Designing rigorous mixed-methods studies

Mixed-methods research is having a moment — and for good reason. Done well, it lets us answer questions neither tradition could tackle alone. Done poorly, it produces studies that satisfy neither audience.

Why mixed methods?#

The case for mixing isn't about more data — it's about answering the right question with the right tool. Most research questions decompose into three layers: scope (what is happening), mechanism (how it happens), and meaning (why it matters).

A four-step framework#

1. Frame the research question — scope, mechanism, and meaning.
2. Choose the integration design — convergent, sequential, or embedded.
3. Plan the sampling strategy — separate samples or nested cases.
4. Design the integration analysis — joint displays and meta-inferences.

Each step prevents a class of failure. Skip framing and the methods don't connect. Skip integration design and the results stay parallel rather than combined.