Qualitative Research

Overview

Systematic procedure for qualitative data analysis including coding, thematic analysis, grounded theory, and interview techniques

Steps

Step 1: Design the qualitative study

Plan the overall approach to the qualitative inquiry:

CLARIFY THE RESEARCH QUESTION

Good qualitative questions:
- Ask “how” or “what” (not “how many” or “does X cause Y”)
- Are open and exploratory
- Focus on meaning, experience, or process
- Cannot be answered with yes/no
Examples:
- “How do first-generation students experience college transition?”
- “What factors shape decision-making in X context?”
- “How do participants make sense of their experiences with Y?”
SELECT THEORETICAL ORIENTATION

Phenomenology:
- Focus: Lived experience of a phenomenon
- Goal: Describe essence of the experience
- Questions: “What is it like to experience X?”
- Analysis: Identify essential structures of experience
Grounded Theory:
- Focus: Process or action
- Goal: Develop theory from data
- Questions: “What process explains X?”
- Analysis: Build theory through constant comparison
Ethnography:
- Focus: Culture-sharing group
- Goal: Describe and interpret cultural patterns
- Questions: “How does this group operate?”
- Analysis: Identify cultural themes and patterns
Case Study:
- Focus: Bounded case or cases
- Goal: In-depth understanding of case
- Questions: “What can we learn from this case?”
- Analysis: Within-case and cross-case analysis
Narrative:
- Focus: Stories and life experiences
- Goal: Understand through storytelling
- Questions: “What story do participants tell about X?”
- Analysis: Restorying, structural analysis
DETERMINE SAMPLING STRATEGY

Purposive sampling approaches:
- Maximum variation: Diverse cases for range of experience
- Homogeneous: Similar cases for depth
- Critical case: Cases that are especially informative
- Theory-driven: Cases to develop/test emerging theory
- Snowball: Participants refer others
Sample size:
- Aim for saturation (no new themes emerging)
- Typical ranges: 5-25 for phenomenology, 20-30 for grounded theory
- Quality of data matters more than quantity
PLAN DATA COLLECTION

Data types:
- Individual interviews
- Focus groups
- Observation/field notes
- Documents and artifacts
- Visual methods
Consider triangulation: Multiple data sources for robustness

Step 2: Develop interview and data collection protocols

Create guides for systematic data collection:

INTERVIEW GUIDE DEVELOPMENT:

Structure
- Introduction (5-10 min)
  - Rapport building
  - Study explanation
  - Consent process
  - Recording permission
- Opening questions (5-10 min)
  - Easy, non-threatening
  - Get participant talking
  - Example: “Tell me about your background with X”
- Main questions (30-45 min)
  - Core topics of inquiry
  - Open-ended, non-leading
  - Ordered from general to specific
  - Include probes for each question
- Closing (5-10 min)
  - Wrap-up questions
  - “Is there anything else?”
  - Thank participant
  - Explain next steps
Question Types

Descriptive questions:
- “Tell me about…”
- “Describe your experience with…”
- “Walk me through…”
Structural questions:
- “How would you categorize…?”
- “What types of X are there?”
- “How do you organize your thinking about…?”
Contrast questions:
- “How does X compare to Y?”
- “What’s the difference between…?”
Evaluation questions:
- “What do you think about…?”
- “How do you feel when…?”
Probes

Elaboration: “Tell me more about that” Clarification: “What do you mean by…?” Example: “Can you give me an example?” Consequence: “What happened then?” Comparison: “How does that compare to…?” Silence: (pause and wait)

FOCUS GROUP GUIDE:

Opening activity
- Icebreaker
- Ground rules
- Introductions
Main discussion
- Opening question (everyone answers)
- Key questions (5-6 main topics)
- Activities if appropriate
Closing
- Summary and member check
- Final thoughts
- Thank participants

Moderator skills:

Balance participation
Handle dominant/quiet participants
Manage disagreements
Keep on track without stifling

OBSERVATION PROTOCOL:

What to observe:

Physical setting
Participants and roles
Activities and interactions
Conversations
Subtle factors (tone, body language)
Your own reactions

Field note format:

Descriptive notes (what happened)
Reflective notes (your thoughts)
Date, time, location
Write up fully immediately after

Step 3: Conduct data collection

Execute systematic data gathering:

INTERVIEWING TECHNIQUES:

Rapport building
- Be warm, genuine, interested
- Small talk before recording
- Explain process clearly
- Acknowledge their expertise on their experience
Active listening
- Full attention on participant
- Minimal interruption
- Nod, “mm-hmm” to encourage
- Reflect back what you hear
Following up
- Use probes to go deeper
- Follow interesting threads
- Don’t rush to next question
- Tolerate silence (they may be thinking)
Managing challenges

Talkative participant:
- “That’s helpful. Let me make sure we cover…”
- Redirect gently
Quiet participant:
- More encouragement
- Longer pauses
- Easier questions first
Emotional content:
- Acknowledge feelings
- Offer to pause
- Have resources available if needed
Off-topic:
- “That’s interesting. To get back to…”
- Gently redirect

RECORDING AND DOCUMENTATION:

Audio recording:

Get explicit consent
Test equipment before
Use backup recorder
Note any issues

During interview:

Brief notes (keywords, follow-ups)
Non-verbal observations
Your own reactions

Immediately after:

Write reflective memo
Note initial impressions
Document context
Ideas for follow-up

TRANSCRIPTION:

Verbatim transcription:

Every word spoken
Include filler words (um, uh)
Note pauses, laughter, emotion
Time stamps helpful

Transcription conventions:

[pause] for silences
[laughs] for non-verbal
… for trailing off
CAPS for emphasis
(inaudible) when unclear

Quality check:

Listen while reading transcript
Correct errors
Verify participant identification

ITERATIVE DATA COLLECTION:

In grounded theory especially:

Analyze as you collect
Let early analysis shape later questions
Theoretical sampling: seek specific cases
Continue until saturation

Step 4: Code the data

Systematically code data to identify patterns:

CODING FUNDAMENTALS:

What is a code?

A label for a segment of data
Captures the meaning of that segment
Building block for themes

Types of codes:

Descriptive codes:

Summarize topic of passage
What is this about?
Example: “family support”

In vivo codes:

Participant’s own words
Preserve authentic voice
Example: “just surviving”

Process codes:

Actions and processes
Use gerunds (-ing)
Example: “negotiating identity”

Emotion codes:

Feelings expressed or inferred
Example: “frustrated”

Values codes:

Beliefs and values
Example: “valuing independence”

FIRST CYCLE CODING:

Read through data first
- Get overall sense
- Note initial impressions
- Don’t code yet
Line-by-line coding
- Code every segment
- Stay close to data
- Don’t interpret too much yet
- Generate many codes
Code the same data multiple ways
- Different codes capture different aspects
- Layer codes (descriptive + process + emotion)
Write memos as you code
- Why this code?
- What does it mean?
- How does it connect to other codes?

SECOND CYCLE CODING:

Pattern coding:

Group related codes into categories
Look for patterns across codes
Reduce many codes to fewer categories

Focused coding:

Use most frequent/significant codes
Apply systematically across data
Test codes against new data

Axial coding (grounded theory):

Identify relationships between categories
Conditions, actions, consequences
Build toward theory

Theoretical coding (grounded theory):

Integrate categories around core category
Develop theoretical framework

CODING MECHANICS:

Tools:

Qualitative software (NVivo, Atlas.ti, Dedoose)
Word processor with comments
Spreadsheets
Physical (printed transcripts, highlighters)

Codebook:

Code name
Definition
When to use/not use
Example quotes

Track decisions:

Why codes were merged/split
How definitions evolved
Audit trail

Step 5: Develop themes

Move from codes to themes through thematic analysis:

THEMATIC ANALYSIS PROCESS:

Familiarization
- Immerse yourself in data
- Read and re-read
- Note initial patterns
Generating initial codes
- Systematic coding (Step 4)
- Note potential themes
Searching for themes
- Examine codes for patterns
- Group codes into potential themes
- Create theme-piles
Reviewing themes
- Check themes against coded data
- Check themes against full dataset
- Refine, split, combine as needed
- Two levels:
  - Do codes form coherent theme?
  - Does theme work across dataset?
Defining and naming themes
- Clear definition of each theme
- Capture essence in name
- Identify subthemes if needed
- Write theme descriptions
Writing up
- Analytic narrative
- Evidence (quotes) for each theme
- How themes relate to research question

THEME CHARACTERISTICS:

Good themes:

Capture something important about the data
Relate to research question
Are not just a code (more abstract)
Have internal coherence
Are distinct from other themes

Theme types:

Semantic: Surface-level, descriptive
Latent: Underlying meanings, interpretive

THEME DEVELOPMENT STRATEGIES:

Pattern identification:

What comes up repeatedly?
What do participants emphasize?
What surprises you?

Constant comparison:

Compare within interview
Compare across interviews
Compare to emerging themes

Negative case analysis:

Look for disconfirming evidence
Cases that don’t fit pattern
Refine themes to account for variation

CONSTRUCTING THE THEMATIC MAP:

Organize themes:

Main themes (3-6 typical)
Subthemes under main themes
Relationships between themes

Visual representation:

Thematic map diagram
Shows structure of analysis
Helps identify gaps

Test the structure:

Does it answer research question?
Is it supported by data?
Does it make sense to others?

Step 6: Apply grounded theory methods (if applicable)

Use grounded theory for theory development:

GROUNDED THEORY BASICS:

Goal: Develop theory that is “grounded” in data Process: Simultaneous data collection and analysis Output: Theoretical framework explaining phenomenon

KEY PROCEDURES:

Open coding
- Initial line-by-line coding
- Generate many codes
- Ask: What is happening here?
- Stay close to data
Axial coding
- Relate categories to subcategories
- Identify relationships
- Use coding paradigm:
  - Conditions: What leads to phenomenon?
  - Actions/Interactions: How do people respond?
  - Consequences: What results?
Selective coding
- Identify core category
- Central phenomenon around which theory revolves
- Relate all categories to core
- Fill in gaps

CONSTANT COMPARISON:

Compare:

Data to data (within and across cases)
Data to codes
Codes to codes
Codes to categories
Categories to categories

Purpose:

Develop properties of categories
Ensure consistency
Identify variation

THEORETICAL SAMPLING:

What it is:

Sampling driven by emerging theory
Seek data to develop/test categories
Not representative sampling

Process:

Initial sampling: Start broadly
As categories emerge: Sample to develop them
Late stage: Sample to saturate and test

Examples:

Category needs more properties? Find cases that vary on dimension
Category seems complete? Find negative cases
Relationship unclear? Sample cases at boundaries

THEORETICAL SATURATION:

Saturation means:

No new properties emerging
No new dimensions emerging
Relationships are well established

Evidence of saturation:

New data fits existing categories
Categories are well-developed
Relationships are clear

MEMO WRITING:

Memos are crucial in grounded theory:

Write continuously throughout
Explore ideas about codes and categories
Develop properties and dimensions
Explore relationships

Types:

Code memos: About specific codes
Theoretical memos: Developing theory
Operational memos: Methods decisions

DEVELOPING THE THEORY:

Theory elements:

Core category (central phenomenon)
Main categories and their properties
Relationships between categories
Conditions under which relationships hold

Theory format:

Visual diagram
Propositional statements
Narrative description

Step 7: Ensure trustworthiness

Demonstrate quality and rigor:

TRUSTWORTHINESS CRITERIA:

Credibility (internal validity equivalent)
- Are findings believable?
- Do they ring true?
Strategies:
- Prolonged engagement: Sufficient time with data
- Triangulation: Multiple sources, methods, researchers
- Member checking: Participants review findings
- Peer debriefing: Colleagues review analysis
- Negative case analysis: Address contradictions
Transferability (external validity equivalent)
- Can findings apply elsewhere?
- Not generalization, but informed transfer
Strategies:
- Thick description: Rich detail about context
- Clear sampling: Who was studied
- Comparison to other contexts: Where might this apply?
Dependability (reliability equivalent)
- Would process produce similar results?
- Is process consistent and documented?
Strategies:
- Audit trail: Document all decisions
- Code-recode reliability: Re-code sample, check consistency
- External audit: Outside review of process
Confirmability (objectivity equivalent)
- Are findings shaped by participants, not researcher bias?
- Is researcher influence acknowledged?
Strategies:
- Reflexivity: Acknowledge researcher position
- Audit trail: Evidence linking data to findings
- Multiple coders: Check for agreement

SPECIFIC TECHNIQUES:

Member checking:

Share findings with participants
Do they recognize their experience?
Note any disagreements
Not requiring agreement, but checking resonance

Peer debriefing:

Colleague reviews your analysis
Challenges assumptions
Asks hard questions
Catches blind spots

Triangulation types:

Data triangulation: Multiple sources
Investigator triangulation: Multiple researchers
Theory triangulation: Multiple perspectives
Method triangulation: Multiple methods

Audit trail contents:

Raw data
Data reduction products (codes, themes)
Process notes (how decisions were made)
Materials about researcher position
Preliminary findings to final report

REFLEXIVITY:

What to reflect on:

Your background and how it shapes interpretation
Your relationship to topic and participants
Assumptions you bring
How your presence affected data

Document in:

Reflexive journal
Methods section
Analysis discussion

Step 8: Write findings and report

Present findings in compelling narrative form:

WRITING QUALITATIVE FINDINGS:

Organize by themes
- Each theme is a section
- Define and explain theme
- Provide evidence (quotes)
- Interpret meaning
Use participant quotes effectively

Good quote use:
- Illustrates the theme
- Shows participant voice
- Integrated into narrative
- Not too long (2-4 sentences ideal)
Quote presentation:
- Introduce: Set up the quote
- Present: The quote itself
- Interpret: What it means
Example: “Participants described feeling caught between worlds. As Maria explained, ‘I don’t fully belong in either place. At home I’m too American, at school I’m too Mexican.’ This experience of cultural liminality shaped…”
Balance description and interpretation
- Describe what participants said/did
- Interpret what it means
- Connect to research question
- Connect to theory/literature
Show complexity
- Present main pattern
- Also show variations
- Address contradictions
- Avoid oversimplification

REPORT STRUCTURE (typical):

Introduction:

Research question
Why it matters
What we don’t know

Methods:

Approach and rationale
Participants and sampling
Data collection procedures
Analysis approach
Trustworthiness strategies

Findings:

Organized by theme
Rich quotes
Interpretation

Discussion:

Relation to research question
Relation to literature
Implications
Limitations
Future directions

WRITING QUALITY:

Voice:

First person is acceptable
Let participants’ voices come through
Write accessibly

Balance:

Enough quotes to support claims
Not so many it’s just quotes
Your analysis is the contribution

Persuasion:

Evidence for each claim
Acknowledge limitations
Show rigor through transparency

When to Use

Exploring new phenomena without existing theory
Understanding participant experiences and perspectives
Investigating the “why” and “how” behind behaviors
Developing theory from data
Complementing quantitative findings with depth
Studying sensitive or complex topics
Generating hypotheses for future quantitative testing
Evaluating programs from stakeholder perspectives

Verification

Research question is appropriate for qualitative inquiry
Sampling is purposeful and justified
Data collection protocols are systematic
Coding is rigorous with clear definitions
Themes are well-defined and supported by data
Multiple trustworthiness strategies employed
Analysis is documented for audit trail
Findings balance description and interpretation

Input: $ARGUMENTS

Apply this procedure to the input provided.