Research methodology · Lesson 3 of 3

Mixed methods, and the approaches outside the families

Combine approaches when one has raised a question the other can answer — and be able to say which shape you are using.

8 min read · Free — no account needed to read

Mixed methods combine both deliberately, and the word that matters is deliberately: doing a survey and some interviews because you could is not mixed methods, it is two studies in one document.

The common shapes: sequential exploratory — qualitative first to work out what the questions should be, then quantitative to measure at scale. Sequential explanatory — quantitative first to find the pattern, then qualitative to understand why it exists. Convergent — both at once, compared at the end, which is strongest when they disagree, because the disagreement is a finding. Embedded — one plays a supporting role inside a study led by the other.

The reason to combine is that one approach answers a question the other raised. The cost is real: you need competence in both, and it takes longer. "For rigour" is not a reason.

Outside the two families sit approaches with their own logic. Action research and participatory research involve the people being studied as partners and aim at change, not only understanding. Historical and archival research works with records made for other purposes. Comparative research treats cases — countries, organisations, periods — as the unit. Design science builds an artefact and evaluates it. Computational and simulation research models a system to explore behaviour that cannot be observed directly. Policy, market and user research are applied traditions with their own conventions and are recognisably the same activity.

See it happen

A survey shows that one department reports much lower satisfaction than the rest of the organisation.

The read that arrives on its own

The survey found the problem department — now act on the scores.

The same situation, with this idea

You recognise a sequential explanatory shape. The survey tells you where and how much; it cannot tell you what is happening, because it only offered the options somebody wrote in advance. A handful of conversations in that department is the second half of the design, and it is what turns a number into something you can act on.

After this, you can

see a result that raises a question your method cannot answer, and know which approach picks it up.

The takeaway

Combine approaches when one has raised a question the other can answer — and be able to say which shape you are using.

Easy to take the wrong way

That mixed methods is automatically stronger. It is stronger when the combination is reasoned and weaker when it splits limited effort across two half-done studies.

Try this

Look at your question and ask what a first approach would leave unanswered. Name the second approach that would pick it up, and the order.