Mixed Methods Research Guide
Combine quantitative and qualitative approaches for comprehensive research. Master design selection, integration strategies, and mixed methods legitimation.
Mixed Methods Design Types
Choose the design that best matches your research questions, timeline, and expertise. Capital letters indicate priority (QUAN vs qual).
Convergent Design
QUAN + QUAL → Merge
Collect quantitative and qualitative data simultaneously, analyze separately, then merge findings.
When to use:
When you want to compare or corroborate findings from different methods
Example:
Survey measuring teacher burnout + interviews exploring burnout experiences
Explanatory Sequential
QUAN → qual
Start with quantitative data collection and analysis, then use qualitative to explain or expand.
When to use:
When quantitative results need deeper explanation or unexpected findings emerge
Example:
Survey identifies factors → Interviews explore why those factors matter
Exploratory Sequential
QUAL → quan
Start with qualitative exploration, then use findings to develop quantitative instrument.
When to use:
When developing new instruments or testing emergent theories
Example:
Interviews identify themes → Survey tests themes across larger population
Embedded Design
QUAN(qual) or QUAL(quan)
One method is embedded within a larger study using a different method.
When to use:
When one method plays a supportive role to the primary approach
Example:
RCT (quantitative) with qualitative process evaluation
Transformative Design
Framework-driven
A theoretical framework (feminist, critical race, disability) guides all design decisions.
When to use:
When research aims to address power imbalances or advocate for change
Example:
Participatory action research combining surveys and focus groups
Multiphase Design
Multiple phases
Multiple sequential phases, each building on previous findings over extended time.
When to use:
For large-scale, multi-year research programs
Example:
Longitudinal study with annual surveys and periodic interviews
Integration Strategies
Integration is what makes mixed methods more than just "doing both." Choose strategies that meaningfully combine your findings.
Merging
Combine quantitative and qualitative databases through side-by-side comparison
Techniques:
- Joint displays
- Side-by-side tables
- Data transformation matrices
Best for: Convergent designs
Building
Results from one phase inform data collection in the next
Techniques:
- Qualitative themes → survey items
- Statistical results → interview protocols
Best for: Sequential designs
Connecting
Link datasets through participant selection or follow-up
Techniques:
- Purposeful sampling from survey
- Case selection for follow-up interviews
Best for: Explanatory sequential
Embedding
Nest one dataset within analysis of the other
Techniques:
- Qualitative themes within experimental conditions
- Quantitative data within case studies
Best for: Embedded designs
Creating Joint Displays
Joint displays are tables or figures that visually integrate quantitative and qualitative findings.
Example: Teacher Burnout Study
| Quantitative Finding | Qualitative Theme | Meta-Inference |
|---|---|---|
| High workload correlated with burnout (r=.67) | "I never have time to breathe" - administrative burden | Workload affects burnout through perceived lack of control |
| Support predicted lower burnout (β=-.45) | "My colleagues keep me sane" - informal support networks | Peer support more protective than formal support structures |
| Experience not significantly related | "Experience helps you cope differently, not less" | Experience changes coping strategies, not burnout levels |
The meta-inference column shows how integration creates insights beyond what either method alone could reveal.
Common Mixed Methods Mistakes
Treating mixed methods as 'doing both' without integration
Plan integration from the start. Where will data connect? How will findings combine? Without integration, it's two studies, not mixed methods.
Choosing design based on preference, not research questions
Let your questions drive design. Sequential designs answer different questions than convergent designs.
Inadequate sample for one strand
Each strand needs appropriate sampling. Don't sacrifice qualitative depth for quantitative breadth or vice versa.
Privileging one method over the other
Even in priority designs (QUAN→qual), both strands contribute. Report and value findings from each.
Superficial integration in discussion only
Use joint displays, matrices, and visual representations. Integration should be systematic, not an afterthought.
Frequently Asked Questions
Need Expert Help with Mixed Methods?
Our methodology specialists can help you design your mixed methods study, plan integration strategies, and ensure rigorous execution.
Related Resources
Explore more resources in our Research Methodology collection