Classical Qualitative Methods

RSM2074 Lecture Week 11

Dr. Chai Jun Ho

The Plan for Today

  • Why qualitative research matters
  • Critique of quantitative approaches
  • Three classical analysis methods
  • Deciding which method fits your research

Why Study Qualitative Methods?

The Subjective World

What does “love” mean to different people? Numbers can’t capture that.

Context Matters

The same behavior in different settings has different meanings.

Complexity

Reality is complex—sometimes you can’t reduce it to a scale of 1-9.

The Problem with Numbers

When Quantification Fails

The Likert Scale Problem:

"Rate your agreement: 1-9"
1 = Strongly Disagree    9 = Strongly Agree

What does a rating of 6 mean?

Interpretation A:

  • Genuine agreement
  • Moderate support
  • Leaning toward yes

Interpretation B:

  • Social desirability bias
  • Fear of negative judgment
  • Hedging one’s position

The same number = fundamentally different psychological states!

The Ambiguity Problem

Four Critiques of Quantitative Approaches

📚 Loss of Meaning

Numbers reduce rich human experience to single digits

🧩 Oversimplification

Complex reality ≠ simple measurements

⚠️ Confounding

We may be measuring the wrong construct entirely

❓ Interpretation Ambiguity

What does the number mean to the participant?

Qualitative vs Quantitative

Quantitative Research

📊 Form of data: Numbers

🔬 Method: Measurement

📈 Analysis: Statistical

🥶 Feel: Cold, bare, mechanical

Classical Qualitative Research

📝 Form of data: Words

👁️ Method: Observation

💡 Analysis: Illustrative

🔥 Feel: Warm, rich, fluid

Methods of Data Collection

Examples of Qualitative Data Sources

🎤 Interviews

👥 Focus Groups

📄 Documents

📹 Recordings

👀 Observations

🃏 Q-Sort

🎤 Interviews

Direct conversations with individual participants, exploring their experiences and perspectives in depth.

✓ Advantages

  • Deep personal insights

  • Flexible probing

  • Rich context

  • Real-time clarification

⚠️ Challenges

  • Time-intensive

  • Interviewer bias

  • Small sample sizes

  • Requires skill

🎤 Example: Employee Wellbeing

Q: “Describe a time you felt valued.”

“My manager asked for my input on the project. Being asked and listened to made me feel my contribution mattered—not the decision itself, but being heard.”

Insights: Emotional depth reveals what “valued” means beyond surface recognition.

👥 Focus Groups

Facilitated group conversations (typically 6-12 people) revealing shared perspectives and social dynamics.

✓ Advantages

  • Multiple viewpoints

  • Group dynamics visible

  • Cost-efficient

  • Natural discussion flow

⚠️ Challenges

  • Dominant voices

  • Conformity pressure

  • Less individual depth

  • Complex analysis

👥 Example: Remote Work Policy

Person A: “Love the flexibility - I’m way more creative at home, fewer interruptions”
Person B: “I actually miss the spontaneous coffee chats and overhearing conversations”
Person A: “But honestly, you can schedule those. I’m just more productive remotely without the commute”
Person C: “Yeah, but some of us live alone. The office is where I see people…”

Insight: Natural disagreement surfaces; group dynamics reveal genuine attitudes and social norms shaping work preferences.

📄 Written Documents

Analysis of letters, emails, journals, organizational documents—naturally occurring text.

✓ Advantages

  • Non-reactive data

  • Authentic language

  • Historical record

  • Private perspectives

⚠️ Challenges

  • Limited scope

  • Selection bias

  • No clarification

  • Different purpose

📄 Example: Internal HR Communications Analysis

Scenario: Analyzing organizational emails over 2 years

Key Extract from Internal Email Exchanges:

“I’ve noticed declining participation in team meetings…People seem hesitant to speak up…Decision-making has become increasingly top-down…something shifted…”

Key Insight: Authentic language and historical patterns visible in document sequence.

📹 Recorded Material

Analysis of recorded content such as interviews, documentaries, podcasts, team meetings, or broadcasts.

✓ Advantages

  • Rich multimodal data

  • Reanalyzable

  • Preserves nuance

  • Authentic setting

⚠️ Challenges

  • Other purposes

  • Selection bias

  • Interpretation dependent

  • Time-intensive

📹 Example: Team Meeting Analysis

20-minute meeting observations:

  • Lead speaks 60% | Juniors <20 sec each

  • Nonverbals: juniors avoid eye contact

  • Language: Lead uses declaratives | Juniors hedge (“I think, but…”)

  • Interruptions: Lead 3x | Juniors 0x

Key Insight: Multimodal data captures power dynamics and hierarchy invisible in meeting minutes alone.

👀 Direct Observations

Systematic recording of behavior in natural settings, observing what people actually do.

✓ Advantages

  • Direct behavior access

  • Ecological validity

  • Nonverbal + environment

  • Spontaneous patterns

⚠️ Challenges

  • Observer effect

  • Time-intensive

  • Selective attention

  • Privacy concerns

👀 Example: Open office environment study

Field Note (Timestamp: 10:30 AM):

“Engineer A mentions debugging issue aloud at desk. Engineer B overhears from 3 desks away, rolls chair over, engages in conversation. Problem solved collaboratively in 5 minutes. Unscheduled interaction—emerged from spatial proximity and accidental overhearing.”

Insight: Physical context triggers spontaneous knowledge-sharing; natural behavior.

🃏 Q-Sort: Systematic Ranking Method

Participants rank statements (-4 to +4) revealing perspective hierarchy without forced choices.

✓ Advantages

  • Systematic ranking

  • Comparable data

  • Holistic view

  • Quantifiable

⚠️ Challenges

  • Time-intensive

  • Pre-written statements

  • Training needed

  • Complex analysis

🃏 Example: Management Values Ranking

+4
    +2
      0
        -2
          -4
            • Listens to my ideas
            • Acts on my ideas
            • Provides development opportunities
            • Gives constructive feedback
            • Makes quick decisions
            • Makes decisions top-down
            • Communicates clearly
            • Celebrates successes publicly

            Insight: Values hierarchy captured systematically; comparable across participants; holistic view of what matters to person.

            Methods of Data Analysis

            Three Classical Qualitative Methods

            🔷 Repertory Grid Analysis - Captures how individuals understand their world - Personal constructs and mental frameworks

            📊 Content Analysis - Systematically codes text data - Quantifies patterns across sources - Counts theme frequency

            🎯 Thematic Analysis - Finds meaningful patterns in data - Keeps data in qualitative form - Extracts core themes

            Repertory Grid Analysis

            Understanding Personal Constructs

            Core Idea: Everyone uniquely understands the world through personal constructs—mental dimensions used to compare things.

            Example: Hospital Psychologists

            • Health-care oriented ↔︎ Finance-focused — Patient wellbeing priority
            • Collaborative ↔︎ Aloof — Relationship style
            • Influential ↔︎ Powerless — Decision-making authority

            Different psychologists rate the same person differently!

            The Repertory Grid Method

            🎯 Step 1: Construct Grid List elements to evaluate (people, objects, concepts)

            🔄 Step 2-3: Generate Constructs Compare three elements; identify similarity-contrast pairs

            📐 Step 4: Organize Arrange constructs by similarity patterns

            ✂️ Step 5: Simplify Remove redundant constructs

            Interactive Repertory Grid Example

            Content Analysis

            Transforming Words into Patterns

            Core Idea: Convert qualitative data into countable categories to identify patterns across sources.

            Example Question: Do hospital managers and industrial managers differ in emphasis?

            Content Analysis in Action

            The Content Analysis Process

            📍 Step 1: Specify Sampling Domain

            Define WHO, WHAT, WHEN of your data

            🏷️ Step 2: Develop Coding System

            Define categories: Train coders; Create units

            📝 Step 3: Data Coding

            Code all data: Verify agreement; Be consistent

            📊 Step 4: Data Analysis

            Quantify: Analyze; Present results

            Connection to Example: Instead of just counting keywords manually, real content analysis involves systematically coding all documents, interpreting meaning in context, and then comparing patterns across sources (e.g., hospital vs. industry documents).

            Example: Management Styles Change

            Finding: Hospitals emphasize hierarchy; industry emphasizes participation

            Thematic Analysis

            Beyond Counting: Finding Meaning

            Content Analysis: Counts how often codes appear

            Thematic Analysis: Identifies big ideas and patterns while keeping data in words

            Goal: Extract core meaningful themes from data

            The Thematic Analysis Process

            Phase 1: Data Engagement
            📖 Familiarization | 🏷️ Code Generation

            Phase 2: Pattern Recognition
            🔍 Search Themes | ✅ Review Themes

            Phase 3: Reporting
            📋 Define Themes | 📄 Produce Report

            Step 1: Data Familiarization

            📖 Read data thoroughly and repeatedly

            🧠 Immerse yourself in the material

            💭 Note patterns, ideas, and observations

            🔎 Get to know all the nuances

            Output: Deep familiarity with the full dataset

            Example: Data Familiarization in Action

            Step 2: Code Generation

            🏷️ Identify features that stand out

            ↔︎️ Mark similarities and differences

            📌 Create initial codes systematically

            📝 Keep notes on code meanings

            Output: A comprehensive list of initial codes

            Example: Code Generation in Action

            Step 3: Search for Themes

            🔗 Group related codes together

            🎯 Look for larger patterns

            🌳 Identify sub-themes

            ✓ Check conceptual sense

            Output: Candidate themes with constituent codes

            Example: Search for Themes in Action

            Step 4: Review Themes

            ✅ Check internal coherence

            🔍 Ensure no overlap

            📊 Verify data support

            🎯 Refine boundaries

            Output: Refined, distinct, well-defined themes

            Example: Review Themes in Action

            Step 5: Theme Definition

            📋 Name each theme clearly

            📝 Write detailed description

            🎯 Define scope & boundaries

            🏷️ Prepare summary statement

            Output: Theme definitions ready for reporting

            Example: Theme Definition in Action

            Step 6: Report Production

            ✍️ Write findings clearly

            💬 Include direct quotes

            🔗 Connect to research question

            📊 Discuss significance

            Output: Final research report with evidence

            Example: Report Production in Action

            Choosing Your Method

            Decision Framework

            Method Comparison

            Method Purpose Sample Data
            Repertory Grid Individual frameworks 1-20 Ratings
            Content Analysis Pattern counts 50+ Text
            Thematic Analysis Meaningful patterns 5-40 Interviews

            Strengths & Limitations

            Strengths of Qualitative Methods

            🔍 Rich understanding of complex phenomena

            👤 Captures variation that numbers hide

            🌍 Ecological validity in real-world contexts

            💡 Discovers unexpected insights not predicted by theory

            📝 Preserves language and context of data

            Limitations of Qualitative Methods

            ⏰ Time-intensive collection and analysis

            👁️ Researcher bias in coding and interpretation

            📊 Hard to generalize beyond studied population

            👥 Small samples limit statistical comparison

            🔬 Subjective interpretation requires rigor

            Addressing Key Criticisms

            “Lack of Rigor”

            Qualitative methods have systematic protocols: coding schemes, inter-rater reliability checks, audit trails, transparent methods

            “Not Generalizable”

            Goal is deep understanding, not prediction. Quality of insight ≠ quantity of participants

            When to Use Qualitative Methods

            Five Key Indicators

            📍 Your goal is understanding meaning not prediction

            🔍 You need deep exploration of complex phenomena

            🗣️ Participants’ perspectives are central to research

            👥 Your sample is small or hard to access

            🧩 You want to discover unexpected patterns in data

            The Five Steps for Qualitative Research (Yin, 1994)

            Step 1: Develop Research Questions

            🎯 Your foundation

            • Open-ended and exploratory
            • Ask “How” or “Why” not “Does”
            • Focused on understanding not prediction
            • Grounded in existing research

            Example: “How do employees experience psychological safety?”

            Step 2: Identify Key Propositions

            💡 Your theoretical framework

            • State what you expect based on theory
            • Guide your data collection
            • Help you interpret findings
            • Connect to existing literature

            Example: “Psychological safety predicts team performance (Edmondson, 1999)”

            Step 3: Specify Units and Contexts

            🔍 Define your scope

            • Units: What are you studying? (individuals? groups? organizations?)
            • Contexts: Where? (school? hospital? workplace?)
            • Justification: Why these choices?

            Example: “Clinical psychologists in hospital settings to understand workplace wellbeing”

            Step 5: Explain Interpretation Criteria

            ✅ Define your standards

            • What counts as evidence?
            • How much evidence is enough?
            • What would contradict your findings?
            • Make standards clear upfront

            Example: “Theme present if 3+ participants mention it across 2+ interviews”

            Thank You

            Remember: There’s no single “best” method—only the best method for your research question.

            Questions?