Mixed methods research combines qualitative and quantitative data to provide a more complete understanding of complex research problems. This guide covers: 1) when to use mixed methods, 2) six core research designs (convergent, explanatory, exploratory, embedded, transformative, multiphase), 3) integration strategies using joint displays, 4) APA JARS-Mixed reporting standards, 5) quality assessment criteria, and 6) common mistakes to avoid. For graduate students conducting rigorous research, mixed methods offers comprehensive insights but requires careful planning and genuine integration.
Introduction: Why Combine Methods?
As a graduate student designing your research study, you’ve likely encountered the classic dilemma: should you use qualitative or quantitative methods? The answer might be both.
Mixed methods research (MMR) has evolved from a novel approach to a mainstream methodology across disciplines. According to Creswell and Plano Clark (2017), mixed methods is now widely accepted in education, health sciences, social sciences, business, nursing, and psychology. The fundamental premise rejects the “paradigm wars” of qualitative versus quantitative, instead adopting a pragmatic approach where the research question dictates the methods, not philosophical allegiance.
However, mixed methods is not simply “sprinkling” open-ended questions into a survey. It requires deliberate design, rigorous execution of both strands, and genuine integration. The hardest but most crucial step is integration—making connections between qualitative and quantitative data to produce meta-inferences that neither method alone could provide.
This comprehensive guide synthesizes current best practices from leading methodologists (Creswell & Plano Clark, Tashakkori & Teddlie, Fetters & Curry) and official standards (APA JARS-Mixed) to help you understand, design, implement, and report high-quality mixed methods research.
What is Mixed Methods Research?
Core Definition
Mixed methods research is both a method and methodology that involves:
“Collecting, analyzing, and mixing both quantitative and qualitative data in a single study or series of studies to provide a more comprehensive understanding of research problems than either approach alone.”
Key Components:
- Two distinct strands: Both qualitative and quantitative data are collected and analyzed separately using appropriate methods
- Genuine integration: Data are combined, connected, or embedded—not merely reported side by side
- Purposeful design: The mixing serves a specific research purpose aligned with the question
- Structured approach: Relies on specific, rigorous designs with clear integration points
Historical Evolution
While researchers combined methods for decades, modern mixed methods emerged as a distinct, systematic approach in the late 1980s and early 1990s. Pioneers like John W. Creswell, Vicki L. Plano Clark, Abbas Tashakkori, and Charles Teddlie developed detailed typologies and frameworks that transformed MMR from ad hoc combinations to sophisticated, fully integrated approaches.
The field has evolved from simple combination to contemporary standards emphasizing genuine integration through tools like joint displays (Fetters & Curry, 2013). The American Psychological Association’s JARS-Mixed guidelines (2020) now provide specific reporting standards, cementing mixed methods as a mature, respected methodology.
When Should You Use Mixed Methods?
Research Questions That Benefit
Mixed methods is particularly valuable when your research question requires BOTH numerical patterns AND rich, contextual understanding. Specifically:
- Explanatory questions: “Why do these patterns exist?” (Quantitative results need qualitative explanation)
- Exploratory questions: “What might be happening here?” (Qualitative insights need quantitative testing)
- Triangulation questions: “Do different methods tell the same story?” (Validation across methods)
- Complementary questions: “What do we get from one method that the other misses?” (Maximizing breadth and depth)
Decision Flow:
Do you need BOTH numerical patterns AND rich, contextual understanding?
├─ Yes → Mixed Methods appropriate
└─ No → Use single method (QUAL or QUAN)
Disciplines Using Mixed Methods
Mixed methods is now widely accepted across diverse fields:
- Education: Program evaluation, student learning outcomes, policy impact
- Health Sciences: Clinical trials with patient experiences, health services research (NIH increasingly requires mixed methods)
- Social Sciences: Community studies, behavioral research, public opinion
- Business: Market research, organizational studies, consumer behavior
- Nursing: Patient care outcomes, implementation research
- Psychology: Intervention studies, behavioral measurement
If you’re in any of these fields, learning mixed methods will enhance your research toolkit significantly.
Advantages and Limitations
Advantages:
- Increased validity through triangulation
- Enhanced understanding with both breadth and depth
- Ability to address complex, multifaceted research problems
- Compensation: one method offsets the weaknesses of the other
- Practical utility: results are often more actionable for stakeholders
Limitations:
- Time and resource intensive (requires expertise in both methods)
- Design, execution, and analysis are more complex
- Philosophical tensions may arise reconciling different paradigms
- Risk of superficial integration if not properly planned
- Some journals still prefer single-method studies (though this is changing)
Six Core Mixed Methods Designs
Choosing the right design is crucial. Here are the six core designs you should know, with examples to help you select appropriately.
1. Convergent Parallel Design
Purpose: Collect and analyze QUAL and QUAN data simultaneously, then merge findings to compare/validate results.
Timing: Concurrent (QUAL ↔ QUAN)
Process:
- Phase 1: Collect both data types at same time
- Phase 2: Analyze separately using appropriate methods
- Phase 3: Integrate/compare results (merge)
- Phase 4: Interpret convergences/divergences
When to Use: When you want to compare findings, validate results, or provide comprehensive description.
Example:
– Quantitative: Survey 200 patients on treatment satisfaction (numerical scores)
– Qualitative: Interview 15 patients about their experiences (themes)
– Integration: Compare themes with survey results in joint display to see if qualitative explanations align with quantitative patterns
2. Explanatory Sequential Design
Purpose: QUAN data collected first, QUAL data collected second to explain initial quantitative results.
Timing: Sequential (QUAN → QUAL)
Process:
- Phase 1: Quantitative collection and analysis
- Phase 2: Identify interesting/confusing results
- Phase 3: Qualitative collection to explain those results
- Phase 4: Integration and interpretation
When to Use: When survey results need explanation, outliers need exploration, or you want to drill down into unexpected patterns.
Example:
- Quantitative: Test scores show gender difference in math performance
- Qualitative: Follow-up interviews with students to understand why the difference exists
- Integration: Qualitative themes explain the quantitative pattern (e.g., classroom dynamics, confidence levels)
3. Exploratory Sequential Design
Purpose: QUAL data collected first to explore phenomenon, then QUAN data to test/measure those findings.
Timing: Sequential (QUAL → QUAN)
Process:
- Phase 1: Qualitative exploration (interviews, focus groups)
- Phase 2: Develop instruments/questions based on findings
- Phase 3: Quantitative testing with larger sample
- Phase 4: Integration and interpretation
When to Use: When phenomena are not well-understood, measurement tools don’t exist, or you need to develop instruments grounded in lived experience.
Example:
- Qualitative: Explore barriers to healthcare access in rural communities through interviews
- Quantitative: Develop and test survey based on emergent themes with larger sample
- Integration: Qualitative insights ground quantitative measurement, ensuring survey items reflect actual experiences
4. Embedded Design
Purpose: One data type has secondary priority within a larger primary method.
Timing: Can be concurrent or sequential
Process:
- Primary method: Main focus (QUAL or QUAN)
- Secondary method: Embedded to address specific sub-question
- Integration: During analysis phase
When to Use: When one method can’t answer all questions, but one method is dominant. Good for students with limited time/resources.
Example:
- Primary: Quantitative RCT of intervention effectiveness
- Embedded: Qualitative interviews with participants to understand mechanisms
- Integration: Qualitative data explains “how” and “why” of quantitative outcomes
5. Transformative Design
Purpose: Integrates social justice/action agenda throughout mixed methods study.
Key Features:
- Guided by transformative paradigm (feminist, critical race theory, disability rights)
- Mixed methods serve advocacy/social change goals
- Community-based participatory research often uses this design
- Focus on marginalized populations and power structures
When to Use: When your research aims to address inequities, empower communities, or drive social change.
6. Multiphase Design
Purpose: Multiple phases where each phase builds on previous ones (often used in program development/evaluation).
Structure: QUAL → QUAN → QUAL → QUAN (iterative cycles)
Example:
- Phase 1: Qualitative needs assessment
- Phase 2: Quantitative survey based on Phase 1
- Phase 3: Qualitative intervention feedback
- Phase 4: Quantitative outcome evaluation
When to Use: Complex, longitudinal projects involving program development, implementation, and evaluation.
Design Selection Decision Tree
What is your primary purpose?
├─ EXPLORATORY (understand phenomenon)
│ └─ Exploratory Sequential (QUAL → QUAN)
├─ EXPLANATORY (explain results)
│ └─ Explanatory Sequential (QUAN → QUAL)
├─ COMPARATIVE/TRIANGULATION
│ └─ Convergent Parallel (QUAL + QUAN simultaneous)
└─ COMPLEMENTARY WITH HIERARCHY
└─ Embedded (one primary, one secondary)
For students starting out, explanatory sequential or embedded designs are often most manageable, as they allow you to build on existing strengths while learning the other method.
Integration Strategies: The Key to Success
Integration is the deliberate connecting of qualitative and quantitative components. It’s the defining feature of true mixed methods research. Without genuine integration, you merely have two separate studies reported together.
The Three-Level Integration Framework
Fetters, Curry, and Creswell (2013) propose integration at three levels:
1. Design Level Integration
What: Choosing designs that inherently integrate methods (convergent, sequential, embedded)
How: During study planning phase, select a design that creates integration points by design
2. Methods Level Integration (The “Mixing”)
This occurs during data collection and analysis through four approaches:
- Connecting: One database informs sampling of the other (Example: Quantitative survey identifies extreme cases for qualitative follow-up)
- Building: Findings from first phase inform second phase data collection (Example: Qualitative themes become items on quantitative survey)
- Merging: Bringing datasets together for side-by-side comparison (Example: Joint displays comparing themes and statistics)
- Embedding: Integrated data collection at multiple points (Example: Qualitative observations during quantitative experiment)
3. Interpretation/Reporting Integration
Approaches:
- Narrative Integration: Weaving findings together in text
- Data Transformation: Converting data types (coding text → numeric categories)
- Joint Displays: Visual side-by-side presentation (most powerful and now required by many journals)
Joint Displays: Critical Tool for Integration
Definition: Tables or figures that display qualitative and quantitative results together to facilitate comparison and meta-inference.
Recommended Structure:
| Theme/Finding |
Quantitative Results |
Qualitative Insights |
Integration Interpretation |
| Student motivation |
72% reported high motivation (M=4.2/5, SD=0.8) |
Themes: autonomy, relevance, instructor feedback |
Mixed methods support that both structured instruction and supportive environment contribute to motivation |
| Program effectiveness |
Post-test scores: +15% improvement (p<.01) |
Participants described “turning point moments” of understanding |
Quantitative gains align with qualitative accounts of specific learning experiences |
Best Practices:
- Use headers that clearly indicate what’s being integrated
- Include both data types in same visual space
- Add “integration” column to indicate convergence/explanation/complementarity
- Always include narrative explanation of what the display shows
- Use color coding or shading to highlight patterns
Source: Fetters et al. (2013) Achieving Integration in Mixed Methods Designs—Principles and Practices (7460+ citations)
Practical Implementation Guide
Sampling Considerations
Key Decision: Same participants vs. different participants?
Sample Types
- Identical Sample: Same individuals participate in both phases
- Advantage: Direct links between datasets
- Disadvantage: Attrition risk, burden on participants
- Best for: Sequential designs where you need participants from Phase 1 in Phase 2
- Parallel Sample: Different but comparable participants for each strand
- Advantage: Reduced burden, can maximize each method’s sampling needs
- Disadvantage: Cannot link individual-level data
- Best for: Concurrent designs where population-level patterns matter
- Nested Sample: Subset of Phase 1 participants selected for Phase 2 (purposive sampling)
- Best for: Sequential designs where you need specific types of cases (outliers, extreme cases)
Sampling Strategies
Quantitative Strand:
- Probability sampling (random, stratified)
- Sample size based on statistical power analysis
- Representative of population
Qualitative Strand:
- Purposeful sampling (criterion, snowball, maximum variation)
- Sample size based on saturation (typically 15-30 participants)
- Information-rich cases
Integration Challenge: Aligning samples that serve different purposes. Solution: For sequential designs, use quantitative phase to identify participants for qualitative phase; for concurrent designs, parallel samples acceptable as long as they’re drawn from same population.
Data Collection Methods
Quantitative Methods: Surveys (questionnaires, standardized scales), experiments/quasi-experiments, secondary data analysis, structured observations, tests/measurements
Qualitative Methods: Individual interviews (semi-structured, unstructured), focus groups, observations (participant/non-participant), document analysis, case studies, field notes
Integration Tip: Coordinate timing—collect both simultaneously for convergent designs, or sequence deliberately for explanatory/exploratory designs.
Analysis Approaches
Step 1: Separate Analysis
Quantitative Analysis: Descriptive statistics (means, frequencies, distributions), inferential statistics (t-tests, ANOVA, regression, correlations), use appropriate software (SPSS, R, Stata)
Qualitative Analysis: Thematic analysis (most common for beginners), content analysis, grounded theory coding, discourse analysis, use qualitative software (NVivo, ATLAS.ti, Dedoose)
Principle: Each strand must stand on its own methodological rigor before integration.
Step 2: Integration Analysis
Comparing Results: Look for:
- Convergence: Agreement between methods
- Complementarity: Different aspects that together provide fuller picture
- Divergence: Conflict that reveals complexity (valuable, not problematic)
Addressing Discordance: When QUAL and QUAN results disagree:
1. Re-examine data quality for each strand
2. Consider contextual factors
3. Explore theoretical explanations
4. Present as valuable insight—divergence often reveals the complexity of the phenomenon
Data Display: Joint Displays in Action
Simple Example Joint Display:
| Theme/Finding |
Quantitative Results |
Qualitative Insights |
Integration |
| Student motivation |
72% reported high motivation (M=4.2/5, SD=0.8) |
Themes: autonomy, relevance, instructor feedback |
Mixed methods support that both structured instruction and supportive environment contribute to motivation |
| Program effectiveness |
Post-test scores: +15% improvement (p<.01) |
Participants described “turning point moments” of understanding |
Quantitative gains align with qualitative accounts of specific learning experiences |
Quality and Validity: Meeting Academic Standards
Mixed methods quality assessment must address both strands AND their integration.
Tashakkori & Teddlie’s Legitimacy Framework
Traditional validity/reliability criteria are insufficient for MMR. They propose legitimation (continuous quality assurance process):
- Sample Integration Legitimation: Ensure both samples allow for meta-inference; sample sizes appropriate; sampling strategies justified and integrated
- Weakness Minimization Legitimation: Demonstrate how one method offsets the other’s flaws (e.g., quantitative provides generalizability, qualitative provides depth)
- Sequential/Conversion Legitimation: For sequential designs, ensure quality of transforming findings from one phase to next
- Paradigmatic Mixing/Commensurability: Manage philosophical differences; show how integration is meaningful despite paradigm differences
- Methodological Connoisseurship: Researcher’s ability to appropriately use both approaches; requires expertise, training, or collaboration
Practical Quality Checklist
For Quantitative Strand
- Sample size justified (power analysis or saturation rationale)
- Appropriate statistical methods selected
- Validity/reliability of instruments reported
- Missing data addressed appropriately
- Statistical assumptions met and verified
For Qualitative Strand
- Sampling strategy described (purposeful, criterion, saturation)
- Data adequacy/saturation discussed
- Trustworthiness measures: credibility (triangulation, member checking), transferability (thick description), dependability (audit trail), confirmability (reflexivity)
- Reflexivity statement (researcher positionality)
- Coding process transparent (with examples)
For Integration
- Explicit integration plan in methods section
- Joint displays used in results (or narrative integration clearly described)
- Integration discussed throughout (not just in discussion section)
- Meta-inferences clearly derived from both datasets
- Discordant findings addressed, not ignored
APA JARS-Mixed Standards
The American Psychological Association’s Journal Article Reporting Standards for mixed methods (2020) are essential reading. Key requirements:
Method Section: Methodological purpose statement, rationale for mixed methods, integration description (when/how/priority), procedures (sequence diagram if sequential)
Results Section: Complete quantitative results, complete qualitative results, integration results with joint displays (required)
Discussion Section: Meta-inferences (synthesized conclusions), implications emerging from integration
Download the complete JARS-Mixed checklist: apastyle.apa.org/jars/mixed-methods
Common Student Mistakes (And How to Avoid Them)
Design and Planning Mistakes
Mistake 1: Insufficient Justification
Problem: “I used mixed methods because it’s trendy” or vague reasoning.
Solution: Explicitly state how the research question requires both QUAL and QUAN: “This question required both [quantitative pattern identification] and [qualitative understanding of mechanisms] because…”
Mistake 2: Choosing Wrong Design
Problem: Using convergent when sequential would better address the question.
Solution: Use the decision flowchart in this guide. Ask: “Does one phase inform the other?” → Sequential; “Do I need to compare/validate at same time?” → Concurrent; “Is one method clearly secondary?” → Embedded.
Mistake 3: Underestimating Time
Problem: Running out of time, rushed analysis, incomplete integration.
Solution: Double time estimates for a single-method study. For theses/dissertations, plan 1.5-2x longer. Consider embedded design if severely time-constrained.
Mistake 4: Lack of Integration Plan
Problem: Collect both datasets but don’t know how to connect them.
Solution: Before data collection, answer these questions:
– When will integration occur (design, data collection, analysis, interpretation)?
– How will datasets be connected (joint displays, transformation, narrative)?
– What is integration priority (equal, QUAL-dominant, QUAN-dominant)?
– Who will perform integration (single researcher or team)?
Analysis and Integration Mistakes
Mistake 7: Superficial Integration
Problem: “Sprinkling” open-ended questions in survey and calling it mixed methods.
Solution: Ensure: distinct QUAL and QUAN strands, genuine mixing with clear integration point(s), joint displays or explicit comparison in results section.
Mistake 8: Merging Too Early or Too Late
Too Early: Integrating before separate analysis is complete
Too Late: Integration only mentioned in discussion, not results
Solution: Integrate at multiple stages: design phase, data collection phase, analysis phase with joint displays, interpretation with meta-inferences
Mistake 9: Ignoring Discordance
Problem: Forcing convergence when QUAL and QUAN disagree.
Solution: Discordance is valuable data, not a problem. Explore why they differ (sampling bias? measurement? context?), present divergence as insight into phenomenon’s complexity, and quote: “When qualitative and quantitative findings diverge, we don’t force agreement; we investigate meaning.”
Writing and Reporting Mistakes
Mistake 10: Missing Critical Elements
Problem: Failing to report: philosophical foundation, integration plan, joint displays, researcher positionality/reflexivity.
Solution: Use APA JARS-Mixed checklist as pre-submission checklist; never omit joint displays.
Writing Up Mixed Methods: APA Standards and Best Practices
APA JARS-Mixed Structure
Abstract: Include “mixed methods” and design type (e.g., “convergent parallel mixed methods design”)
Introduction: Problem statement → Research questions; rationale for mixed methods; brief mention of design
Methods: Overall design; philosophical foundation; participants/sampling with sample alignment justification; data collection procedures (separate but linked); data analysis (separate for each, then integration); integration plan (when/where/how)
Results: Quantitative results; qualitative results; Integration section: Joint display(s) + narrative interpretation
Discussion: Summary of key findings; meta-inferences (integrated conclusions); implications (theoretical/practical); limitations (each strand + integration limitations); future research
Examples by Discipline
Education
Typical Question: “How do teaching methods affect student outcomes AND how do students experience these methods?”
Common Design: Explanatory sequential, convergent
Example: Flipped classroom study—quantitative test scores + qualitative interviews on learning experiences; joint display connects performance differences to engagement themes
Health Sciences
Typical Question: “What is intervention effectiveness AND how do patients experience it?”
Common Design: Embedded RCT with qualitative process evaluation, explanatory, convergent
Example: Diabetes self-management—HbA1c levels + patient interviews; integration contextualizes clinical outcomes with lived experience
Social Sciences
Typical Question: “How do social policies affect behaviors AND how do people make decisions?”
Designs: All designs used; exploratory sequential for understudied populations; convergent for comprehensive community studies
Example: Unemployment impacts—survey statistics + life history interviews; integration explains statistical patterns through narratives
Business/Management
Typical Question: “What are market trends AND how do customers feel about them?”
Common Design: Explanatory sequential, convergent, multiphase
Example: New product adoption—sales figures + focus groups; integration identifies barriers to adoption
Conclusion: Moving Forward with Your Mixed Methods Research
Mixed methods research represents a powerful, sophisticated approach to complex research questions that cannot be adequately addressed by qualitative or quantitative methods alone. Success requires:
- Intentional Design: Don’t combine methods haphazardly; select a design that matches your research question using decision frameworks
- Genuine Integration: Plan, execute, and report integration explicitly—this is the defining feature
- Rigor in Both Strands: Each method must meet disciplinary quality standards independently
- Pragmatic Mindset: Focus on “what works” to answer your question, not philosophical purity
- Resources and Time: Plan realistically—mixed methods is demanding but rewarding
- Use Authoritative Guidance: Follow APA JARS-Mixed, use joint displays, cite key authors (Creswell & Plano Clark, Tashakkori & Teddlie)
Your Next Steps
- Refine your research question: Ensure it genuinely requires both methods
- Consult with your advisor/committee: Get feedback on design appropriateness
- Study the JARS-Mixed standards: Download the checklist—this will be your roadmap for writing
- Develop a detailed protocol: Use the template in Section 13.2 before IRB submission
- Plan integration explicitly: Complete the integration planning worksheet before data collection
- Build a support team: If possible, collaborate with someone strong in the method where you’re weaker
- Start with simpler designs: If you’re new to MMR, consider explanatory sequential or embedded rather than convergent
Recommended Resources:
- Essential Reading: Creswell & Plano Clark (2017). Designing and Conducting Mixed Methods Research (3rd ed.)
- APA Standards: apastyle.apa.org/jars/mixed-methods
- Professional Community: Society for Mixed Methods Research (mixedmethods.org)
- Software: NVivo, ATLAS.ti, Dedoose (mixed methods integration capabilities)
Related Guides
Need help with foundational research skills? Check these related resources:
Get Expert Help with Your Mixed Methods Research
Designing and executing a rigorous mixed methods study is complex. Our team includes methodologists with PhDs in research methodology who can help you:
- Design consultation: Choose the appropriate mixed methods design for your research question
- Protocol development: Create a comprehensive mixed methods research protocol meeting APA standards
- Integration planning: Develop joint displays and meta-inference strategies
- Quality review: Audit your methodology section for completeness and rigor
- IRB support: Ensure ethical approval for both strands
- Complete assistance: Full-service support from design through final write-up
Get expert help with your mixed methods research project. We’ll match you with a methodologist experienced in your discipline.
Summary and Next Steps
Mixed methods research offers the most comprehensive approach to complex research questions by combining the breadth of quantitative data with the depth of qualitative insights. However, success requires:
- Deliberate design using established typologies (convergent, explanatory, exploratory, embedded)
- Genuine integration through joint displays and meta-inferences (not parallel reporting)
- Rigor in both strands meeting respective quality standards
- Adherence to reporting standards (APA JARS-Mixed essential)
- Realistic planning for time and resource requirements
The most common student error is superficial integration—collecting both data types but failing to genuinely connect them. Avoid this by planning integration deliberately before data collection, using joint displays systematically, and addressing integration throughout the manuscript (methods, results, discussion).
Your immediate next steps:
- Read this guide thoroughly, bookmark key sections
- Download the planning template and begin filling it out
- Study the APA JARS-Mixed standards—they’re non-negotiable for publication
- Consult with a mixed methods expert before finalizing your design
- Read 2-3 exemplar mixed methods articles in your target journal to understand expected structure and integration
Mixed methods is a powerful methodology that, when done correctly, produces research with rich insights and practical impact. While it demands more upfront planning and effort than single-method studies, the payoff in comprehensive understanding is substantial. Good luck with your research!
