A mixed methods research design sequences the quantitative and qualitative strands. The three core designs drawn, how each works, and how to choose one.
A mixed methods design names how the quantitative and qualitative strands are sequenced — together, or one after the other. Pick it after collection begins and you are not running a design; you are labeling whatever turned up. For the researchers and evaluators choosing how to structure a mixed methods study.
By Unmesh Sheth · Founder & CEO, Sopact · Updated May 2026
A mixed methods research design is the plan for how a study's quantitative and qualitative strands are sequenced and combined. It names whether the two strands run at the same time or one after the other, which strand leads, and how their findings are brought together. The design is chosen before data collection begins.
There are three core designs — convergent parallel, explanatory sequential, and exploratory sequential — plus named variants like embedded and multiphase. The choice is set by one thing: the research question, and which strand has to come first to answer it.
In the old model, the mixed methods design carried a heavy burden: it was the only thing standing between an integrated study and two parallel reports. Choose wrong, or choose late, and the strands never met. The redefinition lightens it. When every input lands on one record and is read together on arrival, integration is continuous — a property of the data, not of the design. The design choice then becomes a smaller, cleaner question: which strand do you collect first, and why.
Pick the design for the right reason — the sequence your question needs — not as the mechanism that forces integration. Integration is handled upstream, on the record. The full case is on the pillar: mixed methods research, redefined.
Each design is a different sequence of the same two strands. The diagram is the design: which strand runs first, and what the second one does with the result.
Both strands run at the same time and independently. The analysis compares the rating and the narrative for each respondent to see whether they agree.
The quantitative survey runs first. A qualitative follow-up is then designed around the patterns and outliers the numbers revealed.
Qualitative work runs first to surface themes and language. A quantitative survey is then built from those themes and tests them at scale.
One strand leads the study and the other is embedded inside it in a supporting role — for example, qualitative interviews nested within a larger trial.
Multiphase and nested designs are named extensions of these four — more phases, or one design folded inside another — not separate families. Pick a core design first, then extend it if the study needs more phases.
The design is set by what the research question needs first. Five common question shapes, and the design each one points to.
| If your research question is... | The design to choose |
|---|---|
| "Do the numbers and the narratives agree?" | Convergent parallel — run both strands at once and compare them. |
| "The result is clear, but why did it happen?" | Explanatory sequential — quant first, then qual explains it. |
| "Something is happening, but I do not know what to measure." | Exploratory sequential — qual first, then quant tests it. |
| "One strand answers the main question; the other is secondary." | Embedded — one strand leads, the other is nested inside. |
| "The study runs in several phases over a long horizon." | Multiphase — a core design, extended across more phases. |
Ask which strand has to come first to answer your question. If neither does, it is convergent. If the quant leads, it is explanatory. If the qual leads, it is exploratory. The rest is sample size and timing.
Every design above assumes one thing the diagram does not show: that the quantitative and qualitative strands land on the same record, under one persistent ID. A convergent design with two unconnected samples is not convergent; it is two studies. The sequence is the researcher's choice. Keeping the strands connected is the part the data model has to carry.
Each participant carries one persistent ID across the quantitative survey, the qualitative interview, the documents, and the transcripts — whatever order the design collects them in. The convergent comparison, the sequential follow-up, the embedded strand: each one runs on a record that was already connected at collection, not matched together afterward.
Bring your research question. We will walk through which design fits, what sequence it sets, and how the strands stay connected whichever one you pick.
A mixed methods research design is the plan for how a study's quantitative and qualitative strands are sequenced and combined. It names whether the two strands run at the same time or one after the other, which strand leads, and how their findings are brought together. The design is chosen before data collection begins.
There are three core designs. Convergent parallel runs both strands at the same time and compares them. Explanatory sequential runs the quantitative strand first, then qualitative explains the result. Exploratory sequential runs the qualitative strand first, then quantitative tests it. Named variants — embedded, multiphase, and nested — are extensions of these three.
Convergent parallel mixed methods design collects the quantitative and qualitative data at the same time and analyzes each strand independently, then compares the two sets of findings. It fits a research question about whether the numbers and the narratives tell the same story. It is also called a convergent or parallel design.
Explanatory sequential mixed methods design runs the quantitative survey first, then a qualitative follow-up explains the patterns and outliers the numbers revealed. The quantitative findings shape who is interviewed and what is asked. It fits a research question where the measured result is clear but the reason behind it is not.
Exploratory sequential mixed methods design runs the qualitative phase first to surface themes and language, then a quantitative phase tests those themes at scale. The qualitative phase informs the survey instrument. It fits a research question where you do not yet know what to measure, only that something needs naming first.
Embedded mixed methods design has one strand lead the study while the other is nested inside it in a supporting role — for example, qualitative interviews embedded within a larger quantitative trial. The embedded strand answers a secondary question that supports, but does not drive, the main design.
Ask which strand has to come first to answer your research question. If neither does and you want to compare them, choose convergent parallel. If the quantitative result needs explaining, choose explanatory sequential. If you need qualitative work to define what to measure, choose exploratory sequential. The choice is made before collection begins.
Explanatory sequential runs quantitative first, then qualitative explains the numbers. Exploratory sequential runs qualitative first, then quantitative tests the themes. The order is the difference: explanatory starts with a result that needs a reason; exploratory starts with a question that needs to be made measurable.
In a convergent design the two strands run at the same time and are compared. In a sequential design one strand runs first and shapes the second. Convergent answers whether the strands agree; sequential uses one strand to set up or explain the other.
Construct it in four steps: write a three-part research question including an integration question, pick the core design family that matches it, plan the sample relationship between the strands, and decide how the two strands connect at the respondent level. All four are settled before the first response is collected.
A multiphase mixed methods design runs several connected studies in sequence over a longer program, each phase building on the last. It is common in program development and evaluation, where an exploratory phase, a pilot, and a full study each inform the next. It is an extension of the three core designs across more phases.
Yes. A longitudinal mixed methods design carries an integrated quantitative and qualitative design across multiple waves, following the same units over time. It pairs the depth of mixed methods with the change-measurement of a longitudinal design, and is well suited to tracking how an outcome and its causes move together.
This page covers the design types. The pillar holds the methodology and the redefinition; the four other guides cover the analysis, the worked examples, the survey instrument, and the tools.
A working session, not a demo. Bring the research question you need to answer. We walk through which mixed methods design fits, the sequence it sets, the sample relationship between the strands, and how they stay connected on one record whichever design you choose. You leave with a design chosen on purpose and a collection plan to match.
Live walkthrough · 30 min · with Unmesh Sheth, Founder & CEO · bring a research question or a study you are designing
