Join us for thought-provoking discussion with Prof. Drew Dimmery, PhD (Hertie School) on design-based problem solving.
The talk focuses on how to design studies that preemptively address challenges, improving inference and generalization before data collection even begins.
Abstract from the speaker:
How can we design studies to proactively resolve problems before data is even collected? I will discuss three such cases:
First, I'll briefly examine how best to assign treatment when the goal is transporting causal effects to a known population. It turns out that improved generalization just requires a simple modification to the typical rerandomization objective. Paper: [https://arxiv.org/abs/2009.03860](https://arxiv.org/abs/2009.03860 "https://arxiv.org/abs/2009.03860")
Second, I'll look at how a design aimed at improved estimation of heterogeneous treatment effects works. This delves a bit into algorithmic CS theory, showing a connection between designs for good HTE estimation and a ubiquitous graph cutting problem, MAXCUT. Paper: [https://arxiv.org/abs/2010.11332](https://arxiv.org/abs/2010.11332 "https://arxiv.org/abs/2010.11332")
The bulk of the talk will be around a procedure for online (i.e. sequential) assignment of units to treatment, such as in a survey experiment. Specifically, we introduce extensions of the state-of-the-art for online discrepancy minimization from the CS literature, as well as introduce procedures to accommodate multiple treatments and non-uniform treatment probabilities. This procedure works quite well, has provable robustness properties and is even sometimes competitive with _offline_ allocation procedures which are far slower than the proposed algorithm. Paper: [https://arxiv.org/abs/2203.02025](https://arxiv.org/abs/2203.02025)
About the speaker:
Drew Dimmery received his PhD in Politics from New York University in 2016 with a dissertation on methods in causal inference and machine learning, after which he worked for four years on the Adaptive Experimentation team in Facebook’s Core Data Science, developing and implementing methods and tools to improve experimentation at scale. Between 2021 and 2023, he was the Scientific Coordinator at the University of Vienna’s Data Science Research Network, supporting third mission activities of the University alongside his research. He has published in top machine learning journals such as ICML, KDD and AISTATS, as well as general-interest journals such as Science and Nature. Methodologically, his focus of study is on methods for experimentation and causal machine learning. Substantively, he focuses largely on the internet and social media.
Bring your own lunch bag! Light pastries and drinks will be available in case you forget to bring it.
The Data Science Brown Bag Series is an informal and interactive gathering where participants bring their own brown bag lunch and engage in discussions on research and insights the field of data and computational social science (light pastries and drinks will be available if you forget your lunch bag!).
The series provides a platform for data enthusiasts, researchers, and practitioners to share their experiences, best practices, and emerging methodologies and research in using data science to analyze and understand social and political phenomena. The brown bag talk series is for anyone interested in data science and social science to network, learn, and share ideas in a casual and friendly setting.
