5 vivid quotes
   - Analytic commentary connecting themes to RQs

5. Discussion (2-3 pages)
   - Theoretical contributions
   - Practical implications
   - Limitations
   - Future research directions

6. Conclusion (1 page)
   - Core insight
   - Closing reflection

Writing Quality Standards

• [ ] Every claim supported by participant quotes
• [ ] Quotes are vivid, contextualized, and attributed
• [ ] Analytic narrative goes beyond description
• [ ] Contradictions and ambivalences acknowledged
• [ ] Themes are interconnected, not isolated
• [ ] Report answers ALL research questions
• [ ] Implications are specific and actionable

Rigor Criteria (Trustworthiness)

• Credibility: Member checking, peer debriefing, thick description
• Transferability: Rich context, multiple cases, diverse participants
• Dependability: Audit trail, process documentation
• Confirmability: Reflexivity statement, bracketing

Output

• Complete qualitative research report (8000-15000 words)
• Codebook and thematic map as appendix
• Process documentation for audit trail

Key Principles

1. Depth over breadth: 6 rich interviews > 20 shallow ones
2. Participant voice central: Let quotes drive the narrative
3. Embrace contradiction: Real people are inconsistent — that's data
4. Context is everything: Always interpret within participant's context
5. Reflexivity matters: Acknowledge researcher's perspective and influence
6. Saturation not sample size: Stop when new interviews yield no new themes
7. Show the work: Document the analytic journey, not just the destination