Research design: choose the method that actually answers the question

Design follows from the question, not from habit. The most important choice in the project is about credibility: why should peers trust your findings?

Let the question decide the design

If the question is about meaning, process and experience, it points to qualitative designs. If it's about magnitude, association or cause, it points to quantitative ones. Complex questions often call for mixed designs. Choose what answers the question best — not what you're most comfortable with.

Justify robustness and validity

A research design must address how you secure internal and external validity, reliability and — in qualitative research — trustworthiness and transferability. Peers want to see that you know the threats to validity and have designed against them. Being open about the weaknesses strengthens, not weakens, the project.

Write the design so another researcher could repeat the study. Reproducibility is a precondition for your findings to count in the scientific conversation.

Plan for what goes wrong

Real projects meet attrition, missing data and skewed samples. A mature design has contingencies: how do you handle drop-out, non-response or data that doesn't behave as expected? Thinking this through in advance separates a feasible project from a fragile one.

Common pitfalls

• Choosing a method from habit rather than from the question.
• Ignoring the threats to validity.
• Having no contingency for attrition and missing data.

Ready to move on when…

• The design is derived from the research question.
• Validity and reliability are addressed.
• The study is described so it can be reproduced.