How to Write a Methodology Chapter for a Delphi Study

A weak methodology chapter is the single most common reason a Delphi-based thesis comes back with major revisions — not because the study was badly run, but because the chapter doesn't show a reviewer that it was. Examiners and journal reviewers can't see your actual research process; they can only judge the chapter you wrote about it. This guide walks through what a Delphi methodology chapter needs to contain, section by section, so the write-up matches the rigor of the study itself.

What a Delphi Methodology Chapter Actually Needs to Justify

Before describing what you did, a Delphi methodology chapter has to justify why Delphi was the right method at all — this is the section examiners read most critically, because it's where weak methodological reasoning shows first.

Structuring the Chapter Section by Section

Open with your research question, then justify the Delphi method directly against it. State whether you ran a classic multi-round Delphi or a real-time Delphi, and why — this single methodological choice affects everything that follows and reviewers expect it addressed explicitly, not assumed.

Describe your expert definition, recruitment method, and final panel size with justification — not just a number. State your sampling strategy (purposive sampling is standard for Delphi), inclusion and exclusion criteria, and how many panelists you approached versus how many participated.

Explain how Round 1 items were generated — literature-derived, drawn from prior qualitative work, or built from expert interviews — and how the questionnaire was piloted before going to the full panel. A questionnaire built without piloting is a common and easily avoidable weakness reviewers flag.

Walk through what happened round by round: how items were distributed, how long each round stayed open, how feedback was summarized and returned, and your stopping rule. State your consensus threshold and stability criterion here, defined in advance of data collection — not derived afterward to match your results.

Specify exactly how consensus and stability were calculated — percentage agreement, interquartile range, or another measure — and how qualitative comments were handled, if you collected any. This section should read as a plan that existed before data collection, not a post-hoc description of what you happened to compute.

Cover ethics approval, informed consent, anonymity guarantees for panelists, and data protection — GDPR compliance is worth a specific mention if any panelist is based in the EU, since some examiners will check for it directly.

Common Methodology Chapter Mistakes

Writing Style: What Makes a Methodology Chapter Read as Rigorous

The strongest Delphi methodology chapters read like a protocol someone else could follow, not a narrative recap of what happened. Use past tense consistently, state numbers precisely rather than approximately wherever you have them, and avoid evaluative language about your own results in this chapter — judgments about how well the study worked belong in the discussion, not embedded in the method description. A chapter written this way also tends to be shorter and easier to defend in a viva, because there's nothing in it left open to interpretation.

Citing Your Software and Tools Correctly

Whatever platform you use to run the rounds, name it explicitly in your procedure section along with how it shaped your design — for example, whether you ran a real-time Delphi where panelists could see aggregated feedback continuously, or a traditional multi-round version with discrete rounds and a waiting period between them. This single sentence pre-empts a question many examiners ask in a viva: how exactly were rounds administered, and could the process be replicated.

Methodology Chapter Checklist Before Submission

A methodology chapter that walks through design justification, panel selection, instrumentation, procedure, analysis plan, and ethics — in that order, with each choice tied back to your research question — gives reviewers nothing to second-guess on process, which leaves the conversation focused on your actual findings.