Research Design: Qualitative, Quantitative, and Mixed Methods

Philosophical worldviews, three approaches

Every research study rests on philosophical assumptions — beliefs about the nature of reality (ontology), knowledge (epistemology), values (axiology), and methodology. Creswell identifies four major worldviews: Post-positivism (objective reality exists and can be measured, typical of quantitative research), Constructivism (reality is socially constructed, typical of qualitative research), Transformative (research should address power imbalances and social justice), and Pragmatism (use whatever methods work for the problem at hand, typical of mixed methods).

The three research approaches — quantitative, qualitative, and mixed methods — differ fundamentally in their assumptions, strategies, and methods. Quantitative research tests theories by examining relationships among variables measured numerically and analyzed statistically. Qualitative research explores meaning through words, observations, and interpretation. Mixed methods integrates both to provide comprehensive understanding.

Choosing an approach depends on: the research problem (exploring vs testing), personal experience (comfort with statistics vs fieldwork), and audience (disciplinary conventions). A medical researcher testing drug efficacy would use quantitative methods; a sociologist understanding patient experiences would use qualitative methods; a health policy researcher might combine both.

• Four worldviews shape research assumptions: post-positivism, constructivism, transformative, pragmatism
• Quantitative tests theories; qualitative explores meaning; mixed methods integrates both
• Research approach should match the research problem, not personal preference

• Identify the worldview underlying 3 published papers in your field
• Write the same research question in quantitative, qualitative, and mixed methods forms
• Evaluate which approach best fits your thesis topic and justify your choice

Literature map, searching, synthesis

A literature review is not merely a summary of what others have written — it is a critical synthesis that identifies themes, debates, gaps, and methodological approaches in existing research. The review positions your study within the scholarly conversation, demonstrating what is known, what is contested, and what remains unexplored.

Systematic searching uses databases (Scopus, Web of Science, Google Scholar) with carefully constructed search strings combining keywords, Boolean operators (AND, OR, NOT), and filters (date range, language, document type). The PRISMA framework provides a transparent, reproducible process for documenting search strategy, screening criteria, and included/excluded studies.

A literature map is a visual diagram showing how your study relates to existing research. It identifies the major themes, key authors, methodological approaches, and theoretical frameworks in your area, with your proposed study positioned to address an identified gap. Creating this map early guides your writing and ensures comprehensive coverage.

• Literature reviews synthesize and critique, not just summarize
• Systematic searching with PRISMA ensures reproducible and comprehensive coverage
• Literature maps visually organize the scholarly landscape around your study

Variables, hypotheses, theoretical lens

Theory serves different purposes in each approach. In quantitative research, theory is tested — the researcher proposes hypotheses derived from existing theory and tests them against data. Variables (independent, dependent, control, mediating, moderating) operationalize theoretical constructs into measurable quantities. The hypothesis predicts the expected relationship between variables.

In qualitative research, theory may serve as a lens for interpretation (feminist theory, critical race theory) or emerge from the data through grounded theory methods. The theoretical framework contextualizes the study within existing scholarly conversations without predetermining findings.

Mixed methods studies may use theory at multiple stages — a quantitative phase may test theory while a qualitative phase explores mechanisms or contextual factors. The transformative-emancipatory framework uses theory to advocate for marginalized groups throughout the research process.

• Quantitative research tests theory; qualitative research may generate theory
• Variables operationalize abstract concepts into measurable quantities
• Theoretical frameworks position your study within existing scholarly discourse

Academic writing, APA, IRB approval

Academic writing follows disciplinary conventions for structure, citation style, and voice. APA (American Psychological Association) style dominates social and behavioral sciences. Key principles: use past tense for literature review and results, present tense for established findings and conclusions, active voice for clarity, and hedging language for claims beyond the data.

Research ethics requires Institutional Review Board (IRB) approval for studies involving human participants. Key ethical principles from the Belmont Report: respect for persons (informed consent, voluntary participation), beneficence (maximize benefits, minimize harm), and justice (fair distribution of research burden and benefits). Data privacy, confidentiality, and the right to withdraw are non-negotiable protections.

• Academic writing conventions vary by discipline — learn your field's expectations
• IRB approval is required before collecting data from human participants
• Ethical research follows principles of respect, beneficence, and justice

Problem statement, purpose, significance

The introduction establishes the research problem, provides context, states the purpose, and argues for the study's significance. A strong introduction follows a "deficiency model" — it identifies what is missing in existing knowledge and explains how your study addresses that gap. The problem statement is the single most important paragraph in the entire study.

The purpose statement is a concise declaration of the study's intent. Quantitative purpose statements identify the variables, population, and research site. Qualitative purpose statements name the phenomenon being explored and the participants/setting. Mixed methods purpose statements declare the overall intent and the rationale for combining approaches.

• The problem statement identifies the gap your study addresses
• Purpose statements must be clear, concise, and approach-appropriate
• Significance argues why the study matters to theory, practice, and policy

Quantitative, qualitative, mixed methods purpose

Forming testable questions

Research questions narrow the purpose statement into specific, answerable queries. Quantitative questions ask about relationships between variables ("Is there a relationship between X and Y?") or differences between groups ("Is there a difference in Y between Group A and Group B?"). They are accompanied by null hypotheses (no relationship/difference) and alternative hypotheses.

Qualitative questions ask one central question and several sub-questions that explore the phenomenon. They begin with "how" or "what" rather than "why" (which implies causation). The questions are deliberately broad to allow themes to emerge from the data rather than constraining investigation.

Mixed methods questions include both quantitative and qualitative questions plus a "mixing" question that explains how the two strands inform each other.

• Quantitative questions test relationships or differences with hypotheses
• Qualitative questions use "how/what" to explore phenomena openly
• Mixed methods require separate strand questions plus an integration question

Survey, experimental, quasi-experimental

Quantitative research designs include experimental (random assignment to treatment/control groups, establishing causation), quasi-experimental (comparison groups without random assignment), and non-experimental (surveys, correlational studies describing relationships without establishing causation). The choice depends on the research question and practical constraints.

Survey research collects data through questionnaires or interviews using closed-ended questions. Key design decisions include: population definition, sampling strategy (random, stratified, cluster), sample size determination (power analysis), instrument design (Likert scales, multiple choice), and administration method (online, mail, phone). Response rate and sampling bias are critical validity threats.

Experimental designs use random assignment, treatment manipulation, and control groups to establish causal relationships. Internal validity threats (history, maturation, testing effects, attrition) must be controlled. External validity (generalizability) often trades off against internal validity — tightly controlled laboratory experiments may not generalize to real-world settings.

• Experimental designs establish causation through random assignment and control
• Survey research requires careful attention to sampling, instrument design, and response rate
• Internal validity (causal inference) often trades off against external validity (generalizability)

Grounded theory, ethnography, case study, phenomenology

Five major qualitative approaches serve different purposes: Narrative research studies the lives of individuals through their stories. Phenomenology describes the essence of a lived experience shared by several people. Grounded theory develops theory from data through systematic coding and constant comparison. Ethnography studies a cultural group through prolonged immersion. Case study provides in-depth analysis of a bounded system (a person, organization, event).

Qualitative data collection uses interviews (structured, semi-structured, unstructured), observations (participant, non-participant), documents, and audiovisual materials. Purposeful sampling selects information-rich cases rather than representative ones. Saturation — the point where new data no longer reveals new themes — guides sample size decisions, typically 5-25 participants for most approaches.

Data analysis involves coding (assigning labels to data segments), categorizing (grouping codes into themes), and interpreting (explaining what the themes mean). Trustworthiness — the qualitative equivalent of validity — is established through credibility (member checking, triangulation), transferability (thick description), dependability (audit trail), and confirmability (reflexivity).

• Five qualitative approaches serve different research purposes
• Purposeful sampling selects information-rich cases; saturation guides sample size
• Trustworthiness replaces validity/reliability with credibility, transferability, dependability, confirmability

Convergent, explanatory sequential, exploratory sequential

Mixed methods research combines quantitative and qualitative approaches to provide a more complete understanding than either alone. Three core designs: Convergent parallel (collect quantitative and qualitative data simultaneously, then compare/merge results), Explanatory sequential (quantitative first, then qualitative to explain the numbers), and Exploratory sequential (qualitative first to develop an instrument or typology, then quantitative to test it).

The rationale for mixing must be explicit — complementarity (different facets of the same phenomenon), development (one phase informs the next), initiation (discovering paradoxes that lead to new questions), or expansion (extending breadth and range of inquiry). Simply collecting both types of data without a clear integration strategy is not mixed methods — it is parallel studies.

Integration — the point where quantitative and qualitative strands are intentionally combined — is the defining feature of mixed methods. Integration can occur during data collection (embedded design), data analysis (joint display tables), or interpretation (meta-inferences that synthesize both strands). Joint display tables that present quantitative results alongside qualitative themes are particularly effective.

• Mixed methods requires intentional integration, not just parallel data collection
• Three core designs: convergent parallel, explanatory sequential, exploratory sequential
• Joint displays effectively integrate quantitative and qualitative findings