Research Philosophy and Design

Introduction

A brand management dissertation is not simply an extended marketing report. It is a systematic attempt to produce defensible knowledge about a branding problem. Whether the study examines brand equity, consumer trust, influencer credibility, repositioning, employer branding or customer experience, its quality depends on the relationship between the research question, philosophical assumptions, design, evidence and conclusions.

Research philosophy concerns assumptions about reality, knowledge and values. Research design concerns the overall plan through which evidence will be generated and interpreted. Methodology links the two: it explains why particular methods are appropriate for answering a specific question. A well-designed project therefore does not begin with the question, “Should I use interviews or a survey?” It begins with, “What do I need to know, what kind of knowledge would answer this question, and what evidence would justify my claims?”

In brand management, this reasoning is especially important because brands are simultaneously economic assets, cultural meanings, organisational promises and lived consumer experiences. A brand may be measured through awareness scores, interpreted through consumer narratives, evaluated through business outcomes or redesigned through collaborative innovation. No single philosophical perspective is automatically superior; suitability depends on the problem and the claims the researcher intends to make.

Learning Objectives

By the end of this chapter, you should be able to:

• distinguish ontology, epistemology, axiology, methodology and methods;
• compare positivist, interpretivist, pragmatic and critical perspectives;
• select exploratory, descriptive, explanatory, evaluative or design-oriented research approaches;
• justify qualitative, quantitative and mixed-method designs;
• align research questions, sampling, data collection and analysis;
• identify threats to credibility, validity, reliability and ethical integrity; and
• defend a coherent research design in a brand management dissertation proposal.

1. The Architecture of Research Decisions

Research design is a set of connected decisions rather than a list of techniques. Five concepts are useful.

Ontology concerns what reality is assumed to be. A positivist study may treat brand awareness as a phenomenon that can be measured consistently across respondents. An interpretivist study may regard brand meaning as socially constructed and differently experienced by consumer groups.

Epistemology concerns how knowledge can be obtained. A quantitative researcher may privilege systematic observation and measurement. An interpretivist may consider dialogue, context and reflexive interpretation essential to understanding.

Axiology concerns the role of values. Research is never entirely detached from choices about which problem matters, whose voices are included and how findings are used. A critical researcher makes such issues explicit; a positivist researcher attempts to control their influence through standardisation.

Methodology is the logic connecting philosophical assumptions to the research strategy. A case study, experiment, survey, ethnography or design-based inquiry is a methodology or research strategy, not merely a data-collection tool.

Methods are specific procedures, such as interviews, questionnaires, social-media observation, experiments, content analysis or statistical modelling.

These levels should be aligned. For example, a project asking, “How do first-generation university students interpret luxury branding?” is likely to require an interpretive methodology and qualitative methods. A project asking, “Does perceived brand authenticity predict purchase intention?” may justify a quantitative design. A project asking, “How can a retailer improve its loyalty programme for lapsed customers?” may require a pragmatic, mixed-method or design-oriented approach.

2. Major Research Philosophies

Positivism and post-positivism

Positivism assumes that reality can be observed, measured and explained through systematic procedures. It is associated with structured instruments, numerical data, hypothesis testing and an aspiration to minimise researcher influence. In brand management, positivist studies may measure awareness, satisfaction, perceived quality, purchase intention or campaign effectiveness.

A positivist design is appropriate when constructs can be operationalised and when comparison, prediction or estimation is central. For example, a researcher might test whether exposure to sustainability claims is associated with perceived brand trust. The researcher defines variables in advance, specifies a sample, collects standardised responses and uses statistical analysis.

Strict positivism has been criticised for treating social meanings as if they were natural objects and for underestimating the effect of context. Post-positivism responds by recognising that observations are fallible and that complete objectivity is unlikely. It retains systematic measurement while emphasising uncertainty, rival explanations, transparency and cautious claims. Thus, a survey can provide evidence of an association without proving that one variable causes another.

Interpretivism

Interpretivism views social reality as constructed through language, interaction, history and culture. Consumers do not merely receive brand meanings; they interpret, negotiate and sometimes resist them. Interpretivist research seeks contextual understanding rather than universal laws.

Typical methods include semi-structured interviews, focus groups, participant observation, digital ethnography, visual methods and textual analysis. A study of how migrant consumers negotiate the meaning of a national brand, for instance, would attend to identity, memory, community and cultural context.

Interpretivist quality is not assessed primarily through statistical generalisation. Instead, it depends on credibility, depth, transparency, reflexivity and the plausibility of the interpretation. The researcher should show how themes were developed, acknowledge alternative readings and explain the context in which findings apply.

Pragmatism

Pragmatism begins with the research problem and asks what combination of evidence will help address it. It does not require loyalty to a single conception of knowledge. A pragmatic researcher may combine survey results with interviews because numerical patterns show what is happening while qualitative accounts help explain why.

Pragmatism is particularly useful for applied brand projects involving decisions, implementation and evaluation. Suppose a brand team observes declining advocacy. A survey may estimate the scale and correlates of the decline; interviews may reveal service frustrations; customer-journey analysis may identify redesign opportunities. The methods are justified by their contribution to the decision, not by philosophical purity.

Pragmatism does not mean “use any method conveniently.” The researcher must still explain the assumptions, sequencing, integration and limitations of each component.

Critical perspectives

Critical research examines power, inequality, ideology and the consequences of apparently neutral practices. It asks who benefits from a brand system, whose identity is represented, whose labour is invisible and how marketing may reproduce exclusion. Critical perspectives draw on traditions such as critical theory, feminist research, postcolonial scholarship and intersectionality.

A critical brand study might examine whether inclusive advertising changes representation without changing organisational practice, or how algorithmic targeting excludes particular communities. Knowledge is understood as situated, and research often aims not only to interpret the world but also to support more equitable practice.

Critical research requires careful ethical reasoning. Researchers should avoid treating marginalised participants merely as sources of compelling stories. Participation, confidentiality, ownership of data and the potential consequences of publication must be considered. Critical work can use qualitative, quantitative or mixed methods; its defining feature is its attention to power and transformation, not a particular technique.

Comparing the perspectives

The perspectives can overlap. A project may use post-positivist measurement and critical analysis to study unequal campaign outcomes. What matters is that the researcher identifies tensions and explains how they are managed.

3. Choosing a Research Approach

The research approach should follow the question, not the researcher’s preferred technique.

Exploratory research

Exploratory research is suitable when a phenomenon is poorly understood, concepts are unclear or the problem needs initial definition. It asks questions such as: What meanings do consumers attach to a new category? How do employees understand a proposed brand purpose? What factors may influence resistance to rebranding?

Exploration commonly uses flexible interviews, focus groups, observation, online community analysis or an initial review of documents. Sampling is usually purposive rather than statistically representative. The outcome may be a conceptual framework, refined research question, set of themes or propositions for later testing. Exploratory findings should not be presented as prevalence estimates unless an appropriate quantitative design follows.

Descriptive research

Descriptive research establishes what is happening, to whom, where and under what conditions. It may estimate awareness levels, describe customer segments, map touchpoints or document the characteristics of brand advocates.

A descriptive survey requires clear population definition, appropriate sampling, carefully worded measures and attention to non-response. Descriptive work is more than presenting percentages: the researcher must explain the sample, measurement decisions and uncertainty, and distinguish the sample from the wider population.

Explanatory research

Explanatory research investigates relationships and possible causes. It may ask whether perceived value is associated with repurchase intention, whether message framing affects trust or why a campaign produces different responses across segments.

An experiment offers stronger evidence about causal effects when participants are assigned to conditions and competing explanations are controlled. Observational studies can identify associations and use statistical controls, but controls do not automatically establish causality. A dissertation should avoid causal language when the design cannot support it.

Evaluative research

Evaluative research assesses the merit, effectiveness, efficiency or consequences of a programme, campaign, platform or intervention. It may examine whether a loyalty programme achieved its objectives, whether a repositioning improved consideration or whether an internal brand-training programme changed employee behaviour.

Evaluation requires explicit criteria. These may include reach, engagement, conversion, customer experience, equity, inclusion, cost, unintended effects and sustainability. A before-and-after comparison is informative but vulnerable to external events and cannot by itself prove that the intervention caused change. Where feasible, comparison groups, repeated measurements or mixed evidence improve the evaluation.

Design-oriented research

Design-oriented research seeks to create or improve an artefact, service, process, experience or intervention. In brand management, the artefact might be a brand identity system, onboarding journey, loyalty mechanism, packaging concept or stakeholder communication framework.

The process commonly involves problem framing, stakeholder research, ideation, prototyping, testing and iteration. Its contribution includes both the developed solution and knowledge about the design principles that informed it. A design-oriented dissertation must specify how users were involved, how alternatives were assessed and what evidence supports the final design.

Approaches can be combined. For example, an exploratory phase may identify consumer tensions, a descriptive survey may estimate their prevalence, an explanatory experiment may test alternative messages, and an evaluative phase may assess implementation.

4. Qualitative, Quantitative and Mixed-Method Designs

Qualitative design

Qualitative research is appropriate when the objective is to understand meanings, processes, identities, experiences or social contexts. Data may include interview transcripts, field notes, images, advertisements, online posts and organisational documents.

A robust qualitative design specifies the setting, participant criteria, recruitment logic, interview or observation protocol, analytic procedure and reflexive position. Thematic analysis may involve familiarisation, coding, development of candidate themes, review, definition and reporting. These stages are not a mechanical recipe: interpretation requires attention to contradictions, context and negative cases.

Sample size is justified by information power, relevance and analytic sufficiency rather than by a universal numerical rule. Researchers should avoid claiming that a small purposive sample represents all consumers. Instead, they should explain transferability: which comparable contexts may reasonably learn from the findings?

Quantitative design

Quantitative research is appropriate when concepts can be operationalised and the project requires measurement, comparison, estimation or testing. It may use primary surveys, experiments, transaction records, social listening data or syndicated market datasets.

A quantitative proposal should define the unit of analysis, population, sampling frame, variables, measurement scales, data-quality checks and analysis plan. Reliability concerns consistency of measurement; validity concerns whether the measure supports the intended interpretation. A reliable measure can still measure the wrong construct.

For a descriptive proportion, an estimate from a sample is subject to sampling uncertainty. For a sample proportion p̂, a commonly used approximate standard error is:

Here, n is the sample size and the approximation is most defensible when the sample is reasonably independent and the normal approximation is suitable. This formula describes sampling variability under a probability-sampling logic; a very large non-probability sample does not automatically remove selection bias.

Researchers should report effect sizes, uncertainty and practical significance rather than relying only on a significance threshold. Cross-sectional association, for example, does not establish temporal order or causality.

Mixed-method design

Mixed-method research intentionally integrates qualitative and quantitative evidence within one project. Common designs include:

• Sequential exploratory: qualitative exploration informs a later survey or instrument.
• Sequential explanatory: quantitative results are followed by qualitative work to explain unexpected or important patterns.
• Convergent: qualitative and quantitative data are collected in a similar period and compared or integrated.
• Embedded: one method addresses a secondary question within a larger design, such as interviews embedded in an evaluation.

Integration is the defining feature. Merely placing interviews and survey results in separate chapters is not sufficient. Integration may occur through connecting samples, building one phase from another, merging findings in a joint display or developing meta-inferences that explain how the evidence fits together. Contradictions should be investigated, not hidden.

5. From Question to Design: A Practical Alignment Framework

A strong proposal moves through the following sequence:

1. Define the management problem. Distinguish the organisation’s problem from the knowledge gap. “Sales are falling” is a management symptom; the research problem may concern changing perceptions, channel experience or competitive positioning.
2. Review relevant literature. Identify constructs, debates, established measures and gaps. The review should inform—not predetermine—the design.
3. Frame the research question. Use exploratory wording for meanings, descriptive wording for distributions, explanatory wording for relationships, evaluative wording for outcomes and design wording for solutions.
4. Select the philosophical perspective. Explain assumptions about reality, knowledge and values.
5. Choose the approach and design. Specify whether the project is exploratory, descriptive, explanatory, evaluative or design-oriented, and whether it is qualitative, quantitative or mixed.
6. Plan sampling and access. Define who or what will provide evidence and why.
7. Specify collection and analysis. Ensure that each research question has an appropriate source of evidence and analytic method.
8. Address ethics and feasibility. Consider consent, privacy, data security, power relationships, resources and timetable.
9. Plan quality assurance. Include piloting, reflexivity, audit trails, measurement checks, triangulation or peer debriefing as appropriate.
10. Limit the claims. State what the design can establish and what it cannot.

A useful alignment table can expose weaknesses early.

6. Sampling, Data Quality and Research Ethics

Sampling must match the claim. Probability sampling supports statistical generalisation when the sampling frame, selection process and response are adequate. Non-probability sampling—purposive, quota, convenience or snowball—may be appropriate for qualitative, exploratory or hard-to-reach populations, but it limits claims about prevalence.

In brand research, researchers should consider not only customers but also employees, distributors, communities, suppliers and critics. Excluding relevant stakeholders can produce a distorted account of brand value and risk.

Ethical research requires informed consent, voluntary participation, confidentiality where promised, secure storage and a clear explanation of use. Online data are not automatically ethical to collect merely because they are publicly visible. Context, reasonable expectations of privacy, platform rules, re-identification risk and vulnerability must be assessed.

Brand projects often involve commercially sensitive information. Researchers should separate academic judgement from sponsor pressure, declare conflicts of interest and avoid allowing a client to suppress inconvenient findings. Generative AI may assist with transcription, coding support or language editing only under approved governance. Sensitive transcripts should not be uploaded to an external system without permission and appropriate safeguards. AI-generated interpretations must be checked against the source data; fabricated citations, invented participants or unverified summaries are unacceptable.

7. Quality and Limitations

Quantitative quality is commonly discussed through reliability, construct validity, internal validity, external validity and statistical conclusion validity. Qualitative quality may be supported through credibility, dependability, confirmability and transferability. These are not interchangeable checklists. Each reflects the logic of the chosen design.

Common threats include leading questions, common-method bias, non-response, social desirability, selective recruitment, researcher over-interpretation, retrospective memory error and confusing engagement metrics with meaningful brand outcomes. Triangulation can strengthen a study, but disagreement among sources may itself be theoretically important.

Every dissertation should state limitations honestly. A limitation is not a ritual apology; it defines the boundary of justified inference. For example, a single-city interview study may offer rich insight into local brand meanings but cannot estimate national prevalence. A large cross-sectional survey may identify relationships but cannot demonstrate how perceptions change over time.

8. Applied Case: Repositioning a Sustainable Fashion Brand

This is an illustrative case.

Background

GreenThread, a mid-priced fashion brand, has introduced a “responsible style” positioning. Management believes that younger consumers value the claim, but sales and trust scores have not improved. The company wants evidence before redesigning its communication and product information.

Problem

The management question is, “How should GreenThread improve its sustainability positioning?” The research problem is more specific: the organisation does not know whether low response arises from weak awareness, scepticism about credibility, confusing information, price concerns or a mismatch between the promise and consumer identity.

Analysis and design

A pragmatic, sequential mixed-method design is selected. First, twenty-four purposively sampled consumers—current customers, lapsed customers and non-customers—take part in semi-structured interviews. The researcher explores how participants interpret “responsible style,” what evidence they expect and which trade-offs they make.

The interviews identify three possible barriers: vague language, distrust of unsupported claims and the perception that responsible clothing is less fashionable. These themes inform a questionnaire measuring awareness, perceived credibility, fashionability, price concern and purchase consideration. The survey is distributed using a clearly documented sampling strategy, with quotas used to improve demographic balance while recognising that quota sampling does not guarantee population representativeness.

Finally, participants are randomly assigned to view one of three mock product pages: the existing page, a page with specific evidence and traceability information, or a page combining evidence with fashion-led imagery. Purchase consideration and perceived credibility are measured after exposure. Random assignment supports stronger causal inference about differences between the mock pages, although the artificial setting and self-reported outcome limit generalisation to actual purchases.

Solution and outcome

The integrated findings suggest that awareness is adequate but credibility is weak. The evidence-led page improves credibility, while the combined page performs best on both credibility and fashionability. Interviews also reveal that technical supply-chain information is difficult to understand. GreenThread therefore redesigns product pages using concise evidence, accessible explanations and visual styling consistent with the brand’s fashion promise. The company plans a later field evaluation using actual behavioural indicators rather than relying solely on intention measures.

Learning points

The case demonstrates that a management problem may require several forms of evidence. Interviews uncovered meanings and language barriers; the survey assessed the distribution of perceptions; the experiment tested alternative communications. The design is not automatically perfect: the sample, simulated setting and short-term outcome remain limitations. Its strength lies in explicit alignment, integration and appropriately bounded claims.

Best Practices for a Dissertation Proposal

• Begin with a researchable knowledge problem, not a preferred method.
• Define key constructs and distinguish them from managerial labels.
• Justify philosophical assumptions in relation to the question.
• Use literature to establish conceptual and methodological choices.
• Pilot interview guides, questionnaires and experimental materials.
• Keep an audit trail of recruitment, coding, exclusions and design changes.
• Report disconfirming evidence and plausible alternative explanations.
• Separate statistical association from causation and sample description from population inference.
• Treat participants and communities as rights-bearing contributors, not data sources.
• Make the design feasible within the time, access and competence available.

Chapter Summary

Research philosophy shapes assumptions about reality, knowledge and values, while research design translates those assumptions into a defensible plan. Positivist and post-positivist perspectives emphasise measurement and systematic explanation; interpretivism emphasises situated meanings; pragmatism selects methods according to the problem and intended use; and critical perspectives examine power, inequality and transformation.

Exploratory, descriptive, explanatory, evaluative and design-oriented approaches answer different kinds of questions and may be combined across project phases. Qualitative research is suited to meaning and context, quantitative research to measurement and comparison, and mixed-method research to deliberate integration of complementary evidence. A high-quality brand management dissertation aligns its question, philosophy, approach, sampling, methods, analysis, ethics and claims. It also recognises that every design has boundaries and that transparency about those boundaries is a sign of scholarly strength.

Key Terms

Ontology: Assumptions about the nature of reality.

Epistemology: Assumptions about how knowledge can be produced.

Axiology: The role of values in research.

Methodology: The logic connecting philosophical assumptions to research strategy.

Method: A specific procedure for collecting or analysing evidence.

Positivism: A perspective emphasising systematic observation and measurement.

Post-positivism: A view recognising fallibility while retaining systematic inquiry.

Interpretivism: A perspective focused on socially constructed meanings and context.

Pragmatism: A problem-centred perspective judging knowledge partly by practical consequences.

Critical research: Inquiry examining power, inequality, ideology and possible transformation.

Exploratory research: Research used to clarify poorly understood problems or concepts.

Descriptive research: Research documenting characteristics, distributions or conditions.

Explanatory research: Research investigating relationships and possible causes.

Evaluative research: Research assessing the merit or effects of an intervention or programme.

Design-oriented research: Research that creates and tests an artefact, process or intervention.

Qualitative research: Research using non-numerical evidence to understand meanings and processes.

Quantitative research: Research using numerical measurement for estimation, comparison or testing.

Mixed-method research: Deliberate integration of qualitative and quantitative evidence.

Triangulation: Comparing multiple sources, methods, researchers or perspectives.

Reflexivity: Critical examination of the researcher’s influence and assumptions.

Review Questions

Short Answer Questions

1. Distinguish ontology from epistemology.
2. What is the principal purpose of exploratory research?
3. Why does mixed-method research require integration?
4. Name two limitations of a cross-sectional survey.
5. What is meant by reflexivity in qualitative research?

Long Answer Questions

1. Compare positivist, interpretivist, pragmatic and critical perspectives in relation to brand research.
2. Explain how a researcher should choose among exploratory, descriptive, explanatory, evaluative and design-oriented approaches.
3. Discuss the strengths and limitations of qualitative and quantitative designs for studying brand authenticity.
4. Explain how sampling affects the claims that a dissertation can make.
5. Develop an alignment framework for a dissertation investigating declining customer loyalty.

Analytical / Application Questions

1. A brand wants to know why consumers distrust its sustainability claims and how a new message affects consideration. Propose and justify a sequential research design.
2. A survey finds a strong association between influencer exposure and purchase intention. Analyse what can and cannot be concluded, and propose a stronger design for assessing causality.
3. Design a critical research project examining whether a personalised advertising system produces unequal brand experiences across consumer groups.

Multiple Choice Questions with Answers

Question 1

Which philosophical perspective is most directly concerned with power, inequality and transformation?

A. Positivism

B. Interpretivism

C. Pragmatism

D. Critical research

Correct Answer: D. Critical research

Question 2

A study asking how consumers interpret a brand’s promise is primarily what type of research?

A. Exploratory

B. Descriptive

C. Evaluative

D. Experimental

Correct Answer: A. Exploratory

Question 3

What distinguishes mixed-method research from simply using two methods?

A. It always uses a very large sample

B. It integrates qualitative and quantitative evidence

C. It excludes philosophical assumptions

D. It must use an experiment

Correct Answer: B. It integrates qualitative and quantitative evidence

Question 4

Which design generally provides the strongest evidence for an effect of one message over another?

A. Cross-sectional observation without controls

B. Randomised experiment

C. Unstructured observation

D. Convenience interview study

Correct Answer: B. Randomised experiment

Question 5

What does ontology concern?

A. The nature of reality

B. The researcher’s timetable

C. The reliability coefficient only

D. The wording of a consent form

Correct Answer: A. The nature of reality

Question 6

A quota sample can improve demographic balance but does not necessarily provide what?

A. Any participant information

B. A probability-based representative sample

C. A research question

D. Qualitative data

Correct Answer: B. A probability-based representative sample

Question 7

Which approach is most directly focused on assessing whether a campaign achieved its objectives?

A. Evaluative research

B. Exploratory research

C. Ethnography

D. Discourse analysis

Correct Answer: A. Evaluative research

Question 8

Why should a researcher avoid claiming that a cross-sectional association proves causation?

A. Cross-sectional studies never contain numerical data

B. They may not establish temporal order or exclude alternative explanations

C. Associations cannot be reported ethically

D. Causation can only be studied qualitatively

Correct Answer: B. They may not establish temporal order or exclude alternative explanations

Question 9

What is a central ethical concern when using publicly visible online posts?

A. Public data require no ethical assessment

B. The researcher must consider privacy expectations and re-identification risk

C. Online data cannot be analysed

D. Only positive posts may be collected

Correct Answer: B. The researcher must consider privacy expectations and re-identification risk

Question 10

In the illustrative GreenThread case, why were interviews conducted before the survey?

A. To replace the need for sampling

B. To identify themes that could inform the questionnaire

C. To prove the communication caused sales

D. To guarantee national representativeness

Correct Answer: B. To identify themes that could inform the questionnaire