Research Paradigms & Philosophical Foundations
Session at a Glance
Positivism, Interpretivism, Pragmatism, Design Science, Critical Realism — paradigm selection for BBA & BCA research problems
Supervisor allocation; Topic negotiation meetings; Paradigm alignment exercise
2 hrs Lecture + 12 hrs Lab/Project
Supervisor allocated; Topic shortlisted
Learning Objectives
By the end of this session, you will be able to:
- Explain why research philosophy matters — how paradigm assumptions shape every subsequent methodological choice
- Compare and contrast five major research paradigms: positivism, interpretivism, pragmatism, design science, and critical realism
- Identify the ontological, epistemological, and axiological commitments of each paradigm
- Select an appropriate paradigm for a given research problem and justify the choice with reasoned argument
- Align your tentative capstone topic with a paradigm and articulate the fit
Session Planner
Suggested breakdown of the 4-hour contact session.
| Time | Segment | Activity | Mode |
|---|---|---|---|
| 0:00–0:10 | Opening | Recap Week 1; quick onion review; today's roadmap | Whole class |
| 0:10–0:35 | Lecture 1 | Why philosophy matters; Positivism & Interpretivism — core beliefs, examples from BBA & BCA | Lecture |
| 0:35–0:55 | Lecture 2 | Pragmatism, Design Science Research (DSR), Critical Realism — the fuller paradigm landscape | Lecture |
| 0:55–1:10 | Activity | Paradigm diagnosis: classify 6 research scenarios with justification | Pairs |
| 1:10–1:25 | Discussion | Share classifications; debate edge cases; facilitator synthesis | Whole class |
| 1:25–1:40 | Break | — | — |
| 1:40–2:00 | Lab Briefing | Supervisor allocation process; topic negotiation worksheet; paradigm alignment exercise | Demo |
| 2:00–3:30 | Lab Work | Meet supervisors (or review profiles); complete topic negotiation worksheet; paradigm alignment | Individual |
| 3:30–3:50 | Discussion | Share preliminary topics; paradigm peer-check; facilitator observations | Whole class |
| 3:50–4:00 | Exit Ticket | Self-assessment + paradigm commitment for your topic | Individual |
1. Why Philosophy Matters in Research
A paradigm is a set of basic beliefs that define, for the researcher, the nature of the world (ontology), what can be known about it (epistemology), and how to go about investigating it (methodology). It is the philosophical lens through which you see your research problem — often implicitly. Making it explicit is what separates a rigorous capstone from a naive one.
1.1 Three Foundational Questions Every Researcher Must Answer
| Question | Philosophical Term | What It Asks |
|---|---|---|
| What is the nature of reality? | Ontology | Is there one objective reality out there, or are there multiple realities constructed by people? Can social and technical phenomena be studied as "things" with properties, or must they be understood as fluid and context-dependent? |
| How can we know about reality? | Epistemology | What counts as valid knowledge? Can we discover laws through measurement and hypothesis testing? Or must we interpret meanings through immersion and dialogue? What is the researcher's relationship to what is being studied — detached observer or engaged participant? |
| What is the role of values? | Axiology | Should research be value-free (objective, unbiased), or is all research value-laden (shaped by the researcher's beliefs, culture, and choices)? How should ethics and values influence the research process? |
Imagine you're looking at the same city through different lenses. A positivist uses a telescope from a distance — measuring, counting, testing relationships between variables. An interpretivist walks the streets and talks to residents — understanding lived experience from the inside. A design scientist sketches a new bridge — creating something that didn't exist before and evaluating whether it works. A pragmatist switches lenses depending on what the problem demands. The city is the same; what you see depends on how you choose to look.
1.2 The Cost of Not Thinking About Paradigms
When students skip the philosophical layer, three common failures occur:
- Methodological incoherence: Using interpretivist language ("I want to understand lived experience...") but deploying only Likert-scale surveys with no qualitative depth.
- Mismatched evaluation criteria: Evaluators applying positivist standards (generalizability, sample size) to an interpretivist case study — or vice versa.
- Shallow contribution: Findings that don't answer the right kind of question because the researcher never clarified what kind of knowledge they were trying to produce.
2. Positivism — The Empirical Tradition
Ontology: A single, objective reality exists independent of our perception.
Epistemology: Knowledge is discovered through observation, measurement, and hypothesis testing — the researcher is a detached, neutral observer.
Axiology: Research should be value-free; personal beliefs must not influence findings.
Typical methods: Surveys, experiments, structured observation, secondary data analysis, benchmarking.
2.1 Core Commitments
- Causality and laws: The goal is to identify cause-and-effect relationships that generalize beyond the specific sample — "If X increases, Y decreases, all else equal."
- Hypothesis testing: Start with theory, derive testable hypotheses, collect data, confirm or reject. This is the hypothetico-deductive model.
- Measurement: Concepts must be operationalized into measurable variables. "Customer satisfaction" becomes a score on a validated scale. "System performance" becomes throughput and latency metrics.
- Replicability: Another researcher following your procedure should get similar results. This is the gold standard of credibility.
- Generalizability: Findings from the sample should apply to the population — this requires careful sampling and statistical inference.
2.2 Positivism in Practice
| Dimension | BBA Example | BCA Example |
|---|---|---|
| Typical question | "Does ESG disclosure score predict firm profitability among NIFTY 500 companies?" | "Does model compression ratio predict inference accuracy degradation on mobile devices?" |
| Data | Financial ratios, ESG scores, market data — all numeric, from databases | Accuracy scores, latency measurements, compression ratios — all numeric, from benchmark runs |
| Analysis | Regression, correlation, t-tests, ANOVA | Statistical comparison tests, correlation, regression, performance metrics |
| Output | "ESG disclosure has a significant positive effect on ROA (β = 0.34, p < 0.01), controlling for firm size and leverage." | "8-bit quantization reduces accuracy by 2.3% on average (p < 0.001, paired t-test) while reducing model size by 73%." |
| Limitation | Cannot explain why ESG affects profitability — only that it does | Cannot explain why certain layers are more sensitive to quantization — only that they are |
Choose positivism when: (a) clear, measurable variables exist, (b) you can collect data from a sample that represents a population, (c) your goal is to test a theory or hypothesis, and (d) you need findings that generalize. It is the dominant paradigm in finance, marketing analytics, software engineering experiments, and ML benchmarking.
3. Interpretivism — The Meaning-Making Tradition
Ontology: Reality is socially constructed — there are multiple realities, each shaped by people's experiences, culture, and context.
Epistemology: Knowledge is created through understanding meanings, interpretations, and lived experiences — the researcher is an engaged interpreter, not a detached observer.
Axiology: Research is inherently value-laden; the researcher's perspective shapes the inquiry — reflexivity is essential, not a weakness.
Typical methods: In-depth interviews, ethnography, phenomenology, case study, narrative inquiry, focus groups.
3.1 Core Commitments
- Meaning over measurement: The goal is to understand what things mean to people, not to measure how much of something exists. "Why do developers contribute to open source?" rather than "How many hours do developers contribute?"
- Context is everything: People and phenomena cannot be understood in isolation. The same behaviour in different contexts may have entirely different meanings.
- The researcher as instrument: You are not a neutral data collector — your background, assumptions, and interactions with participants shape the data. Reflexivity (examining your own role) is a requirement, not an option.
- Rich, thick description: The output is not a p-value but a narrative — detailed accounts that make the phenomenon vivid and intelligible to the reader.
- Transferability, not generalizability: Findings may resonate with other contexts, but the reader judges applicability, not the researcher.
3.2 Interpretivism in Practice
| Dimension | BBA Example | BCA Example |
|---|---|---|
| Typical question | "How do first-generation women entrepreneurs in Tier-2 Indian cities experience and navigate gendered expectations while building their ventures?" | "How do agile development teams experience and adapt to the introduction of AI coding assistants in their daily workflow?" |
| Data | 20 in-depth interviews, field notes, participant diaries | 30 interviews, 6 months of meeting observations, team retrospectives |
| Analysis | Thematic analysis, interpretative phenomenological analysis (IPA), narrative analysis | Thematic analysis, grounded theory coding, ethnographic narrative |
| Output | "Four superordinate themes: (1) performing legitimacy, (2) strategic invisibility, (3) leveraging kinship networks, (4) reframing gender as advantage." | "Teams progress through three phases: initial enchantment, integration friction, and emergent partnership — each characterized by distinct emotional and practical negotiations." |
| Limitation | Findings are context-bound; cannot claim statistical generalizability | Observer presence may have altered team behaviour; findings specific to the studied teams |
Choose interpretivism when: (a) your research question starts with "How" or "What is the experience of...", (b) you want to understand a phenomenon in depth from participants' perspectives, (c) context and meaning are central (not peripheral) to the question, and (d) the existing theory is thin — you need to build rich understanding before you can measure.
4. Pragmatism — The "What Works" Tradition
Ontology: Reality is both objective and constructed — the pragmatist sidesteps the debate and focuses on practical consequences.
Epistemology: Knowledge is what has practical value — "truth is what works." The best method is the one that answers the question, regardless of its philosophical origins.
Axiology: Values drive the inquiry; the researcher chooses what to study based on what matters practically.
Typical methods: Mixed methods — survey + interview, experiment + focus group, quantitative metrics + qualitative user feedback.
4.1 Core Commitments
- Methodological pluralism: You are not forced to choose between quantitative and qualitative — use both if the question demands it. Pragmatism is the philosophical home of mixed-methods research.
- Abductive logic: Move back and forth between data and theory, surprising facts and possible explanations. Neither purely deductive (positivist) nor purely inductive (interpretivist).
- Problem-centred: The research question drives everything. Start with the problem, then ask: "What combination of methods will give me the most complete answer?"
- Practical consequences as truth criterion: A finding is "true" if it works in practice — if it solves the problem, informs the decision, or improves the situation.
A BBA student studying fintech adoption among MSMEs might survey 400 owners (quantitative — positivist) to identify key adoption drivers, then interview 15 of them (qualitative — interpretivist) to understand why those drivers matter and how trust is constructed through peer networks. The philosophical justification: "The research question has both a 'what' dimension (drivers of adoption) and a 'why' dimension (the meaning of trust), and neither method alone can answer both." This is the pragmatist's strength — complementarity.
5. Design Science Research (DSR) — The Artefact Tradition
Ontology: Knowledge can be created through the act of designing — a working artefact is itself a knowledge claim.
Epistemology: "We know what we can build and make work." Knowledge is embodied in artefacts (algorithms, systems, models, frameworks, methods) and their evaluated performance.
Axiology: Research should solve real problems — utility and relevance are central values. The artefact must address a problem that matters to some community.
Typical methods: Artefact design & development, experimental evaluation, benchmarking, case-based evaluation, simulation, user studies.
5.1 The DSR Process Model (Peffers et al. 2007)
5.2 Artefact Types (Hevner et al. 2004)
| Artefact Type | What It Is | BBA Example | BCA Example |
|---|---|---|---|
| Constructs | Vocabulary and symbols used to describe problems and solutions | "Customer journey friction" — a new construct for describing points of drop-off in B2B sales | "Technical debt sentiment" — a construct for measuring developer perception of code quality erosion |
| Models | Representations of reality that aid problem understanding and solution design | A predictive model of startup failure based on founder, market, and funding variables | An architectural model for privacy-preserving federated learning in healthcare IoT |
| Methods | Defined processes or algorithms for performing tasks | A structured method for conducting due diligence on social enterprise investments | A novel consensus algorithm for permissioned blockchain networks |
| Instantiations | Working systems, tools, or prototypes that demonstrate feasibility | A functional ESG scoring dashboard for SME supply chain assessment | A working prototype of a load-balancing middleware for edge computing |
| Design Theories | Generalized knowledge about how to design artefacts of a class | Design principles for creating effective financial literacy interventions for low-income populations | Design principles for building explainable AI interfaces for clinical decision support |
DSR is the most important paradigm for BCA students (it's the dominant paradigm in computer science and information systems research) and increasingly relevant for BBA students designing frameworks, models, or decision tools. The key distinction: building something is development; building something AND rigorously evaluating it AND extracting design principles that generalize is DSR. Your capstone's contribution is not the artefact itself — it's the knowledge embedded in the artefact and validated through evaluation.
6. Critical Realism — The Depth Tradition
Ontology: Reality has three layers: the real (underlying mechanisms and structures that cause events), the actual (events that occur, whether we observe them or not), and the empirical (what we actually observe and experience).
Epistemology: We can study the real layer, but our knowledge of it is always fallible and socially mediated — hence "critical" realism, not naive realism.
Axiology: Research should aim for emancipation — revealing hidden structures that constrain people and organizations.
6.1 The Iceberg Metaphor
Positivism studies the tip of the iceberg (what is observable and measurable). Interpretivism studies how people experience the tip. Critical realism asks: what is beneath the water — what hidden structures and mechanisms generate the observable surface?
- BBA example: You observe a gender pay gap in an organization (empirical layer). The events include hiring, promotion, and compensation decisions (actual layer). The real layer includes patriarchal social structures, unconscious bias in evaluation processes, and institutional policies that systematically disadvantage women — even when no individual intends to discriminate.
- BCA example: You observe that certain demographic groups receive higher error rates from a facial recognition system (empirical layer). The events are the classification decisions (actual layer). The real layer includes training data imbalances, historical biases embedded in datasets, and engineering choices that optimized for majority-group performance.
Critical realism is well-suited for capstone projects that study complex socio-technical systems — situations where technology and human organization interact (e.g., "Why did the ERP implementation fail despite following best practices?" or "Why do developers resist adopting static analysis tools despite evidence they find bugs?"). It is less common at the undergraduate level but powerful when the research question demands going beneath the surface.
7. Paradigm Selection Guide — Which Paradigm Fits Your Problem?
The diagram below is a decision aid. Start with your research question and work through the questions to identify the paradigm that best fits.
→ YES → Design Science Research
→ NO → Continue to Q2
Q2: Does your research question ask about measurable relationships between variables (cause-effect, correlation, difference between groups)?
→ YES, and only this → Positivism
→ NO, or not only this → Continue to Q3
Q3: Does your research question ask about meanings, experiences, interpretations, or social processes from people's perspectives?
→ YES, and only this → Interpretivism
→ YES, but I also need to measure relationships → Pragmatism (mixed methods)
→ I need to go deeper — what hidden structures cause what I'm seeing → Critical Realism
Paradigm Comparison at a Glance
| Paradigm | Core Question | BBA Example RQ | BCA Example RQ | Signature Method |
|---|---|---|---|---|
| Positivism | "What causes what?" | Does CSR disclosure affect firm profitability among NIFTY 500 firms? | Does 8-bit quantization reduce BERT's accuracy on sentiment classification? | Large-N survey or controlled experiment with statistical testing |
| Interpretivism | "What does it mean?" | How do rural women entrepreneurs experience access to microfinance? | How do open-source maintainers experience burnout and recovery? | In-depth interviews with thematic analysis |
| Pragmatism | "What works, and why?" | What factors drive fintech adoption among MSMEs, and how is trust constructed? | Which CI/CD practice most improves deployment frequency, and how do teams adapt it? | Survey + follow-up interviews (mixed methods) |
| Design Science | "Can we build something that solves this?" | Can we design a predictive model for early-stage startup failure, and how well does it perform? | Can we design a privacy-preserving contact tracing protocol, and how does it perform under realistic load? | Artefact design + experimental evaluation |
| Critical Realism | "What hidden structures generate what we see?" | What underlying mechanisms explain persistent gender pay gaps despite equal-pay policies? | What structural factors in software organizations cause static analysis tools to be adopted but not used? | Multi-level case study with retroduction |
Think Deeper — Cross Questions
Discuss in pairs before sharing with the class.
A BCA student says: "I don't need philosophy — I'm building a sentiment analysis model, not doing philosophy." How would you respond? What philosophical assumptions is this student already making (without realizing it) by choosing to benchmark accuracy as the evaluation criterion?
Can a BCA student building a system (a DSR project) also adopt an interpretivist stance? Think of a scenario where understanding user experience or developer behaviour is essential to evaluating the artefact. How would you justify this paradigmatic combination?
A BBA student surveys 300 consumers about brand loyalty and reports: "65% of respondents are loyal to Brand X, therefore Brand X has strong customer loyalty." What paradigm is this student operating within, and what is the ontological problem with this conclusion? (Hint: think about what the survey actually measured vs. what it claims.)
Critical realism distinguishes the "real" from the "empirical." Give an example from your own discipline where the surface-level data (empirical) might mislead you about what is actually going on (real). How would critical realism change how you study it?
Quick Check — Paradigm Diagnosis
For each scenario, select the most appropriate paradigm. Click to reveal the answer.
1. A student wants to study whether remote work affects software developer productivity. She plans to collect commit frequency, pull request throughput, and lines of code from 50 GitHub repositories — 25 from remote-first teams and 25 from co-located teams — and compare means using t-tests.
2. A student designs and builds a novel load-balancing algorithm for Kubernetes clusters, implements it as a custom scheduler plugin, and evaluates it against the default scheduler using throughput and latency benchmarks under three workload patterns.
3. A student investigates fintech adoption among MSMEs. She administers a survey to 400 owners (measuring adoption drivers quantitatively) AND conducts 15 in-depth interviews (to understand how trust is constructed through peer networks). She integrates both in her findings chapter.
4. Through 20 in-depth interviews with first-generation women entrepreneurs in Tier-2 cities, a student explores how they experience and navigate gendered expectations. She uses interpretative phenomenological analysis (IPA) and produces rich themes.
5. A student observes that a company's diversity training program has no effect on hiring outcomes despite high trainee satisfaction scores. She investigates the underlying organizational structures, incentive systems, and cultural norms that may explain why surface-level training doesn't translate to behavioural change.
6. A student runs a controlled experiment with 60 developers to test whether AI coding assistants (GitHub Copilot) increase the number of security vulnerabilities in code. Half the participants use Copilot, half don't. All submissions are analysed with static analysis tools and manual code review.
Knowledge Check — Interactive Quiz
Test your understanding of research paradigms.
Q1. A researcher believes that "the social world cannot be understood by subsuming it under causal laws — it must be interpreted." Which philosophical stance does this statement reflect?
Q2. Which paradigm is the philosophical home of mixed-methods research — justifying the use of both quantitative and qualitative methods in a single study?
Q3. In the DSR process model (Peffers et al.), which step comes immediately AFTER "Design and Development"?
Q4. Critical realism differs from positivism primarily because it:
Q5. A BBA student writes: "I am studying the relationship between digital marketing spend and customer acquisition cost among Indian D2C brands. I will collect monthly spend and CAC data from 100 brands and run a regression analysis." Which paradigm is this student operating within?
Lab Activity — Supervisor Allocation & Topic Negotiation
Part A: Supervisor Allocation (First 2 lab hours)
- Supervisor profiles: Review the list of available supervisors, their research interests, and their discipline alignment (BBA/BCA/both).
- Preference submission: Submit your top 3 supervisor choices with a one-paragraph justification for each — explain why their expertise aligns with your tentative research area.
- Allocation: The course coordinator allocates supervisors based on preferences, workload balance, and disciplinary fit.
- First meeting: If allocated during this session, schedule your first supervisor meeting for Week 3. Prepare a one-page topic brief to bring to that meeting.
Part B: Topic Negotiation Worksheet
Complete this worksheet individually. It will form the basis of your first supervisor meeting and your Week 3 problem statement draft.
1. Broad Domain
What broad area interests you? (e.g., consumer behaviour, cybersecurity, fintech, ML, organizational culture, HCI, entrepreneurship, software architecture)
2. Why This Domain?
What draws you to this area? (personal experience, industry exposure, coursework, a specific problem you've observed, a gap you suspect exists)
3. Tentative Research Question(s)
Try to formulate 1–2 possible research questions. They don't need to be final — just a starting point. Use the formats: "What factors influence..." / "How does X affect Y...?" / "How do people experience...?" / "Can we design a... that...?"
4. Preliminary Paradigm Fit
Based on today's lecture, which paradigm seems most appropriate for your tentative question? Justify in 2–3 sentences.
5. What You Already Know
What existing knowledge, skills, or experience do you bring to this topic? (coursework, internships, personal projects, work experience, technical skills)
6. What You Need to Learn
What specific knowledge, skills, or resources will you need to acquire to pursue this topic? (statistical methods, a programming language, domain knowledge, access to data or participants)
7. Potential Supervisor(s)
Which faculty members have expertise in this area? (Check the supervisor profiles list.)
Part C: Paradigm Alignment Exercise
For your tentative research topic from Part B, complete the following:
- Ontological assumption: What kind of reality are you assuming? Is your phenomenon objectively measurable, or is it socially constructed? Or both?
- Epistemological stance: How will you know what you claim to know? Through measurement and testing? Through interpretation and understanding? Through building and evaluating?
- Paradigm declaration: Write one paragraph declaring your paradigm and justifying it with reference to your research question. This will go into your methodology chapter.
Example: "This study adopts a pragmatist paradigm. The research question — 'What factors drive fintech adoption among MSMEs, and how is trust constructed through peer networks?' — has both a quantitative dimension (identifying adoption drivers) and a qualitative dimension (understanding trust construction). Neither a purely positivist nor a purely interpretivist approach would adequately address both. Pragmatism justifies the convergent mixed-methods design that this study employs."
Exit Ticket
Complete before leaving. Submit to your facilitator.
- Name all five paradigms covered today and write one sentence describing each.
- Which paradigm do you think best fits your tentative capstone topic? Why?
- Rate your confidence in distinguishing positivism from interpretivism: Not confident / Somewhat confident / Confident / Very confident
- One paradigm-related question you still have:
- Your tentative capstone topic in one sentence (refined from the lab worksheet):
Key Takeaways — Week 2
Your paradigm is not an afterthought — it shapes what you study, how you study it, what counts as evidence, and how your contribution is judged. Choose consciously, justify explicitly.
Positivism tests causal laws. Interpretivism interprets meaning. Pragmatism uses what works. Design Science creates and evaluates artefacts. Critical Realism uncovers hidden structures.
Design Science Research frames artefact creation as knowledge contribution. It is the dominant paradigm in CS/IS research and increasingly used in business for framework and model development.
Your paradigm must align with your research question, which must align with your methods, which must align with your analysis. Incoherence is the fastest way to lose marks in your proposal defence.
Facilitator Notes
Preparation Checklist
- Prepare supervisor profiles list with research interests, past supervisions, and discipline alignment (BBA/BCA/both).
- Print the Topic Negotiation Worksheet for each student (or share as editable document).
- Prepare 2–3 worked examples of paradigm alignment — one BBA, one BCA, one cross-disciplinary — showing how RQ → paradigm → method flows logically.
- For DSR segment: have at least one published DSR paper ready to show (e.g., a short excerpt from a MIS Quarterly or ACM publication that clearly follows the Peffers process).
- Coordinate with the course coordinator about the supervisor allocation process — is it happening this week or next?
Common Student Difficulties
- Philosophy feels abstract and irrelevant: Counter with the "cost of not thinking about paradigms" examples from Section 1.2. Show real dissertation excerpts where paradigm incoherence led to poor evaluation.
- "I'm just building an app — I don't need a paradigm": This is the most important misconception to correct for BCA students. Walk through a DSR example showing how even a "build" project has philosophical commitments (What counts as "working"? How do you know it's better than alternatives?).
- Confusing pragmatism with "anything goes": Emphasize that pragmatism is not methodological anarchy — it requires rigorous justification of why mixing methods produces better knowledge than a single method.
- Critical realism overload: For most undergraduate capstones, the first four paradigms suffice. Critical realism can be mentioned as an advanced option but shouldn't be forced.
Pacing Tips
- The five paradigms can feel like too much content in one session. Prioritize positivism, interpretivism, and DSR — these cover 90%+ of undergraduate capstones. Pragmatism and critical realism can be lighter.
- The paradigm diagnosis exercise (6 scenarios) is high-value — budget enough time for discussion of edge cases.
- If supervisor allocation isn't finalized this week, focus lab time on the Topic Negotiation Worksheet and Paradigm Alignment Exercise — these are valuable regardless.