EXPERIMENT Distinguishing Between Correlation and Causality скачать в хорошем качестве

EXPERIMENT Distinguishing Between Correlation and Causality

Join us for a special CBE 50th Anniversary guest lecture with economist and experiment designer Dr. David Reiley, who will share insights from a career spent using field experiments to understand human behavior and measure causal effects. David will open the session with stories from his childhood that illustrate the difference between causality and correlation, adapted from his theater show EXPERIMENT. These anecdotes set the stage for a deeper exploration of how experimental thinking has shaped both his personal life and professional work. He will then walk through two major experiments he conducted in the tech industry. The first is his foundational advertising‑effectiveness experiment at Yahoo!, where he demonstrates how a naïve observational estimator performs poorly (three times the magnitude—and in the opposite direction) compared to a randomized experimental estimator. He also discusses when difference‑in‑differences methods can work well, and when they fail. Next, David will present his ad‑load experiment at Pandora, where he estimated a nonparametric demand curve for listening hours as a function of ads per hour. In this example, both cross‑sectional and panel observational estimators perform poorly, with the latter affected by a subtle reverse‑causality issue. This anniversary‑year talk will be an engaging and accessible look at the power of experimentation in real‑world settings—and why causal inference matters. Featuring: David Reiley, PhD David Reiley is Adjunct Professor of Information at the University of California at Berkeley, where he developed an online course on experiments and causal inference for the master's degree program in data science. He is a pioneer in the use of field experiments in economics. Economics has historically been an observational rather than an experimental science, but David has been a leader in the movement to introduce experiments wherever possible. He designed field experiments to test theories of auction bidding and charitable fundraising before he began to focus on experimental measurement of the effects of advertising on consumer behavior. Brought to you by: Department of Economics, College of Business and Economics, The Foundation for WWU & Alumni