Research Methodology Best Practices for Postgraduate Studies

Most dissertation problems that surface in month four were actually created in week one, at the methodology stage. A well-designed methodology doesn't guarantee an easy dissertation, but a poorly designed one guarantees a difficult one, regardless of how hard you work afterward. Methodology is where research succeeds or fails, long before the first data point is collected.

Start from the research question, not the method you already know

A common mistake is choosing a method because it's familiar (usually a survey, because it's fast to design) and then retrofitting a research question to fit it. Strong methodology chapters do the reverse: the question dictates the method. If your question asks "how much" or "how many," you likely need a quantitative design. If it asks "why" or "how people experience," you likely need a qualitative one. If it asks both, you need a genuinely integrated mixed-methods design, not two separate mini-studies bolted together.

The four decisions that anchor a defensible methodology

• Philosophical position: your assumptions about the nature of the knowledge you're generating (positivist, interpretivist, pragmatist), stated explicitly rather than implied.
• Research design: the overall strategy (case study, survey, experimental, ethnographic, grounded theory) that logically follows from your philosophical position and question.
• Sampling strategy: who or what you're studying, how you selected them, and, critically, why that selection is defensible for your specific question, not just convenient.
• Data collection and analysis method: the specific instruments and analytical techniques, chosen because they answer your question, not because a template exists online.

Examiners rarely fail a dissertation because the method was quantitative or qualitative. They fail dissertations where the method doesn't logically follow from the question.

The methodology chapter examiners actually respect

A strong methodology chapter doesn't just describe what you did, it justifies why alternatives were rejected. If you chose semi-structured interviews, briefly explain why a survey wouldn't have answered your question as well. This comparative justification, done in two or three sentences per decision, is what separates a methodology chapter that reads as considered from one that reads as default.

Common weaknesses we see repeatedly

Sample sizes justified by convenience rather than by power calculation or theoretical saturation; validity and reliability (or trustworthiness, for qualitative work) mentioned in a single throwaway paragraph instead of woven through the design; and ethics treated as a compliance form to complete rather than a design consideration that shapes recruitment and consent from the start.

The takeaway

Let the question choose the method, state your philosophical position explicitly, justify every design decision against the alternatives you rejected, and treat ethics as part of the design rather than paperwork. Get the methodology right and the rest of the dissertation has somewhere solid to stand.