In a study, I want to go for Fuzzy AHP technique where the objective is to find the priority (relative importance) of a set of attributes defining a system.What should be a reasonable sample size (number of expert) to conduct the survey?
It depends on the subject you work on it. Sometimes there are a few experts in an especial subject and 4 or 5 ideas can be enough. Sometimes you can gather more experts' ideas. Therefore it's logical to have more ideas. As you know you can give weights to different ideas too.
These articles are really helpful. Thanks for sharing. Could you tell me based on your experience that what should be the minimum size of the sample for FAHP?
You are right that the sample size depends on the size and complexity of the design. To be specific, my objective is to assess a set of criteria for defining 'pedestrian safety' in urban area and I want it to be evaluated by the subject expert. Say I have 6 attributes to define it. In that case what you think should be a reasonable sample size?
Before estimating a minimum sample size for FAHP it is necessary to delimit the subsidiary attributes and elements to attributes and subjects of such so that there is no duplication between them, that is complete the next formal. Thus one could initialize the simplest échoantillon size equal to the number of these attrubuts and then make adjustments for future evolution.