BEGIN:VCALENDAR VERSION:2.0 PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4// BEGIN:VEVENT UID:20260427T025710EDT-3170ounf3O@132.216.98.100 DTSTAMP:20260427T065710Z DESCRIPTION:Abstract: Forecasts about research findings affect critical sci entific decisions\, such as what treatments an R&D lab invests in\, or whi ch papers a researcher decides to write. But what do we know about the acc uracy of these forecasts? We analyze a unique data set of all 100 projects posted on the Social Science Prediction Platform from 2020 to 2024\, whic h received 53\,298 forecasts in total\, including 66 projects for which we also have results. We show that forecasters\, on average\, over-estimate treatment effects\; however\, the average forecast is quite predictive of the actual treatment effect. We also examine differences in accuracy acros s forecasters. Academics have a slightly higher accuracy than non-academic s\, but expertise in a field does not increase accuracy. A panel of motiva ted repeat forecasters has higher accuracy\, but this does not extend more broadly to all repeat forecasters. Confidence in the accuracy of one's fo recasts is perversely associated with lower accuracy. We also document sub stantial cross-study correlation in accuracy among forecasters and identif y a group of 'superforecasters'. Finally\, we relate our findings to resul ts in the literature as well as to expert forecasts.\n\n \n\nEva Vivalt is an Assistant Professor in the Department of Economics at the University o f Toronto. Her research explores questions related to development and labo ur\, and has studied cash transfers in the U.S.\, the stumbling blocks to evidence-based policy decisions\, including methodological issues\, how ev idence is interpreted\, and the use of forecasting.\n\n \n DTSTART:20260402T163000Z DTEND:20260402T180000Z SUMMARY:Forecasting Social Science: Evidence from 100 Projects URL:/isid/channels/event/forecasting-social-science-ev idence-100-projects-372071 END:VEVENT END:VCALENDAR