Design-based Mixed Methods Matrix — Mixed Methods Design Classification Framework
Design-based Mixed Methods Matrix Framework · Also known as: mixed methods design matrix, MM design typology matrix, mixed methods design framework, DBMM matrix
The design-based mixed methods matrix is a systematic framework for selecting and structuring mixed methods research designs. It organises key design decisions — purpose, timing of data strands, point of integration, and weighting of quantitative versus qualitative components — into a coherent matrix that guides researchers toward a defensible, transparent design. The framework draws on the typology traditions of Creswell and Plano Clark and Greene's purposes-based approach.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use the design-based mixed methods matrix when your research question genuinely requires both quantitative and qualitative evidence and you need a principled way to choose among competing design options. It is especially valuable for doctoral candidates and applied researchers who are new to mixed methods and need a structured decision aid. The matrix is appropriate at the proposal stage, before any data are collected. Do not use it when a mono-method design is sufficient — if your question is purely explanatory (a survey will do) or purely exploratory (interviews will do), mixing adds complexity without benefit. Also avoid forcing a mixed design when resources allow only one strand to be executed rigorously; a well-executed mono-method study is preferable to a poorly integrated mixed design.
Strengths & limitations
- Provides a structured, transparent decision procedure for a notoriously complex design choice.
- Grounded in widely adopted typologies (Creswell & Plano Clark; Greene), making design rationale communicable to reviewers and committees.
- Forces explicit decisions about integration point and weighting before data collection, reducing post-hoc rationalisation.
- Applicable across disciplines — education, health sciences, social sciences, policy research — wherever mixed evidence is needed.
- Reduces the risk of mixing for the sake of mixing by anchoring design choice to a named purpose.
- The canonical typology covers four major designs; hybrid or emergent designs may not fit neatly into any cell, requiring adaptation.
- The matrix does not resolve the deeper paradigmatic tension between post-positivist and constructivist assumptions — it is a design tool, not a philosophical reconciliation.
- Sequential designs add substantial time and cost; the matrix makes the trade-off visible but cannot eliminate it.
- Novice researchers may treat the matrix as a mechanical algorithm and underestimate the iterative, judgement-intensive nature of integration.
Frequently asked
Is the design-based mixed methods matrix a method or a framework?
It is a design framework — a decision aid for selecting and structuring a mixed methods study. It does not prescribe how to collect or analyse data within each strand; it governs the overall architecture of the study, specifying timing, weighting, and integration strategy across quantitative and qualitative components.
Do I have to follow one of the four canonical designs, or can I create a custom design?
The four canonical designs (convergent parallel, sequential explanatory, sequential exploratory, embedded) cover the most common configurations, but mixed methods scholars acknowledge that complex studies sometimes require hybrid or transformative designs. If your study departs from the canonical cells, you should name and justify the departure explicitly rather than leaving the design unlabelled.
What does weighting mean in the matrix, and how do I decide?
Weighting refers to the relative emphasis given to the quantitative and qualitative strands in terms of data volume, analytical depth, and interpretive authority. The decision follows from the research purpose: if you are primarily testing a hypothesis and qualitative data serve an explanatory supplement, the quantitative strand is dominant (QUAN + qual). If you are building theory and quantitative data serve a scaling function, the qualitative strand is dominant (QUAL + quan). Equal weighting is also a legitimate choice in genuinely convergent designs.
Can I use this framework with a single dataset, or does mixed methods always require two separate data collection phases?
Some mixed methods designs — particularly convergent parallel designs and certain embedded designs — collect both strands concurrently, but they remain two distinct strands (e.g., a survey with closed items analysed quantitatively and open-ended items analysed qualitatively). A single questionnaire can generate both strands. Sequential designs, by definition, require separate collection phases. The matrix helps clarify which configuration matches your question and timeline.
How does the matrix relate to the paradigm wars in research methodology?
The matrix operates at the design level, not the paradigm level. Researchers adopting a pragmatist stance typically find the matrix most comfortable, as pragmatism holds that the research question — not philosophical allegiance — should drive method choice. Researchers committed to a strong post-positivist or constructivist paradigm may find that certain matrix cells sit uneasily with their epistemological assumptions; in those cases the matrix should be used critically rather than mechanically.
Sources
- Creswell, J. W., & Plano Clark, V. L. (2011). Designing and Conducting Mixed Methods Research (2nd ed.). Sage. ISBN: 978-1412975179
- Greene, J. C. (2007). Mixed Methods in Social Inquiry. Jossey-Bass. ISBN: 978-0787983826
How to cite this page
ScholarGate. (2026, June 3). Design-based Mixed Methods Matrix Framework. ScholarGate. https://scholargate.app/en/research-design/design-based-mixed-methods-matrix
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Design-based ResearchField Methods↔ compare