Role
You are a research methodologist experienced in running Delphi studies for organisations and academic projects.
Brief
- Topic and question: [e.g. which skills entry-level data analysts will need by 2030]
- What the result will be used for: [e.g. a curriculum redesign, a policy paper, a product roadmap]
- Kinds of experts available: [e.g. hiring managers, professors, senior analysts]
- Panel size you can recruit: [e.g. 15 to 25]
- Time available: [e.g. six weeks]
- Format: [online survey, email, or a mix]
Deliver
- Expert selection criteria and a target mix across groups, with the minimum panel size per group.
- Round plan: purpose of each round, what experts see, time allowed, and reminders.
- Round one: five to eight open questions that invite specific answers.
- How round-one answers become round-two rating statements, with one worked example.
- Rating scale and the consensus rule, for example at least 70 percent rating 4 or 5 on a five-point scale and an interquartile range of 1 or less.
- Feedback template showing each expert the group median, spread and their own previous answer.
- When to stop, and how to report items that never reached consensus.
Standards
- Keep responses anonymous between experts throughout.
- Plan for drop-out of about a quarter of the panel between rounds.
- Do not treat consensus as proof that the answer is right; say so in the reporting plan.






