Skip to main content
Fig. 1 | BMC Public Health

Fig. 1

From: Using prediction polling to harness collective intelligence for disease forecasting

Fig. 1

Main features of the forecasting user interface. Participants were asked to provide, for each question, a probability for each possible answer. The forecasting widget forced all probabilities to add up to 100%. After they made their first forecast in a question, participants were shown a current (but sub-optimal) aggregation of the collective (crowd) forecast for comparison, and could update their own at will. If they chose to, participants could also share their forecasts and rationales in the discussion forum

Back to article page