Evaluation of a Performance-Based Expert Elicitation: WHO Global Attribution of Foodborne Diseases

摘要: For many societally important science-based decisions, data are inadequate, unreliable or non-existent, and expert advice is sought. In such cases, procedures for eliciting structured judgments (SEJ) increasingly used. This raises questions regarding validity reproducibility. paper presents new findings from a large-scale international SEJ study intended to estimate the global burden of foodborne disease on behalf WHO. The involved 72 experts distributed over 134 panels, with panels comprising thirteen average. Elicitations were conducted in five languages. Performance-based weighted solutions target interest formed each panel. These weights based individual expert’s statistical accuracy informativeness, determined using between ten fifteen calibration variables experts' field known values. Equal combinations also calculated. main conclusions performance are: (1) does provide method attribution diseases; (2) equal weighting per panel increased acceptable levels, but at cost informativeness; (3) performance-based while retaining accuracy; (4) due constraints experts’ accuracies generally lower than other studies, (5) there was negative correlation informativeness which attenuated as improved, revealing that least accurate drive correlation. It shown, however, has ability yield statistically informative judgments, thereby offsetting this contrary influence. present suggest application large scale feasible, motivate development enhanced training tools remote elicitation multiple, internationally-dispersed panels.