Data Analytics Networking Night with Andrew Gelman & Shira Mitchell

Principal Analytics Prep hosts a networking night for people excited about data. We're honored to have two amazing speakers, Professor Andrew Gelman (Columbia) and Dr. Shira Mitchell (Mathematica Policy Research). The talks cover two essential topics in real-world analytics: story-telling with data (Andrew); and causal inference without true experiments (Shira).

Come hear these great talks, and meet other data scientists and analytics people. Food and refreshments will be served.

--------------------------------------------------------------------------------------------------

Andrew Gelman is Higgins Professor of Statistics, Professor of Political Science and director of the Applied Statistics Center at Columbia University. His books include Bayesian Data Analysis, Teaching Statistics: A Bag of Tricks, Data Analysis Using Regression and Multilevel/ Hierarchical Models, Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do, and A Quantitative Tour of the Social Sciences. Among his numerous publications is an article in Slate critical of the "power pose" research, co-authored with Principal Analytics Prep founder, Kaiser Fung.

Shira Mitchell is a statistician at Mathematica Policy Research, where she works on surveys, small area estimation, and causal inference for health and labor policy. Shira received a PhD in biostatistics and a BA in mathematics from Harvard University. Her dissertation was a collaboration with the Human Rights Data Analysis Group, using hierarchical models to estimate the numbers of casualties in Colombia's armed conflict. She did her postdoctoral fellowship at Columbia University, working with Andrew Gelman and Jeffrey Sachs.

-------------------------------------------------------------------------------------------------

Andrew's Talk: "Statistics: Learning From Stories"

Here is a paradox: In statistics we aim for representative samples and balanced comparisons, but stories are interesting to the extent that they are surprising and atypical. The resolution of the paradox is that stories can be seen as a form of model checking: we learn from a good story when it refutes some idea we have about the world. We demonstrate with several examples of successes and failures of applied statistics.

Shira's Talk: "The Millennium Villages Project"

The Millennium Villages Project (MVP) was a 10-year project of the Earth Institute at Columbia University, the United Nations Development Programme, and Millennium Promise, implemented in 10 sub-Saharan African sites, aimed at achieving the Millennium Development Goals—eight globally endorsed targets that address the problems of poverty, health, gender equality, and disease. We evaluated MVP's effect on development indicators. The greatest challenges for causal inference include: a nonrandomized design, limited baseline data for candidate control areas, and the assignment of treatment to only ten sites, limiting effective sample sizes. We fit a hierarchical Bayesian model, in Stan, that partially pools across multiple sites and multiple outcomes to ameliorate the problem of "multiple comparisons", and compare to results from a classical analysis.