There are no hard and fast rules governing the use of expert judgements: how many experts to use, how they are found and chosen, or even more important, how differences among them are reconciled. The methods by which the opinions are gathered can run the gamut from having a colleague scribble some comments on a rough draft of the questionnaire to holding a conference of recognized leaders in the field, with explicit rules governing voting. Most approaches usually fall between these two extremes; somewhere in the neighborhood of three to ten people known to the scale developer as experts are consulted, usually individually. Since the objective is to generate as many potential items as possible for the scale, those suggested by even one person should be considered, at least in the first draft of the instrument.
The advantage of this approach is that if the experts are chosen carefully, they probably represent the most recent thinking in an area. Without much effort, the scale developer has access to the accumulated knowledge and experience of others who have worked in the field. The disadvantages may arise if the panel is skewed in some way, and does not reflect a range of opinions. Then, the final selection of items may represent one particular viewpoint, and there may be glaring gaps in the final product.