Qualitative Research Design

Learning Objectives

By the end of this session, you will be able to:

• Select an appropriate qualitative research approach — phenomenology, ethnography, grounded theory, case study, or narrative inquiry — based on the nature of your research questions and the type of knowledge you seek to produce
• Design a purposive sampling strategy with explicit inclusion/exclusion criteria and a defensible plan for determining when data saturation has been reached
• Construct a semi-structured interview or focus group protocol that moves from broad opening questions to focused probes, avoiding leading questions while ensuring all RQs are addressed
• Apply Lincoln and Guba's (1985) trustworthiness criteria — credibility, transferability, dependability, and confirmability — to design strategies that ensure rigour in qualitative research
• Articulate the rationale for choosing a qualitative (rather than quantitative) design, demonstrating understanding of when qualitative methods are the appropriate tool for the research question

Session Planner

Suggested breakdown of the 4-hour contact session.

Time	Segment	Activity	Mode
0:00–0:08	Opening	Recap Week 9; transition: "Last week we covered designs for measuring and testing. This week: designs for understanding and interpreting." The quantitative-qualitative complementarity.	Whole class
0:08–0:30	Lecture 1	Five qualitative traditions: phenomenology, ethnography, grounded theory, case study, narrative inquiry — what each reveals, when to use each, key exemplars in BBA and BCA	Lecture
0:30–0:50	Lecture 2	Qualitative sampling — purposive strategies (maximum variation, homogeneous, critical case, snowball, theoretical); saturation: what it is, how to plan for it, how to demonstrate it	Lecture
0:50–1:10	Activity	Tradition matching: given 6 research scenarios, select the most appropriate qualitative tradition and justify the choice. Debate cases where multiple traditions could apply.	Pairs
1:10–1:25	Discussion	Share and debate tradition choices; facilitator highlights the reasoning behind each match and addresses common mismatches	Whole class
1:25–1:40	Break	—	—
1:40–2:05	Lecture 3	Data collection — interviews (structured, semi-structured, unstructured), focus groups, observation, document analysis; interview protocol development; trustworthiness criteria (Lincoln & Guba)	Lecture + Demo
2:05–2:20	Demo	Live demonstration: a 3-minute "bad interview" (closed questions, leading probes, interrupting) vs. a 3-minute "good interview" (open questions, active listening, follow-up probes). Class identifies the differences.	Whole class
2:20–3:50	Lab Work	Part A: Select and justify qualitative approach; Part B: Develop interview/focus group protocol; Part C: Design purposive sampling strategy with saturation plan	Individual/Pairs
3:50–4:00	Exit Ticket	Submit qualitative approach justification, draft protocol, and sampling plan	Individual

1. The Qualitative Research Landscape

Qualitative research is the systematic inquiry into meaning — how people make sense of their world, experience phenomena, construct identities, and navigate social contexts. If quantitative research asks "how much" and "what is the relationship," qualitative research asks "what is the experience of," "how do people understand," and "why do they act as they do." The goal is not measurement but interpretation — producing rich, contextualised accounts that illuminate the human dimensions of a phenomenon.

1.1 When to Choose Qualitative — The Decision Criteria

Choose Qualitative When...	Example RQs (BBA)	Example RQs (BCA)
You need to understand a phenomenon from the perspective of those who experience it — their meanings, interpretations, and lived experience	"How do first-generation women entrepreneurs in Tier-2 Indian cities experience and navigate institutional barriers to business growth?"	"How do visually impaired users experience and navigate mobile banking applications — what are their coping strategies, frustrations, and moments of empowerment?"
The phenomenon is complex, context-dependent, and poorly captured by pre-defined variables or scales	"How do middle managers in Indian family businesses balance professional management practices with family expectations and informal power structures?"	"How do software teams transitioning from waterfall to agile actually adapt their collaboration practices — what informal norms emerge, and how do they differ from prescribed Scrum ceremonies?"
You want to understand a process — how something unfolds over time, what sequences and turning points characterise it	"How does the consumer trust-building process unfold in online second-hand luxury markets in India — what are the critical incidents that build or erode trust?"	"What is the process by which open-source contributors move from peripheral participation (filing bugs) to core membership (committing code) — what learning, socialisation, and legitimation occurs at each stage?"
The existing literature offers inadequate constructs or theories for the context you are studying, and you need to build theory from the ground up	"What constitutes 'career success' for gig economy workers in India — how do their definitions differ from the corporate career models assumed by existing career theories?"	"How do Indian AI startups navigate the tension between global AI ethics frameworks (developed in Western contexts) and local data practices, infrastructure constraints, and cultural norms?"
You need to explore a sensitive, hidden, or marginalised phenomenon where trust and rapport are essential and standardised instruments would be inappropriate	"How do employees experience and make sense of workplace bullying in Indian IT companies — what prevents reporting, and how do targets cope?"	"How do women developers in Indian tech startups experience and navigate gender-based discrimination in male-dominated engineering teams?"

2. Five Qualitative Traditions — Choosing the Right Lens

Creswell and Poth (2018) identify five major qualitative approaches, each with its own philosophical roots, disciplinary origins, central question, and analytical procedures. Your choice of tradition shapes everything: what you look for, how you collect data, how you analyse it, and what kind of knowledge you produce. Selecting the wrong tradition for your RQs produces research that is internally inconsistent — like using a microscope when you need a telescope.

2.1 The Five Traditions at a Glance

Tradition	Central Question	Disciplinary Origin	Key Procedures	Typical Output
Phenomenology	"What is the lived experience of [phenomenon] for [people]?" — the essence and structure of a conscious experience	Philosophy (Husserl, Heidegger, Merleau-Ponty)	In-depth interviews with 5–25 people who have experienced the phenomenon; bracketing (setting aside researcher preconceptions); horizontalisation (treating every statement as equally significant); clustering into themes; composite description of the essence	A rich, textured description of what the experience IS like and HOW it is experienced — its invariant structure or essence
Ethnography	"What is the culture of [group] — their shared beliefs, practices, language, and behaviours?"	Anthropology (Malinowski, Geertz)	Prolonged immersion in the field (months); participant observation; in-depth interviews; collection of artefacts and documents; thick description (Geertz, 1973); analysis of cultural themes	A portrait of a culture-sharing group — how they make meaning, what they value, how they organise social life
Grounded Theory	"What theory emerges from the data to explain [process/action/interaction] for [people] in [context]?"	Sociology (Glaser & Strauss, 1967; Strauss & Corbin; Charmaz)	Theoretical sampling (sampling guided by emerging theory); constant comparison (comparing data with data, data with codes, codes with categories); open, axial, and selective coding; memo-writing throughout; theoretical saturation	A substantive theory — a set of interrelated concepts and propositions that explain the process or phenomenon, grounded in the data
Case Study	"What is the case [person, organisation, event, programme] — what are its boundaries, its dynamics, and what can we learn from it?"	Multiple (Stake, Yin, Merriam)	Bounding the case (defining what is inside and outside); multiple data sources (interviews, documents, observation, artefacts); within-case and cross-case analysis; pattern matching; explanation building	An in-depth, multi-faceted understanding of a bounded system — the case in its real-world context
Narrative Inquiry	"What stories do [people] tell about [experience] — and what do these stories reveal about identity, meaning, and social context?"	Literature, Sociology, Education (Clandinin & Connelly, Riessman)	Collecting stories through interviews, journals, letters, or autobiographies; restorying (organising stories into a coherent chronology — the three-dimensional narrative space: temporality, sociality, place); analysing for narrative elements (plot, character, setting, turning points, resolution)	A narrative account — stories restoried into a coherent framework that reveals how individuals make sense of their lives through narrative

2.2 Matching Tradition to Research Question — Worked Examples

Research Question	Best-Fit Tradition	Why This Tradition?
"How do Indian women in C-suite positions experience and navigate gender bias in boardroom decision-making?"	Phenomenology	The RQ asks about a lived experience — the felt, subjective reality of navigating bias. We want the essence of that experience across multiple women, not a theory of how bias operates or a cultural portrait of boardrooms.
"How do delivery partners for food-tech platforms in Mumbai develop and sustain informal mutual-support networks to cope with algorithmic management?"	Ethnography	The RQ asks about a culture-sharing group (delivery partners) with shared practices, language, and norms. Understanding their world requires immersion — observing them at gathering points, riding along, understanding their argot and rituals.
"How do Indian fintech startups navigate the process of building consumer trust — what stages, strategies, and turning points characterise this process?"	Grounded Theory	The RQ asks about a process — trust-building unfolds over time. We need a theory of that process grounded in data from startup founders, not an existing theory tested deductively. The output will be a process model of trust-building in this context.
"How did Zomato's 'Gold' membership programme reshape its competitive positioning, customer engagement metrics, and unit economics between 2018 and 2024?"	Case Study	The RQ examines a bounded system (Zomato Gold programme, 2018–2024) from multiple angles. The case is intrinsically interesting — it is not a "sample" of loyalty programmes; it is the specific case we want to understand in depth.
"How do Indian software engineers who left corporate jobs to become indie developers narrate their career transition — what stories do they tell about risk, identity, and success?"	Narrative Inquiry	The RQ asks about stories — how people construct coherent narratives about radical career change. The unit of analysis is the story itself — its plot, characters, turning points, and the identity work the narrative performs for the teller.

3. Qualitative Sampling & Saturation

Qualitative sampling operates on a fundamentally different logic from quantitative sampling. The goal is not statistical representativeness — it is information richness. You select participants not because they "represent" a population in a statistical sense, but because they can provide deep, relevant, and varied information about your phenomenon of interest. The sample size is not determined by power analysis but by saturation — the point at which additional data ceases to yield new insights.

3.1 Purposive Sampling Strategies

Strategy	What You Do	Use When...	Example
Maximum Variation	Deliberately select participants who differ widely on characteristics relevant to your RQs — age, experience, organisation size, geography, etc.	You want to capture the full range of variation in a phenomenon; you want findings that reflect diverse perspectives; your RQs concern shared patterns ACROSS difference	Studying digital payment adoption: select participants ranging from daily users to non-adopters, across urban and rural locations, across income levels, across age groups. The shared themes that emerge across this diversity are particularly robust.
Homogeneous	Select participants who are similar on key characteristics to enable deep, focused exploration of a specific subgroup	You want to understand a particular subgroup in depth; variation would dilute the focus; the phenomenon is specific to a defined population	Studying the experience of first-time mothers returning to IT jobs after maternity leave in Bangalore — all participants share key characteristics (first child, IT sector, Bangalore, returned within last 2 years), enabling deep exploration of this specific experience.
Critical Case	Select a case or participant that is strategically important — if a pattern holds here, it is likely to hold elsewhere; if it doesn't hold here, it is unlikely to hold anywhere	You want to test a proposition or theory in a setting where it faces its strongest test; resources are limited and you need maximum analytical leverage from a small sample	Testing whether a new agile methodology improves team productivity: select the team with the most experienced developers and strongest technical practices. If agile doesn't improve productivity for THEM, it is unlikely to improve it for less experienced teams.
Theoretical Sampling	Let emerging theory guide who you sample next — as categories and propositions develop, seek participants who can elaborate, refine, or challenge them. This is specific to Grounded Theory.	You are building theory from data (Grounded Theory); your sampling evolves as your analysis progresses; you cannot pre-specify the entire sample before data collection begins	Initial interviews with startup founders suggest "investor pressure" as a key category. You then sample founders who: (a) experienced extreme investor pressure (to elaborate the category), (b) had supportive investors (to understand negative cases), and (c) bootstrapped without investors (to explore boundary conditions).
Snowball / Chain	Start with a few participants who meet criteria; ask them to refer others who also meet criteria; continue until target sample is reached	The population is hidden, stigmatised, or difficult to access through formal channels; trust is essential and referrals from insiders build credibility	Studying whistleblowers in Indian corporations — no sampling frame exists, and trust is critical. Initial participants identified through media reports and professional networks; each is asked to refer others they know who have also blown the whistle.

3.2 Data Saturation — How Much is Enough?

Saturation is the point at which additional data collection produces no substantially new insights, themes, or theoretical categories. It is the qualitative equivalent of statistical power — the criterion by which you determine that your sample is adequate. Unlike power analysis, saturation cannot be calculated in advance. But it can — and must — be planned for and demonstrated.

3.3 Demonstrating Saturation in Your Methodology Chapter

"Data was collected until saturation was reached" is insufficient. You must demonstrate HOW you determined saturation — what evidence supports your claim that additional data would not yield new insights. Include in your methodology chapter:

1. Your planned sample range: "I plan to conduct 12–20 interviews, with the final number determined by thematic saturation."
2. Your saturation tracking procedure: "After each interview, I will maintain a saturation log: new codes generated, existing codes elaborated, new themes identified. When three consecutive interviews produce no new codes or themes, I will conduct two additional interviews to confirm saturation."
3. Your saturation evidence: In your findings chapter, include a saturation table or figure: number of new codes generated per interview or focus group. Show the curve flattening. This is the evidence that supports your saturation claim.

4. Qualitative Data Collection — Generating Rich Data

Qualitative data collection is more accurately described as data generation — unlike picking up pre-existing data points, you are actively co-creating data through interaction with participants. The quality of what you generate depends on your skill as an interviewer, observer, or facilitator, not just on your research design.

4.1 The Semi-Structured Interview — The Workhorse Method

The semi-structured interview is the most widely used qualitative method in capstone research. It occupies the productive middle ground between the structured interview (fixed questions, fixed order — essentially a verbally administered survey) and the unstructured interview (a conversation with no predetermined questions). A semi-structured interview uses an interview protocol — a guide with key questions and probes — but allows the conversation to follow participant leads, explore unexpected paths, and adapt to each participant's unique perspective.

4.2 The Art of the Probe

Probes are follow-up questions or prompts that deepen, clarify, or expand a participant's response. They transform a surface-level answer into rich data. Mastering probes is the single most important interviewing skill.

Probe Type	Purpose	Example
Silent Probe	Signal that you are listening and waiting for more. The most underused and powerful probe.	Remain silent after the participant finishes speaking. Nod. Maintain eye contact. Count to five in your head before asking the next question.
Echo Probe	Repeat the last few words the participant said, with a rising intonation — inviting elaboration	Participant: "...and then I realised the whole system was designed without ever talking to actual users." Interviewer: "Without ever talking to actual users...?"
Detail Probe	Ask for specifics — when, where, who, what exactly happened, what was said	"Can you walk me through that meeting — who was there, what was said, what did you do?"
Example Probe	Ask for a concrete instance of an abstract claim	"You said the culture was 'toxic' — can you give me a specific example of something that happened that felt toxic?"
Elaboration Probe	Ask the participant to expand on a brief or superficial response	"Could you tell me more about that?" "What did that mean to you?" "How did that affect how you thought about...?"
Clarification Probe	Check your understanding; verify you interpreted correctly	"Let me make sure I understand — are you saying that [paraphrase], or did I misunderstand?"
Emotion Probe	Explore the affective dimension of an experience	"How did you feel when that happened?" "What was the emotional impact of that decision on you and your team?"
Contrast Probe	Explore variation by asking about comparison or difference	"How was that experience different from what you expected?" "You mentioned two very different reactions from your managers — what do you think explains the difference?"

4.3 Other Qualitative Data Collection Methods

Method	What It Is	Best For	Practical Considerations
Focus Groups	A facilitated group discussion with 6–10 participants, lasting 60–90 minutes, exploring a shared experience or topic through group interaction	When group interaction will generate insights that individual interviews won't — shared meaning-making, disagreement and debate, collective sense-making; when the topic is not highly sensitive or stigmatised	Harder to schedule than interviews (6–10 people simultaneously); dominant participants can skew the discussion; requires skilled facilitation; generates a lot of data quickly; recordings are harder to transcribe (multiple speakers, overlapping talk)
Observation	Systematically watching and recording behaviour, interactions, and context in a natural setting; can be participant (you are part of the setting) or non-participant (you are an outsider observing)	When you need to understand what people DO, not just what they SAY they do; when behaviour is routine or tacit and participants may not be able to articulate it; central to ethnography	Requires access to the setting (gatekeepers); can be time-intensive; observer effect (people change behaviour when observed); field notes must be written immediately after observation; ethical considerations if observation is covert
Document Analysis	Systematic review and analysis of existing texts — organisational documents, policies, reports, emails, meeting minutes, social media posts, websites, news articles	When documents can provide context, corroborate or challenge interview data, or reveal organisational discourses and practices that participants may not discuss; useful for triangulation	Documents were produced for a purpose — not for your research; they reflect organisational interests and authorship; access may be restricted; authenticity and credibility must be evaluated; large volumes require systematic sampling
Diaries / Journals	Participants record their experiences, thoughts, or behaviours over a period of time, following a structured or unstructured format provided by the researcher	When you need to capture experiences as they unfold (not retrospectively); when the phenomenon is intermittent or private; when retrospective accounts would be distorted by memory	High participant burden → high dropout; requires participant training and motivation; data quality varies enormously; produces longitudinal data that requires longitudinal analysis

5. Trustworthiness — Rigour in Qualitative Research

Quantitative research establishes quality through validity, reliability, and objectivity. Qualitative research — operating on different philosophical assumptions — requires different criteria. Lincoln and Guba's (1985) trustworthiness framework is the most widely cited and provides four parallel criteria: credibility, transferability, dependability, and confirmability.

Quantitative Criterion	Qualitative Parallel (Lincoln & Guba)	Definition	Strategies to Establish
Internal Validity	Credibility	The findings accurately represent the participants' realities — they are credible to the people who experienced the phenomenon. "Do the findings ring true?"	Prolonged engagement: Spending enough time in the field to understand the context and build trust. Triangulation: Using multiple data sources, methods, or investigators to cross-check findings. Member checking: Returning findings to participants and asking: "Does this represent your experience accurately?" Peer debriefing: Discussing findings with a disinterested peer who challenges assumptions and interpretations.
External Validity	Transferability	The findings can be transferred to other contexts — not by the researcher (who cannot know all possible contexts), but by the reader, who assesses whether the findings apply to their context based on the rich description provided.	Thick description: Describing the context, participants, and phenomenon in sufficient detail that a reader can judge the similarity between the research context and their own context. The researcher's job is to provide the raw material for transferability judgements; the reader's job is to make those judgements.
Reliability	Dependability	The research process is logical, traceable, and clearly documented — another researcher could follow the "audit trail" and understand how conclusions were reached (not necessarily replicate them, since qualitative research is context-dependent).	Audit trail: Maintaining a complete record of raw data, analysis products (codes, categories, themes), process notes (methodological decisions and rationale), and personal reflections (reflexive journal). Inquiry audit: Having an external reviewer examine the audit trail and confirm that findings are grounded in the data.
Objectivity	Confirmability	The findings are shaped by the participants and the data, not by the researcher's biases, motivations, or preconceptions. The researcher has practised reflexivity — examining their own influence on the research.	Reflexive journal: Keeping a separate journal documenting your assumptions, expectations, emotional reactions, and how your background may influence your interpretations. Audit trail: The same audit trail that supports dependability also supports confirmability — it allows an external reviewer to trace findings back to data, checking that interpretations are grounded.

5.1 Reflexivity — The Researcher as Instrument

In qualitative research, YOU are the primary research instrument. Your background, beliefs, experiences, and identity shape what you notice, how you interpret it, and what you report. Reflexivity is the practice of critically examining your own role in the research — not eliminating your influence (impossible), but making it visible so that readers can assess it.

Think Deeper — Cross Questions

Discuss in pairs before sharing with the class.

Quick Check — Tradition Matching

For each research question, select the most appropriate qualitative tradition.

Knowledge Check — Interactive Quiz

Test your understanding of qualitative research design.

Lab Activity — Developing Qualitative Instruments

Part A: Select and Justify Your Qualitative Approach (30 min)

1. Revisit your RQs. Do they ask about lived experience, culture, process, a bounded system, or stories? This determines your tradition.
2. Select the tradition that best matches your RQs using the decision framework in Section 2. If multiple traditions could apply, note the trade-offs and justify your choice.
3. Write a 200–300 word justification for your methodology chapter, using the template below.

Part B: Develop Your Interview or Focus Group Protocol (60 min)

1. Map your RQs to protocol sections: Each RQ should correspond to a block of questions in your protocol. If an RQ has no corresponding protocol questions, it will not be answered by your data.
2. Draft your protocol following the four-phase structure (Section 4.1). Write the opening script verbatim. Write 8–12 core exploration questions, each with 2–3 planned probes. Write 3–4 closing questions, including the reflective summary.
3. Review each question against the probe types (Section 4.2). Are you asking mostly "what" and "how" questions (good) or "why" questions (can feel interrogating)? Does your protocol include mostly open questions with planned probes?
4. Pilot one question with a peer — does the question generate rich narrative or a short, thin answer? Revise questions that produce thin responses.

Part C: Design Your Sampling and Saturation Plan (30 min)

1. Define your inclusion/exclusion criteria precisely. What must a participant have experienced, known, or been to qualify for your study?
2. Select your purposive sampling strategy from Section 3.1. Justify with reference to your RQs and tradition.
3. Specify your planned sample range based on tradition norms, information power, and saturation literature (Section 3.2).
4. Describe your saturation procedure using the template in Section 3.3.
5. Write a 200–300 word sampling section for your methodology chapter.

Exit Ticket

Submit with your qualitative approach justification and protocol.

Key Takeaways — Week 10