Robust weighted aggregation of expert opinions in futures studies

Abstract Expert judgments are widespread in many fields, and the way which they collected procedure by aggregated considered crucial steps. From a statistical perspective, expert subjective data must be gathered treated as carefully scientifically possible. In elicitation phase, multitude of experts is preferable to single expert, techniques based on anonymity iterations, such Delphi, offer advantages terms reducing distortions, mainly related cognitive biases. There two approaches aggregation given panel experts, referred behavioural (implying an interaction between experts) mathematical (involving non-interacting participants using formula). Both have disadvantages, with approach, main problem concerns choice appropriate formula for both normalization aggregation. We propose new method aggregating processing Delphi method, aim obtaining robust rankings outputs. This makes it possible normalize aggregate opinions while modelling different sources uncertainty. use uncertainty analysis approach that allows contemporaneous functions, so result does not depend specific formula, thereby solving choice. Furthermore, we can also model weighting system, reflects expertise well opinion accuracy. By combining ranking procedure, protocol covering elicitation, used construction Delphi-based future scenarios. The very flexible applied any judgments, i.e. those outside context futures studies. Finally, show validity, reproducibility potential through its application regard Italian families.