A general methodology question about reaching consensus in the Delphi method:
When we have a Likert scale questionnaire for our experts to fill, the consensus criteria is items that have a mode above 5, and we have 11 items in total. In the first round, many items have reached a mode above 5. Do we exclude them for the next round and only ask about the other items which haven't reached the consensus criteria? Or can we do the opposite and exclude the items which haven't reached the criteria and include the items that have reached the criteria to narrow them down, as the goal is to limit the number of items to 4-5?
Thank you in advance