The Complete Guide to Qualitative Research Methodology

Qualitative research is not simply research without numbers. It is a family of approaches designed to understand meaning, experience, process, context, identity, interaction and interpretation. A defensible qualitative study begins with a question that requires depth rather than measurement, and it proceeds through a coherent chain of philosophical assumptions, research design, data generation, analysis and interpretation.

1. What qualitative research is 2. Why researchers use it 3. Philosophical foundations 4. Ontology and epistemology 5. Induction, deduction and abduction 6. Core characteristics 7. Strengths and limitations 8. When to use it 9. When not to use it 10. Examiner expectations

1. What Is Qualitative Research?

Qualitative research investigates how people understand, experience and respond to the world around them. It is especially useful when the researcher needs to explore how a phenomenon unfolds, how individuals assign meaning to it, how context shapes behaviour, or why apparently similar people respond differently.

The data may include interview transcripts, focus-group discussions, observations, field notes, diaries, organisational documents, policy texts, images, online interactions or audio-visual material. What makes the study qualitative is not the format of the data alone. The defining feature is the analytical purpose: the researcher interprets patterns of meaning rather than merely counting occurrences.

Key principle: A qualitative study should be chosen because the research question demands contextual understanding, not because the researcher wishes to avoid statistics.

For example, a quantitative study might measure whether employee engagement declined after digital transformation. A qualitative study might explore how employees experienced that transformation, how they interpreted management decisions, and why the same change created enthusiasm in one department but resistance in another.

2. Why Qualitative Research Matters

Many important research problems cannot be understood adequately through variables alone. Numbers may reveal that a pattern exists, but they often cannot explain the processes, interpretations and institutional conditions that produced it. Qualitative research gives researchers access to those layers.

Meaning: How participants understand an event, identity or relationship.
Process: How decisions, adaptations or social changes develop over time.
Context: How organisational, cultural, political or historical conditions shape experience.
Complexity: How several influences interact rather than operate independently.
Voice: How underrepresented participants describe issues in their own terms.
Theory development: How new concepts or explanations can emerge from close engagement with data.

A strong qualitative study often makes the familiar unfamiliar. It exposes assumptions that surveys may take for granted and identifies mechanisms that a statistical association cannot reveal.

3. Philosophical Foundations

Research philosophy is not decorative language placed at the beginning of a methodology chapter. It explains what the researcher believes can be known, how knowledge can be produced, and what role interpretation plays in that process. The philosophy should be visible in the design choices.

3.1 Positivism

Positivism assumes that reality exists independently of the researcher and can be observed through systematic measurement. It is more commonly associated with quantitative designs, although qualitative data can sometimes be used within a largely positivist project—for example, structured coding of responses using predefined categories.

A purely positivist stance is usually a poor fit when the aim is to explore multiple interpretations, socially constructed meanings or context-dependent experiences.

3.2 Interpretivism

Interpretivism begins from the position that social reality is understood through the meanings people create and share. Researchers therefore seek to understand events from participants' perspectives rather than treating behaviour as an objective response detached from interpretation.

An interpretivist researcher studying remote work would not ask only whether productivity changed. They might examine what productivity meant to employees, how monitoring altered trust, and how home circumstances affected their experience of organisational expectations.

3.3 Constructivism

Constructivism emphasises that knowledge and social realities are produced through interaction, language, institutions and history. Categories such as leadership, professionalism, risk or success are not treated as fixed objects with universal meanings. Their meanings are constructed and negotiated.

Constructivist research therefore pays close attention to how participants describe reality and how those descriptions are shaped by their social positions and contexts.

3.4 Critical Realism

Critical realism distinguishes between what people experience, what actually happens, and the deeper mechanisms or structures that generate events. It accepts that a reality exists independently of individual interpretation, while also recognising that researchers can access it only imperfectly.

This approach is especially valuable when a researcher wants to explain why an outcome occurs under some conditions but not others. For example, a policy may formally apply to all employees, yet underlying power structures, resource inequalities and managerial practices may produce different outcomes across sites.

3.5 Pragmatism

Pragmatism prioritises the research problem and the practical consequences of knowledge. It permits researchers to combine assumptions and methods when doing so helps answer the question. Pragmatism is often associated with mixed methods, but it can also guide qualitative studies focused on practical organisational or policy problems.

Common mistake: Pragmatism does not mean that any method can be used without justification. The researcher must still explain why each method produces evidence relevant to the research problem.

Position	View of reality	Research emphasis	Typical qualitative fit
Positivism	Single reality that can be observed	Measurement, regularity, explanation	Limited or structured qualitative use
Interpretivism	Reality understood through meanings	Experience and interpretation	Interviews, case studies, ethnography
Constructivism	Reality socially produced	Meaning-making and interaction	Narrative, discourse, reflexive thematic analysis
Critical realism	Reality exists but is imperfectly known	Mechanisms, structures and context	Explanatory case studies, realist analysis
Pragmatism	Knowledge judged partly by usefulness	Problem solving and consequences	Applied qualitative and mixed-method research

4. Ontology and Epistemology Explained Simply

Ontology concerns what exists and what reality is like. Epistemology concerns how we can know about that reality. These concepts are connected but not interchangeable.

Suppose a researcher studies organisational culture. An ontological question is whether culture exists as a relatively stable feature of the organisation or whether it is continually produced through interactions. An epistemological question is how the researcher can know that culture—through surveys, interviews, observation, documents, or some combination.

Research problem

View of reality

View of knowledge

Methodology

Methods

Examiners look for consistency across this chain. A thesis may claim that reality is multiple and socially constructed, but then present interview findings as if they were objective facts that apply universally. That contradiction weakens the methodology.

4.1 Objectivist and Relativist Tendencies

An objectivist ontology treats social phenomena as having an existence that is not dependent on any one person. A relativist ontology holds that reality can differ across people, contexts and interpretive communities. Most qualitative studies fall somewhere between rigid objectivism and absolute relativism.

The important task is not to select fashionable terminology. It is to state a defensible position that fits the research question and to apply it consistently.

5. Inductive, Deductive and Abductive Reasoning

5.1 Inductive Reasoning

Induction moves from detailed observations toward broader patterns, concepts or explanations. The researcher does not begin with a hypothesis that must be tested. Instead, patterns are developed through engagement with the data.

Induction does not mean entering the study with an empty mind. Researchers always bring prior knowledge and assumptions. A credible inductive study demonstrates openness to unexpected patterns and avoids forcing data into an existing framework.

5.2 Deductive Reasoning

Deduction begins with theory or an existing conceptual framework and examines how the data correspond with it. Qualitative deduction may involve using predefined themes, sensitising concepts or propositions to guide coding and interpretation.

This can be appropriate when evaluating implementation against an established framework or examining how a known theory appears in a new context. The danger is that the framework may silence important ideas that fall outside it.

5.3 Abductive Reasoning

Abduction moves iteratively between data and theory. It is often triggered by a surprising observation that existing explanations cannot fully account for. The researcher develops the most plausible explanation, returns to the data, compares alternatives and refines the interpretation.

Abduction is highly useful in doctoral research because it acknowledges that analysis is rarely a straight line. Researchers often begin with existing concepts, encounter anomalies, and then modify or extend their explanations.

Reasoning	Starting point	Main movement	Primary risk
Inductive	Detailed data	Data to pattern or concept	Claiming to be theory-free
Deductive	Theory or framework	Theory to data	Forcing data into prior categories
Abductive	Surprising finding or incomplete explanation	Repeated movement between data and theory	Selecting an attractive explanation without testing alternatives

6. Core Characteristics of Strong Qualitative Research

Contextual depth: Findings are interpreted within the setting in which they were produced.
Participant meaning: The study takes participants' interpretations seriously without treating every account as unquestionable truth.
Researcher reflexivity: The researcher examines how their role, assumptions and relationships shaped the research.
Iterative design: Data collection and analysis may inform one another.
Purposeful sampling: Participants or cases are selected because they can illuminate the phenomenon.
Analytical transparency: Readers can follow how raw material became codes, categories, themes or explanations.
Rich evidence: Claims are supported with extracts, observations and contextual detail.
Interpretive contribution: The findings move beyond description to explain significance, pattern or mechanism.

7. Strengths and Limitations

7.1 Strengths

Qualitative research can reveal mechanisms hidden beneath surface patterns, preserve complexity, explore emerging topics, include marginalised perspectives and generate theory. It is particularly effective where concepts are contested or insufficiently defined.

7.2 Limitations

Qualitative findings are often context-dependent, and the analytical process requires considerable judgement. Data collection and analysis can be time-intensive. Researcher influence cannot be eliminated, although it can be examined reflexively. Claims of broad prevalence or population-level effect usually require additional quantitative evidence.

Examiner insight: A limitation is not automatically a weakness. It becomes a weakness when the researcher ignores it, makes claims the design cannot support, or fails to explain how quality was protected.

8. When Qualitative Research Is Appropriate

Qualitative research is usually appropriate when the question asks:

How do people experience a phenomenon?
How are meanings, identities or practices constructed?
Why does a process unfold differently across contexts?
How do participants interpret organisational or policy change?
What mechanisms may explain an observed outcome?
How does a little-understood phenomenon operate?
How can a new concept or theory be developed?

Words such as explore, understand, experience, interpret, construct, negotiate and how often indicate qualitative intent, but wording alone is not decisive. The evidence required by the question is what matters.

Worked Example

Weak question: “What is the impact of leadership on employee performance?” This implies a measurable causal or predictive relationship and may be better suited to quantitative analysis.

Qualitative alternative: “How do employees interpret leadership practices during organisational restructuring?” This asks about meaning, experience and context.

9. When Qualitative Research Should Not Be Used

Qualitative research is generally unsuitable as the sole design when the principal aim is to estimate prevalence, quantify an effect, compare group means, predict an outcome, test a causal hypothesis or generalise statistically to a defined population.

It should also not be chosen merely because:

the researcher has a small sample;
statistical methods feel difficult;
interviews appear easier than survey design;
the supervisor prefers qualitative research;
the researcher wants flexibility without documenting decisions.

Interviews can generate large, complex datasets. A small number of interviews does not automatically make a study manageable or rigorous.

10. What Examiners Look For

Examiners commonly evaluate whether:

The research question genuinely requires qualitative evidence.
The philosophical position is explained and aligned with the design.
The selected methodology is more than a label.
Sampling decisions are purposeful and justified.
Data collection is sufficiently detailed to be evaluated.
The analytical process is transparent and systematic.
The researcher demonstrates reflexivity.
Findings are supported by evidence rather than assertion.
Alternative explanations and negative cases are considered.
Claims remain within the limits of the design.

Frequent examiner criticism: “The chapter states that an interpretivist approach was adopted, but it does not show how interpretivism influenced sampling, interviewing, analysis or the presentation of findings.”

A Practical Alignment Checklist

Element	Question to ask
Problem	Does the problem require understanding of meaning, context or process?
Question	Can the question be answered through participants' accounts, observation or texts?
Philosophy	Does the philosophical position fit the kind of knowledge being claimed?
Methodology	Does the methodology provide a coherent logic for the study?
Sampling	Were information-rich cases selected for a clear reason?
Analysis	Can the reader see how interpretations were developed?
Claims	Are conclusions appropriately bounded by context and evidence?

11. Choosing the Right Qualitative Methodology

Selecting a qualitative methodology is not a matter of choosing the method that sounds most sophisticated. The methodology must match the nature of the research problem, the type of knowledge sought, the unit of analysis, the role of context and the intended contribution. A case study, phenomenological study and grounded theory study may all use interviews, but they are not interchangeable because they pursue different analytical goals.

Decision rule: Choose the methodology by asking what the study must explain—not by asking which data-collection method you prefer.

11.1 Case Study Research

Case study research investigates a bounded case in depth and within its real-life context. The case may be an organisation, programme, policy, community, event, team, project, decision process or other clearly delimited system. The purpose is not merely to collect several forms of data. The purpose is to understand the case as a coherent whole and to explain how its context shapes what happens within it.

A strong case study defines its boundaries explicitly. These boundaries may be geographical, organisational, temporal, functional or conceptual. A weak case study often labels an organisation as a “case” without explaining why it is analytically significant or where the case begins and ends.

When case study is appropriate

The research asks how or why a contemporary phenomenon occurs.
The phenomenon cannot be separated easily from its context.
The study focuses on a bounded organisation, programme, event or process.
Multiple sources of evidence can illuminate the case.
The researcher seeks analytical rather than statistical generalisation.

Single-case and multiple-case designs

A single-case design may be appropriate when the case is critical, unusual, revelatory, extreme or longitudinal. A multiple-case design compares several cases to identify recurring mechanisms, contextual differences or contrasting outcomes. Multiple-case research is not automatically stronger. It requires a clear replication logic and enough depth in each case.

Case study choice	Best used when	Main risk
Single case	One case is unusually revealing or theoretically significant	Treating an ordinary convenient site as intrinsically important
Embedded case	The case contains meaningful sub-units such as departments or teams	Losing sight of the overall case
Multiple cases	Comparison helps test or refine explanations	Producing shallow mini-cases
Longitudinal case	Change must be examined over time	Insufficient temporal evidence

Examiner criticism: “The thesis claims to use case study methodology, but the case is neither bounded nor analysed as a case. The organisation is merely the location from which participants were recruited.”

11.2 Phenomenology

Phenomenology explores how people experience a phenomenon and the meanings they attach to that experience. It is concerned with lived experience rather than with measuring frequency or identifying organisational outcomes. The researcher seeks to understand the essential qualities, structures or interpretations of an experience as described by participants.

Phenomenology is appropriate when the phenomenon itself is central—for example, living with chronic uncertainty, experiencing professional identity loss, becoming a first-generation doctoral student, or navigating bereavement after workplace trauma.

Descriptive and interpretive phenomenology

Descriptive phenomenology aims to describe the essence of an experience as faithfully as possible, often emphasising bracketing or disciplined reflection on prior assumptions. Interpretive or hermeneutic phenomenology accepts that interpretation is unavoidable and examines how experience is shaped by language, history and context.

Common requirements

Participants have direct experience of the phenomenon.
Interviews are sufficiently deep and reflective.
Analysis focuses on experience rather than topic summaries.
The researcher demonstrates reflexivity.
Claims remain grounded in participants' accounts.

Common mistake: Using “phenomenology” as a label for any interview study. A phenomenological study must analyse lived experience, not merely collect opinions.

11.3 Grounded Theory

Grounded theory is designed to generate an explanatory theory of a social process grounded in systematically collected and analysed data. It is especially appropriate when existing theories do not explain how participants move through a process, respond to conditions or manage a recurring problem.

The defining features are not simply open coding and interviews. Grounded theory normally involves concurrent data collection and analysis, constant comparison, memo writing, theoretical sampling and progressive development of categories until an explanatory model emerges.

Core grounded theory processes

Initial coding: Examining data closely and identifying actions, meanings or incidents.
Constant comparison: Comparing incident with incident, code with code, and category with category.
Memo writing: Recording analytical ideas, relationships and questions.
Theoretical sampling: Collecting additional data to refine emerging categories.
Category integration: Connecting major categories into an explanatory framework.
Theoretical saturation or sufficiency: Determining when additional data no longer materially develop the core categories.

Major traditions

Classic grounded theory, Straussian grounded theory and constructivist grounded theory differ in their philosophical assumptions and coding procedures. A researcher should state which tradition is being followed and avoid combining incompatible procedures without explanation.

Examiner criticism: “The study claims grounded theory but recruited the full sample before analysis, used no theoretical sampling and produced themes rather than a theory.”

11.4 Ethnography

Ethnography examines culture, shared practices, meanings and social interaction through prolonged engagement in a natural setting. Observation is usually central, although interviews, documents and artefacts may also be used. Ethnography seeks to understand not only what participants say but also what they routinely do, how they interact and which unwritten norms organise their world.

Traditional ethnography often involves extended fieldwork. Focused ethnography may examine a more specific issue over a shorter period, particularly in professional, healthcare or organisational settings. Digital ethnography investigates online communities, platforms and digitally mediated practices.

When ethnography is appropriate

The research focuses on a culture-sharing group.
Routine practice and informal norms matter.
Observation can reveal what interviews alone cannot.
The researcher can gain meaningful access to the field.
Context and interaction are central to interpretation.

Researcher position

Ethnographers may participate to different degrees, from relatively detached observation to active membership. The researcher must explain their role, access, relationships, ethical responsibilities and influence on the field.

Common mistake: Calling a study ethnographic because the researcher visited a site or observed a few meetings. Ethnography requires sustained attention to culture, practice and interaction.

11.5 Narrative Inquiry

Narrative inquiry studies how people make sense of experience through stories. It treats stories not merely as containers of facts but as structured acts of meaning-making shaped by audience, identity, time and cultural expectations.

The analysis may focus on what is told, how the story is organised, why it is told in a particular way, how identities are constructed, and how broader social narratives influence personal accounts.

Possible units of narrative analysis

A life history
A professional career story
A story of illness, migration or transformation
Several accounts of the same event
Institutional or policy narratives
Stories circulated within an organisation

Narrative inquiry is especially useful when temporality matters. It can show how participants connect past events, present identity and imagined futures.

Examiner criticism: “The researcher collected stories but fragmented them into decontextualised themes, thereby losing the narrative structure that justified the methodology.”

11.6 Action Research

Action research combines inquiry with deliberate efforts to improve practice. Researchers and participants identify a problem, plan an intervention, act, observe the consequences and reflect before beginning another cycle. The goal is both practical improvement and knowledge generation.

Diagnose

Plan

Act

Observe

Reflect

Action research is appropriate in professional settings where the researcher has a legitimate role in change—for example, education, healthcare, public administration or organisational development. It requires transparency about dual roles, power relationships and how claims are generated from local cycles of action.

Key quality questions

Was the problem defined collaboratively?
Were participants involved meaningfully rather than symbolically?
Did reflection alter later action?
Is the intervention documented clearly?
Does the study produce knowledge beyond a simple project report?

11.7 Participatory Action Research

Participatory action research extends action research by placing stronger emphasis on shared ownership, power redistribution and knowledge produced with communities rather than about them. Participants may act as co-researchers in defining questions, collecting data, interpreting findings and deciding action.

This methodology is particularly appropriate where research aims to support communities experiencing exclusion, inequality or institutional disadvantage. However, participatory language should not be used when participants have little influence over the study.

Ethical warning: Participation creates responsibilities. Researchers should be honest about which decisions participants genuinely control and which remain constrained by funding, ethics or institutional requirements.

11.8 Qualitative Description

Qualitative description provides a rich, relatively low-inference account of participants' experiences, perceptions or practices. It is appropriate when the research aims to produce a clear descriptive summary rather than develop theory, uncover the essence of lived experience or interpret deep cultural structures.

It is often valuable in applied health, service evaluation and policy research where decision-makers need an accessible account of what participants experienced and what practical issues emerged.

Qualitative description is not methodologically inferior. Its quality depends on whether the descriptive aim is explicit, sampling is appropriate, analysis is systematic and claims are proportionate.

11.9 Generic Qualitative Research

A generic qualitative study uses qualitative methods without fully adopting a recognised methodological tradition. This can be defensible when the research problem does not require the distinctive commitments of phenomenology, grounded theory, ethnography or narrative inquiry.

However, “generic qualitative” must not become an excuse for methodological vagueness. The researcher should still explain the philosophical position, sampling logic, data-generation method, analytical strategy and quality procedures.

11.10 Historical Qualitative Research

Historical qualitative research examines past events, institutions, practices and interpretations through documents, archives, oral histories and other evidence. It requires source criticism: researchers must evaluate provenance, purpose, context, authenticity, silences and bias.

The goal is not to compile a chronology. Strong historical research develops an interpretation of change, continuity, causation or contested meaning over time.

11.11 Comparative Qualitative Research

Comparative qualitative research examines similarities and differences across cases, settings, groups or periods. Comparison can reveal mechanisms that are invisible within a single context. The cases should be selected through a clear logic—for example, most-similar, most-different, typical, contrasting or theoretically significant cases.

The researcher must balance within-case depth with cross-case comparison. Jumping too quickly to a matrix of themes may erase each case's internal logic.

12. Methodology Comparison Table

Methodology	Primary purpose	Typical unit	Common evidence	Expected output
Case study	Understand a bounded case in context	Organisation, programme, event or process	Interviews, documents, observation, records	Contextual explanation of the case
Phenomenology	Understand lived experience	Shared phenomenon	Deep interviews, diaries, reflective accounts	Essence or interpretation of experience
Grounded theory	Develop theory of a process	Actions, interactions and processes	Iterative interviews and observations	Substantive explanatory theory
Ethnography	Understand culture and practice	Culture-sharing group or setting	Prolonged observation, interviews, artefacts	Cultural interpretation
Narrative inquiry	Understand experience through stories	Story, life, career or event	Narrative interviews, texts, biographies	Narrative interpretation
Action research	Improve practice while generating knowledge	Local practice or organisation	Cycles of action, observation and reflection	Practical change and transferable learning
Participatory action research	Generate knowledge and action collaboratively	Community or stakeholder group	Co-produced data and action cycles	Shared learning, empowerment and change
Qualitative description	Provide a rich descriptive account	Experience, service or practice	Interviews, focus groups, documents	Accessible descriptive findings
Generic qualitative	Explore meaning without adopting a full tradition	Topic, group or setting	Interviews, focus groups, documents	Themes or interpretive categories

13. Decision Framework: Which Methodology Should You Choose?

Is the study centred on one bounded system in context? Consider case study.
Is the study centred on lived experience? Consider phenomenology.
Is the aim to develop a theory of a process? Consider grounded theory.
Is the focus a culture-sharing group and its practices? Consider ethnography.
Are stories, temporality and identity central? Consider narrative inquiry.
Does the study seek to improve practice through cycles of action? Consider action research.
Must participants share control over knowledge and action? Consider participatory action research.
Is a clear, practice-oriented description sufficient? Consider qualitative description.
Does no established tradition fully fit? A carefully justified generic qualitative design may be appropriate.

Best practice: Write one sentence completing this structure: “This study uses [methodology] because the research aims to [specific analytical purpose], which requires [distinctive feature of methodology].” If the sentence is weak, the methodology may be poorly matched.

14. Worked PhD Examples

Example A: Digital transformation in a public organisation

Question: How did employees across three departments experience and interpret the implementation of an AI-enabled workflow system?

Possible design: A multiple-case study if each department is treated as a bounded case and contextual comparison is central. Phenomenology would be less suitable if the primary goal is to explain departmental variation rather than the essence of individual experience.

Example B: Identity after career disruption

Question: How do senior professionals experience and make sense of involuntary career transition?

Possible design: Interpretive phenomenology if lived experience is central; narrative inquiry if the researcher examines how participants construct career stories across time.

Example C: Adoption of a new professional practice

Question: What process explains how novice teachers adopt, adapt or reject AI-supported assessment practices?

Possible design: Grounded theory because the aim is to develop an explanatory model of a process.

Example D: Informal norms in an emergency department

Question: How do clinicians negotiate formal protocols and informal workarounds during periods of extreme pressure?

Possible design: Focused ethnography because routine practice, interaction and cultural norms are central.

15. Common Methodology Selection Errors

Choosing methodology after data have already been collected.
Using “case study” merely because the study takes place in one organisation.
Calling any interview study phenomenological.
Claiming grounded theory while using a fixed sample and thematic analysis.
Calling short-term observation ethnography.
Collecting stories but analysing only cross-sectional themes.
Calling research participatory when participants only answer questions.
Combining methodologies without explaining compatibility.
Describing methods but never articulating methodology.
Selecting a prestigious label that does not fit the research question.

16. Examiner Checklist for Methodology Choice

Examiner question	Evidence expected
Why this methodology?	A direct link to the research aim and question
What makes the study an example of this methodology?	Distinctive design and analytical features
Which tradition or version is used?	Clear conceptual and procedural positioning
How did methodology shape sampling?	Sampling logic consistent with the design
How did it shape data collection?	Methods suited to the type of knowledge sought
How did it shape analysis?	An analytical process aligned with the methodology
What claims can the methodology support?	Bounded and proportionate conclusions

Final examiner test: If the name of the methodology were removed from the chapter, could a knowledgeable reader still identify it from the research design, sampling, data generation and analysis? If not, the methodology may exist only as a label.

17. Designing a Qualitative Study

A strong qualitative study is designed through alignment. The research problem, research questions, methodology, sampling strategy, data-generation method and analytical approach must all point in the same direction. Many weak qualitative theses contain individually reasonable choices that do not form a coherent whole.

Alignment principle: Every major design choice should be justified by the type of understanding the research question requires.

18. Developing Strong Qualitative Research Questions

Qualitative research questions should invite exploration of meaning, process, experience, interpretation, context or interaction. They should be focused enough to guide the study, but open enough to allow participants and evidence to reveal complexity.

18.1 Characteristics of a strong qualitative question

It addresses a clear research problem.
It requires depth rather than numerical estimation.
It identifies the phenomenon, context or participants of interest.
It avoids assuming the answer in advance.
It can be answered through qualitative evidence.
It is consistent with the selected methodology.
It is feasible within the available time and access.

18.2 Useful qualitative question forms

Purpose	Useful wording	Example
Experience	How do participants experience…?	How do first-generation doctoral students experience academic belonging?
Meaning	What meanings do participants attach to…?	What meanings do public-sector employees attach to algorithmic monitoring?
Process	How does a process unfold…?	How do managers adapt to the introduction of AI-supported decision systems?
Interpretation	How do participants interpret…?	How do nurses interpret organisational expectations during service redesign?
Context	How does context shape…?	How does institutional culture shape ethical decision-making?
Construction	How is something constructed or negotiated…?	How is professional credibility negotiated in interdisciplinary teams?

18.3 Weak and improved questions

Weak: What is the effect of leadership on employee motivation?

Improved qualitative version: How do employees interpret leadership practices during organisational restructuring?

Weak: Does remote work improve productivity?

Improved qualitative version: How do employees and managers define and negotiate productivity in hybrid working arrangements?

Weak: What are the factors affecting student retention?

Improved qualitative version: How do students describe the experiences and institutional conditions that shape their decisions to remain at or leave university?

Common mistake: Replacing “does” with “how” does not automatically make a question qualitative. The underlying purpose must genuinely require interpretation or contextual understanding.

19. Aligning Research Questions with Methodology

Different qualitative methodologies answer different kinds of questions. The methodology should not be selected independently of the research question.

Methodology	Question focus	Example question
Case study	A bounded case in context	How did three municipal agencies implement an AI-supported procurement system?
Phenomenology	Lived experience	How do employees experience involuntary redeployment after automation?
Grounded theory	Process and theory generation	What process explains how supervisors respond to repeated project failure?
Ethnography	Culture, norms and practice	How do informal norms shape clinical decision-making in an emergency unit?
Narrative inquiry	Stories, identity and temporality	How do entrepreneurs narrate identity after business closure?
Action research	Change through iterative action	How can a department improve feedback practices through collaborative cycles of intervention?
Qualitative description	Accessible account of experience or practice	How do patients describe barriers to using a new digital health service?

20. Building a Coherent Research Aim and Objectives

The research aim states the overall purpose of the study. Objectives divide that purpose into concrete analytical tasks. In qualitative research, objectives should not read like a list of data-collection activities.

20.1 Weak objectives

To conduct interviews with managers.
To analyse interview data.
To review company documents.

These describe procedures rather than intellectual objectives.

20.2 Stronger objectives

To explore how managers interpret the introduction of algorithmic decision support.
To identify organisational conditions that shape acceptance or resistance.
To compare how interpretations differ across departments.
To develop an explanatory framework linking implementation context, professional identity and adoption behaviour.

Examiner test: Objectives should reveal what the researcher intends to understand, explain, compare or develop—not merely what the researcher intends to do.

21. Purposive Sampling

Purposive sampling selects participants, cases or documents because they can provide information relevant to the research question. It is the dominant sampling logic in qualitative research. The goal is not numerical representativeness but information richness.

21.1 Why purposive sampling is appropriate

A qualitative researcher may deliberately recruit people with direct experience of a phenomenon, responsibility for a process, exposure to contrasting conditions or knowledge of a particular institutional context. Selection is therefore guided by analytical relevance.

21.2 Common forms of purposive sampling

Strategy	Purpose	Example
Criterion sampling	Select all cases meeting a defined condition	Employees who used a new AI system for at least six months
Maximum variation	Capture diverse perspectives across important dimensions	Participants from different departments, grades and locations
Homogeneous sampling	Study a relatively similar group in depth	Early-career nurses in one specialist unit
Typical-case sampling	Examine an ordinary or representative case	A mid-sized public agency with average implementation outcomes
Extreme or deviant-case sampling	Learn from unusual success, failure or experience	A project that succeeded despite severe resource constraints
Critical-case sampling	Select a case with strong theoretical or practical significance	A site where a policy should work if it can work anywhere
Expert sampling	Recruit participants with specialised knowledge	Regulators, system architects or policy designers

21.3 Writing a defensible purposive sampling justification

A strong justification should specify:

the unit being selected;
the criteria for inclusion;
why those criteria matter analytically;
which dimensions of variation are important;
how access and feasibility affected selection;
how the sample supports the research question.

Weak justification: “Purposive sampling was used because participants were easy to access.”

Stronger justification: “Participants were selected purposively because they had direct responsibility for implementing the digital workflow and had used the system for at least six months. Maximum variation was introduced across department, managerial level and implementation outcome to capture contrasting experiences.”

22. Criterion Sampling

Criterion sampling includes participants or cases that meet one or more predefined conditions. The criteria should be derived from the research question rather than chosen merely to narrow recruitment.

22.1 Good inclusion criteria

Direct experience of the phenomenon
Minimum duration of exposure
Relevant role or responsibility
Participation during a defined period
Membership of a clearly specified setting or group

22.2 Exclusion criteria

Exclusion criteria should protect relevance and ethics. For example, employees who joined after an implementation may be excluded from a study of transition experience. However, excessive criteria may create an artificially narrow sample.

Common mistake: Criteria are sometimes reported without explaining why they matter. Each criterion should have an analytical or ethical purpose.

23. Maximum Variation Sampling

Maximum variation sampling deliberately includes cases that differ across dimensions likely to shape the phenomenon. The aim is not to represent every subgroup statistically. It is to identify patterns that persist across variation and explain differences where they do not.

23.1 Choosing variation dimensions

Relevant dimensions may include:

age or career stage;
organisational role;
department or location;
length of experience;
implementation success;
gender or social position;
type of service or institution.

The researcher should avoid collecting demographic variety merely for appearance. Variation should be linked to a plausible analytical reason.

24. Homogeneous Sampling

Homogeneous sampling selects participants who share key characteristics. It is useful when the aim is to examine a specific group's experience in depth or create a focused group discussion.

For example, a study may recruit only newly appointed school principals because the research concerns the first year of leadership transition. The narrowness of the sample is then a deliberate strength rather than a limitation.

25. Snowball Sampling

Snowball sampling asks participants or gatekeepers to identify other potential participants. It is especially useful for hidden, dispersed or difficult-to-access populations.

25.1 Strengths

Facilitates access where formal sampling frames do not exist.
Uses trust within social networks.
Can identify knowledgeable participants unknown to the researcher.

25.2 Risks

Participants may refer people similar to themselves.
Networks may reproduce social or organisational bias.
Confidentiality can be difficult where referrals reveal group membership.
Gatekeepers may control who is included.

Examiner concern: Snowball sampling should not be presented as automatically appropriate. The researcher must discuss network bias and steps taken to diversify recruitment.

26. Convenience Sampling

Convenience sampling recruits participants because they are readily available. It is often criticised because ease of access may replace analytical relevance. However, convenience can be a practical component of recruitment when acknowledged transparently and combined with clear eligibility criteria.

26.1 When convenience may be defensible

The study is exploratory and access is genuinely constrained.
Participants still meet analytically relevant criteria.
The limitations of the accessible setting are discussed.
Claims are appropriately bounded.
Additional purposive variation is introduced where possible.

Weak statement: “Convenience sampling was used because the participants were available.”

More defensible statement: “Access was limited to two participating organisations; within those sites, criterion and maximum variation sampling were used to recruit employees with relevant experience across roles and departments.”

27. Theoretical Sampling

Theoretical sampling is most closely associated with grounded theory. It involves selecting new participants, incidents or data sources in response to emerging analytical categories. The purpose is to develop the theory, not to make the sample demographically balanced.

For example, if early interviews suggest that middle managers act as translators between policy and frontline practice, the researcher may recruit additional middle managers or examine meetings where translation occurs.

Important distinction: Theoretical sampling cannot normally be planned fully before data collection because it depends on what emerges during analysis.

28. Sampling Cases, Documents and Events

Qualitative sampling is not limited to people. Researchers may sample:

organisations or departments;
policy documents;
meeting records;
online discussions;
critical incidents;
time periods;
observational settings;
media texts or visual materials.

The same logic applies: the researcher should explain why the selected materials can illuminate the research question.

29. Recruitment Strategy

Recruitment is the practical process through which the intended sample is approached. A sound recruitment plan should specify who will make contact, what information participants will receive, how voluntariness will be protected and how power relationships will be managed.

29.1 Recruitment channels

Direct invitation
Organisational gatekeepers
Professional associations
Community organisations
Research advertisements
Online forums or social platforms
Participant referral

29.2 Gatekeepers

Gatekeepers can facilitate access, but they may also shape the sample or create pressure to participate. Researchers should prevent managers from selecting only favourable participants and should ensure that individuals can decline confidentially.

Examiner question: “How did you ensure that organisational gatekeepers did not determine whose voices were included?”

30. Package 3.1 Design Checklist

Design element	Check
Research question	Requires qualitative understanding rather than numerical estimation
Methodology alignment	Question wording and analytical purpose fit the chosen methodology
Research aim	States the overall intellectual purpose clearly
Objectives	Describe what will be understood, explained or developed
Sampling logic	Participants, cases or materials are selected for analytical relevance
Selection criteria	Inclusion and exclusion criteria are justified
Variation	Relevant similarities and differences are considered deliberately
Recruitment	Access, gatekeepers and voluntariness are addressed
Claims	Intended conclusions fit the scope and composition of the sample

Package 3.1 outcome: By this stage, the study should have a defensible question, aligned methodology, clear aim and objectives, and a purposive sampling strategy that identifies who or what can provide the evidence required.

31. Sample Size in Qualitative Research

Qualitative sample size is determined by the study's purpose, methodological design, participant diversity, quality of data and analytical depth. It is not determined through statistical power calculations because qualitative research does not usually estimate population parameters.

Core principle: The appropriate sample is large enough to answer the research question convincingly and small enough to permit deep, rigorous analysis.

31.1 Why there is no universal number

A sample of eight participants may be sufficient for a tightly focused phenomenological study involving long, detailed interviews with a relatively homogeneous group. The same number may be inadequate for a multi-site case study comparing several organisational roles and settings. Therefore, sample size should be justified in relation to the design rather than defended by quoting a generic number.

31.2 Sample size is not a quality indicator by itself

More interviews do not automatically produce stronger research. A large sample can create superficial analysis when the researcher lacks time to examine contradictions, context and meaning. Conversely, a very small sample may produce unsupported claims when the study addresses a broad or diverse population.

32. Factors That Influence Sample Size

32.1 Scope of the research question

Narrow questions usually require fewer participants than broad questions. A study of one specific professional transition within one organisation may require fewer participants than a study comparing experiences across occupations, institutions and countries.

32.2 Methodological tradition

Each qualitative methodology places different demands on sampling. Phenomenology often privileges depth of lived experience. Grounded theory may require iterative recruitment as categories develop. Case study research may include relatively few cases but many evidence sources within each case.

32.3 Homogeneity and heterogeneity

A homogeneous sample may reach adequate depth with fewer participants because participants share key characteristics. A heterogeneous sample generally requires more participants to explore meaningful variation across roles, settings or experiences.

32.4 Quality and depth of data

Long, reflective interviews with highly knowledgeable participants may contribute more information than brief interviews with participants who have limited exposure to the phenomenon. The researcher should consider information richness rather than counting interviews alone.

32.5 Analytical strategy

Fine-grained narrative, discourse or phenomenological analysis is labour intensive and may justify a smaller sample. Broad thematic mapping may accommodate a larger sample, although analytical depth must still be protected.

32.6 Number of comparison groups

Studies comparing departments, occupations, demographic groups or sites need sufficient evidence within each group. A total sample may appear adequate while individual comparison groups remain too thin to support credible interpretation.

32.7 Access, ethics and feasibility

Practical constraints matter but should not be disguised as methodological principles. Researchers should state constraints honestly and narrow the research claims where recruitment is limited.

33. Common Sample-Size Myths

Myth	Why it is misleading
Qualitative studies always need 20–30 interviews	No number is appropriate across all questions, methodologies and populations.
A larger sample is always stronger	Excessive data can reduce analytical depth and coherence.
Ten participants are enough for phenomenology	Adequacy depends on depth, focus, participant relevance and the form of phenomenology used.
Saturation proves the sample is valid	Saturation must be defined, monitored and demonstrated rather than merely asserted.
Published sample sizes can be copied	Other studies may have different aims, contexts and analytical strategies.
Recruitment targets must never change	Qualitative sampling may evolve in response to emerging analysis.

34. Indicative Sample-Size Considerations by Methodology

The following guidance is not a set of mandatory numerical rules. It shows how the sampling logic changes across methodologies.

Methodology	Primary consideration	Typical planning logic
Phenomenology	Depth and commonality of lived experience	A focused group with rich first-hand experience
Grounded theory	Development of categories and explanatory theory	Initial sample followed by theoretical sampling
Case study	Depth within each bounded case	Few cases with multiple participants and evidence sources
Ethnography	Range of practices, roles and field situations	Sampling people, events, settings and time periods
Narrative inquiry	Completeness and depth of individual stories	Small number of information-rich narratives
Qualitative description	Coverage of relevant experiences and practical variation	Enough participants to produce a useful descriptive account
Action research	Participation in the local change process	Those directly involved in action and reflection cycles

Important: Do not present indicative ranges as universal rules. Explain why the planned sample is appropriate for this particular study.

35. Planning an Initial Sample

Researchers often need to provide an initial recruitment target before data collection begins. This is acceptable when the number is presented as a reasoned starting point rather than a guaranteed final total.

35.1 A defensible planning sequence

Define the exact phenomenon and population.
Identify the methodology and analytical approach.
Decide whether comparison across groups or sites is required.
Assess expected participant diversity.
Estimate the likely depth of each interview or evidence source.
Set an initial recruitment target.
Explain how adequacy will be reviewed during analysis.

35.2 Example justification

“The study will initially recruit approximately 15–20 participants. This target reflects the focused research question, the relatively homogeneous participant group and the use of in-depth interviews. Recruitment and analysis will proceed concurrently, and the final sample will be determined by the adequacy and richness of the developing categories rather than by the numerical target alone.”

36. Information Richness Versus Numerical Coverage

Information-rich participants have relevant experience, can describe it in detail and can illuminate the research question. Strategic selection of such participants may be more valuable than recruiting a larger number with weak or indirect knowledge.

Researchers should nevertheless avoid equating seniority with information richness. Frontline workers, service users, marginalised participants and dissenting voices may reveal processes that formal experts cannot see.

37. Depth Versus Diversity

Qualitative sampling frequently involves a trade-off between studying a narrow group deeply and including wider variation. The correct balance depends on the study aim.

Design priority	Sampling implication
Understand one shared experience deeply	Use a more homogeneous and focused sample
Compare experiences across roles or settings	Increase variation and ensure evidence within each subgroup
Develop an explanatory process	Recruit iteratively as emerging categories require
Describe service-user perspectives for practice	Include the principal forms of relevant experience

38. Worked Examples

Example A: Focused phenomenological study

A study explores the lived experience of first-year doctoral candidates returning to education after a 15-year career. A relatively small, homogeneous sample may be defensible because the phenomenon and participant criteria are tightly focused and interviews are intensive.

Example B: Multi-department organisational study

A study compares perceptions of AI monitoring among managers, technical staff and frontline employees across four departments. A small overall sample would be inadequate if it leaves only one or two participants in each subgroup. The sampling plan must support both within-group depth and cross-group comparison.

Example C: Grounded theory study

A researcher begins with participants involved in digital transformation projects. Early analysis indicates that informal intermediaries play a critical role. Additional participants are then recruited specifically to develop this category. The final sample cannot be determined fully in advance because theoretical sampling is part of the design.

39. Examiner Expectations

Examiners normally expect the researcher to explain:

why the planned sample fits the research question;
how methodology influenced sample size;
whether important variation was captured;
how recruitment developed during data collection;
how the researcher judged that the evidence was adequate;
what limitations remain because of the sample.

Weak thesis statement: “Twenty interviews were conducted because previous studies used similar samples.”

Stronger thesis statement: “The initial target was based on the focused question, participant homogeneity and anticipated interview depth. The adequacy of the sample was reviewed through concurrent analysis, attention to variation and assessment of whether additional interviews materially developed the findings.”

40. Package 3.2.1 Checklist

Question	Check
Is the research question narrow or broad?	The sample reflects the study's actual scope
Is the participant group homogeneous?	Expected similarity or diversity is explained
Are comparisons planned?	Each group or case has adequate evidence
How deep will each data source be?	Interview or observational richness is considered
Does the methodology require iteration?	The plan permits sampling to evolve
Is the target provisional?	The final sample will be based on analytical adequacy
Are limitations acknowledged?	Claims remain proportionate to the evidence

41. Saturation in Qualitative Research

Saturation is often used to justify when qualitative data collection can stop, but the term is frequently used too loosely. Researchers should specify what kind of saturation they are assessing, how it was monitored and why additional data were judged unlikely to change the analysis materially.

Core principle: Saturation is not a magic point reached automatically after a certain number of interviews. It is an analytical judgement supported by evidence.

42. Data Saturation

Data saturation refers to the point at which additional data no longer produce substantially new information relevant to the research question. It is most useful in focused qualitative studies where the phenomenon, participant group and analytical scope are clearly defined.

42.1 What data saturation does not mean

It does not mean that every possible perspective has been captured.
It does not mean that participants repeat exactly the same words.
It does not mean that all variation has disappeared.
It does not mean that the researcher has stopped hearing anything interesting.
It does not mean that a predetermined interview quota has been reached.

42.2 Evidence of data saturation

Researchers may demonstrate saturation by showing that later interviews confirm, elaborate or challenge existing patterns without introducing materially new dimensions. A simple saturation table or interview-by-interview coding log can help make this judgement transparent.

Interview stage	Analytical development	Interpretation
Early interviews	Many new codes and tentative categories	Sampling should continue
Middle interviews	Categories deepen and variation becomes clearer	Targeted recruitment may be needed
Later interviews	Few genuinely new insights; mostly confirmation or refinement	Saturation may be approaching
Final interviews	No material change to core findings	Stopping may be justified

43. Code Saturation and Meaning Saturation

Code saturation occurs when no substantially new codes are identified. Meaning saturation occurs later, when the researcher has developed a sufficiently rich understanding of the dimensions, variation and significance of those codes.

Important distinction: A study may reach code saturation while still lacking meaning saturation. Knowing the main topics is not the same as understanding how and why they matter.

43.1 Example

After ten interviews, a researcher may repeatedly identify the codes “loss of autonomy,” “monitoring anxiety” and “managerial mistrust.” This may suggest code saturation. However, further interviews may reveal that loss of autonomy is experienced differently by junior staff, technical experts and managers. Meaning saturation requires enough depth to interpret these differences.

44. Theoretical Saturation

Theoretical saturation is associated primarily with grounded theory. It is reached when the major categories are well developed, relationships among them are established and further data no longer add important properties to the emerging theory.

44.1 Theoretical saturation requires

concurrent data collection and analysis;
constant comparison;
memo writing;
theoretical sampling;
development of category properties and relationships;
an integrated explanatory model.

A study cannot credibly claim theoretical saturation if all participants were recruited in advance, data were analysed only after collection ended, or the final output is a set of descriptive themes rather than an explanatory theory.

Examiner criticism: “The thesis claims theoretical saturation, but no theoretical sampling occurred and the categories were not developed beyond thematic description.”

45. Saturation Across Different Methodologies

Saturation is not equally appropriate for every qualitative methodology. Some approaches use different standards of adequacy.

Methodology	How adequacy is commonly judged
Grounded theory	Theoretical saturation or sufficient development of core categories
Qualitative description	Coverage of relevant experiences and practical variation
Case study	Depth, triangulation and adequate explanation within each case
Phenomenology	Richness and depth of the lived experience rather than simple repetition
Narrative inquiry	Completeness and interpretive depth of individual narratives
Ethnography	Sufficient immersion to understand practices, meanings and variation
Discourse analysis	A sufficiently developed corpus to support interpretation of discursive patterns

46. When Saturation Is a Poor Fit

Saturation may be a poor fit when the study aims to preserve uniqueness, examine rare cases, analyse individual life stories, investigate evolving events or explore highly diverse experiences. In such studies, researchers should use a more appropriate concept such as depth, completeness, information richness, case adequacy or analytical sufficiency.

47. Practical Saturation Monitoring

A defensible monitoring process can be built into the research design.

Begin coding early rather than waiting until all data are collected.
Maintain a record of new codes and category development after each interview or batch.
Identify underdeveloped categories.
Recruit participants who can illuminate gaps or contrasting experiences.
Assess whether later data materially alter the findings.
Document the decision to stop and its limitations.

47.1 Saturation log template

Data source	New codes	New category properties	Contradictions	Sampling implication
Interview 1	High	Initial concepts	None yet	Broaden sample
Interview 5	Moderate	Categories developing	Role differences emerging	Recruit contrasting roles
Interview 10	Low	Relationships clearer	One negative case	Test negative case
Interview 14	Minimal	No material development	Existing variation confirmed	Consider stopping

48. Negative Cases and Saturation

A claim of saturation should not depend only on repeated confirmation. Negative or deviant cases can expose weaknesses in emerging explanations. Researchers should actively look for evidence that challenges the dominant pattern before deciding that the analysis is adequate.

Good practice: Saturation is stronger when the researcher has tested categories against contrasting participants, settings or incidents rather than repeatedly interviewing very similar people.

49. Common Saturation Errors

Claiming saturation without defining it.
Using saturation as a synonym for sample size.
Stopping because recruitment became difficult.
Equating repetition with analytical depth.
Ignoring subgroup variation.
Claiming theoretical saturation in a non-grounded-theory study.
Making the decision after data collection without a documented process.
Failing to consider contradictory evidence.

50. Reporting Saturation in a Thesis

A strong methods section should explain when analysis began, what was monitored, how new information was assessed, whether targeted recruitment occurred and why later data no longer changed the interpretation materially.

50.1 Weak reporting

“Saturation was reached after 15 interviews.”

50.2 Stronger reporting

“Interviews were coded concurrently with data collection. After interview 11, no substantially new codes were identified, although additional interviews were conducted to explore variation across managerial levels. Interviews 12–15 refined existing categories but did not alter their properties or relationships. Recruitment therefore ended when both code recurrence and category depth were judged sufficient for the study aim.”

51. Examiner Checklist for Saturation

Examiner question	Evidence expected
What type of saturation is claimed?	Clear definition appropriate to the methodology
How was saturation monitored?	Concurrent coding, logs, memos or category tracking
Was variation explored?	Evidence from contrasting participants or cases
Were negative cases considered?	Challenges to emerging explanations were examined
Why did recruitment stop?	An analytical rather than purely practical justification
Are limitations acknowledged?	Recognition that saturation is a reasoned judgement, not certainty

Final examiner test: Could another researcher understand how the stopping decision was made and judge whether it was reasonable? If not, the saturation claim is too vague.

52. Information Power in Qualitative Sampling

Information power offers an alternative to relying only on saturation. The central idea is that the more relevant information a sample holds for the research question, the fewer participants may be needed. Sample adequacy therefore depends not only on how many people are recruited, but on how well the sample, data quality and analytical strategy fit the study aim.

Core principle: A smaller sample may be sufficient when the study aim is narrow, participants are highly relevant, interviews are rich, theory is well specified and the analytical strategy is intensive.

53. The Five Main Dimensions of Information Power

Dimension	Higher information power	Lower information power
Study aim	Narrow and focused	Broad and exploratory
Sample specificity	Participants have direct and highly relevant experience	Participants are diverse or only indirectly connected
Use of theory	Established theory guides focused inquiry	Little prior theory and wide conceptual exploration
Quality of dialogue	Rich, reflective and detailed interviews	Brief, superficial or constrained accounts
Analysis strategy	In-depth case-oriented analysis	Broad cross-case mapping across many groups

53.1 Narrow versus broad study aims

A narrowly defined study usually requires fewer participants because each interview contributes directly to a specific question. A broad study examining several phenomena, contexts or stakeholder groups will usually need more evidence.

53.2 Sample specificity

Participants with direct, sustained and relevant experience contribute more information than participants selected only because they are available. Specificity can be increased through clear inclusion criteria, purposive recruitment and attention to contrasting cases.

53.3 Role of theory

A study guided by a clear theoretical framework may require fewer participants because data collection and analysis are more focused. However, theory can also constrain discovery if it is applied too rigidly. Researchers should explain whether theory is used deductively, inductively or abductively.

53.4 Quality of dialogue

Interview quality depends on trust, participant experience, question design, probing skill and the time available for reflection. Twenty weak interviews may contain less usable information than ten intensive interviews with knowledgeable participants.

53.5 Analytical strategy

In-depth phenomenological, narrative or case-oriented analysis often works with smaller samples because each case receives substantial attention. Broader thematic comparison across many groups generally requires a larger and more varied sample.

54. Sample Adequacy

Sample adequacy asks whether the evidence is sufficient to support the intended claims. It is a broader concept than saturation because it includes relevance, diversity, depth, negative cases and the fit between data and methodology.

54.1 Signs of an adequate sample

Participants or cases are directly relevant to the research question.
Important dimensions of variation are represented.
Data are sufficiently rich to support interpretation.
Major categories are supported by several pieces of evidence.
Contradictory or deviant cases have been explored.
The sample fits the methodology and analytical approach.
Claims are proportionate to the breadth of the evidence.

Examiner insight: An adequate sample is not necessarily a large sample. It is a sample that allows the researcher to answer the research question convincingly and transparently.

55. Analytical Sufficiency

Analytical sufficiency is reached when the researcher has enough evidence to develop a coherent and defensible interpretation. It focuses on whether the analysis is sufficiently developed rather than whether all possible data have been exhausted.

55.1 Questions for judging analytical sufficiency

Are the central themes or categories clearly defined?
Are relationships among themes explained?
Is there adequate evidence for each major claim?
Have alternative explanations been considered?
Have negative cases refined the analysis?
Does the interpretation answer the research question fully?
Would additional data likely change the core explanation materially?

56. Information-Rich Cases

Information-rich cases are cases from which the researcher can learn substantially about the phenomenon. They may be typical, critical, extreme, contrasting, expert, longitudinal or unusually revealing.

Case type	Why it may be information-rich	Example
Typical case	Shows how the phenomenon usually operates	An average-performing public agency
Extreme case	Reveals mechanisms under unusual success or failure	A digital project with exceptional adoption
Critical case	Tests a proposition under especially important conditions	A site where a policy should work if it can work anywhere
Contrasting case	Clarifies why outcomes differ	Two departments with opposite responses to the same reform
Expert case	Provides specialised knowledge	Regulators or implementation architects
Longitudinal case	Shows change over time	A team followed before, during and after restructuring

57. When to Stop Recruiting Participants

Recruitment may reasonably stop when the sample is sufficiently relevant, the important forms of variation have been explored, the analysis is well developed and additional data are unlikely to alter the core findings materially.

57.1 A practical stopping framework

Review the research question and intended claims.
Assess whether all important participant groups or case types are represented.
Examine whether new interviews still generate important codes or explanations.
Identify weak categories and recruit strategically where needed.
Test the developing interpretation against negative cases.
Document why further recruitment would add limited analytical value.

Do not confuse stopping with exhaustion: Recruitment should not end merely because the researcher is tired, time is limited or access has become difficult. Practical constraints may matter, but they should be reported honestly.

58. Sample-Size Justification Templates

58.1 Focused interview study

“The initial sample target reflected the narrow study aim, the specificity of the participant group and the anticipated depth of the interviews. Recruitment continued alongside analysis until the sample provided sufficient information power and additional interviews no longer altered the core interpretation materially.”

58.2 Multiple-group study

“The sample was designed to support comparison across the principal stakeholder groups. Recruitment continued until each group contained sufficient evidence for within-group interpretation and cross-group comparison, with particular attention to contradictory and minority perspectives.”

58.3 Case study

“Sample adequacy was judged at the level of the case rather than by participant count alone. Interviews, documents and observations were collected until the case could be described and explained in sufficient contextual depth, and the major findings were supported through multiple sources of evidence.”

58.4 Grounded theory

“An initial purposive sample was followed by theoretical sampling. Recruitment ended when the core categories were sufficiently developed, their relationships were established and additional data no longer contributed important properties to the emerging theory.”

59. Worked PhD Examples

Example A: Narrow phenomenological study

A researcher investigates the lived experience of senior surgeons returning to practice after long-term illness. The aim is narrow, participants are highly specific and interviews are intensive. A relatively small sample may possess high information power.

Example B: Broad public-sector comparison

A study explores AI adoption across multiple agencies, occupational groups and regions. Because the aim is broad and the sample heterogeneous, more participants and stronger within-group coverage are required.

Example C: Case-oriented organisational study

A researcher examines one failed digital transformation programme using interviews, meeting records, policy documents and observations. Adequacy is judged through the depth and triangulation of the case evidence rather than the interview count alone.

Example D: Theory-building study

A grounded theory project begins with frontline employees but later recruits middle managers and system designers because emerging categories suggest that translation between policy and practice is central. Sample size evolves through theoretical sampling.

60. Defending Sample Size in a Viva

In a viva, examiners may ask why the sample was not larger, how adequacy was judged or whether important voices were excluded. A strong response should connect the sample to the study aim, methodology, participant specificity, data quality and analytical strategy.

60.1 A strong viva response structure

State the purpose of the study.
Explain why the selected participants or cases were relevant.
Describe the depth and quality of the evidence.
Explain how adequacy was reviewed during analysis.
Discuss variation and negative cases.
Acknowledge remaining limitations.

Example viva answer: “I did not treat sample size as a numerical target. I evaluated the information power of the sample, the richness of the interviews and the adequacy of evidence across the main participant groups. Later interviews refined existing interpretations but did not materially change the core explanation. I have nevertheless limited my claims to the contexts represented.”

61. Examiner Decision Framework

Examiner question	Strong evidence	Warning sign
Why this number?	Justification linked to aim, sample specificity and analysis	A generic rule copied from another study
Why these participants?	Clear analytical relevance	Availability was the main criterion
Was important variation covered?	Deliberate inclusion of contrasting cases	Major groups are represented by one participant
How was adequacy judged?	Concurrent analysis, logs and category development	Sample size fixed without analytical review
Were negative cases considered?	Contradictions were actively examined	Only confirming evidence is presented
Are claims proportionate?	Conclusions are bounded by context	Broad generalisation from a narrow sample

62. Package 3.2.3 Checklist

Question	Check
Is the study aim narrow or broad?	The sample reflects the scope of the question
Are participants highly specific?	Relevance to the phenomenon is clear
Is the dialogue rich?	Interviews or observations provide sufficient depth
Does theory focus the study?	The role of theory is explained
Does the analysis fit the sample?	Depth and breadth are balanced appropriately
Has important variation been covered?	Contrasting and negative cases are considered
Can stopping recruitment be defended?	The decision is analytical and documented

Conclusion

Qualitative research is strongest when every decision is connected. The research question should require depth and context; the philosophical assumptions should explain what can be known; the methodology should guide the design; and the analysis should demonstrate how interpretation emerged from evidence. This package has developed the foundations of qualitative study design, including research questions, methodology alignment, aims, objectives, purposive sampling and recruitment. Package 3.2.1 examined qualitative sample-size planning and adequacy. Package 3.2.2 added data saturation, code saturation, meaning saturation, theoretical saturation, negative cases and practical saturation monitoring. Package 3.2.3 now adds information power, sample adequacy, analytical sufficiency, information-rich cases, stopping rules, justification templates and viva defence guidance.

Evaluate the methodological coherence of your research

Use Averon Research to identify gaps between your research questions, methodology, sampling, analysis and claims.

Evaluate Your Research