How reliable are published claims in social & behavioral science? For example, does 1 minute of power posing really empower you? Does collective bargaining by teachers really reduce students’ academic achievement? To address the low replication rate of published papers in these fields, DARPA’s SCORE program rates the reliability of published claims. As one part of SCORE, we run surveys & prediction markets on at least 3,000 research claims from social & behavioral science.
We will briefly review previous results and present preliminary results from our work (over 1,400 claims forecast so far). For examples, our forecasters expect Economics > Psychology > Marketing, and we now have enough data to compare scores of different journals, based on a few dozen articles from each. We will also conduct live crowd forecasts on papers/claims of interest to the audience — please send suggestions to Deb Schenaker at email@example.com no later than COB Jan 28, 2020.
Dr. Kathryn Blackmond Laskey is Professor of Systems Engineering and Operations Research and Associate Director of the C4I and Cyber Center at George Mason University. She teaches and performs research on multisource information fusion, decision theoretic knowledge representation and reasoning method-ology, data analytics, and decision support. She has worked in diverse application areas, including modeling the emplacement of improvised explosive devices, detecting insider threats, predicting air-craft delays, managing terrorist risk at public facilities, and planning military engagements. She serves on the Board of the International Society of Information Fusion and the Washington Metropolitan Area chapter of INCOSE. She is past board chair of the Association for Uncertainty in Artificial Intelligence.
Dr. Charles Twardy worked with the C4I and Cyber Center from 2008-2015, when he led the George Mason DAGGRE and SciCast teams for the IARPA ACE and ForeST forecasting challenges (2011-2015). Their novel combinatorial prediction market showed sustained 35% gains over baseline methods, with much greater expressivity. He then worked for a year supporting the Defense Suicide Prevention Office on data-handling protocols and sound approaches to hot-spot detection. In late 2016 he joined Jacobs* where he worked on DARPA big data programs in counter-trafficking and cyber-security, supported another team for IARPA CREATE (crowdsourced argumentation), and some quick-turn machine learning. He now leads the Replication Markets team for DARPA SCORE. (*Actually, Sotera, which was acquired in 2017 by KeyW and in 2019 by Jacobs. It’s an exciting time!)
1:00 pm - 2:00 pm
C4I Center ENGR 4705