I am planning to implement the Delphi method as part of my project and a few queries came to mind:
1. does the Delphi method have to be iterative including at least 2 rounds or would just one round be enough?
2. what is the best way to gather data from the experts panel? Are semi-structured interviews a recommended way?
3. should the members of the targeted panel meet up together at any stage of the Delphi method (e.g. summarising and discussing their responses) or can the lead researcher analyse the data straight after their collection in absence of any iterative process (e.g. use of thematic analysis of interview's responses).
Hi Kostas,
There is no 'optimal' form of Delphi. It depends on what you want to achieve and what occurs - but there is a 'traditional and classical' form of Delphi.
1. The Delphi process is iterative - but, as with your third question, that doesn't mean that the panel need to meet up. You can never have one round in a Delphi. The first round must be quantitative from where your second round quantitative statements emerge from - or do you mean 'how many rounds after the first qualitative round'? That depends on the levels of consensus that you have set and want to to achieve.
2. The first qualitative round is usually semi-structured. the following quantitative rounds are usually Likert-scale related and quantitative.
3. No - the panel do not meet up. The 'consensus' is anonymous. A form of thematic analysis occurs in the first round in order to form the later round statements.
The following chapter may assist - as it is a form of mixed methods research. Obviously - just the sections on Delphi.
No problem Kostas - happy to assist where I can!! I've just noticed an error from my last response though. I think that it might be the RG spell-checker kicking in. In response 1 - it should state that the 'first round must be qualitative'.
Best,
Dean
Dear all,
I have nothing to add to what Dean already very judiciously said.
- Badges
- Science topic
- Similar topics
- Statistics
- Survey