I am a post-doctoral researcher in the Institute of Data Science at the University of California, Berkeley School of Information. In 2016 I received my PhD in Political Science from the University of California, San Diego. I also studied at the University of Michigan where I graduated with a B.A. in Political Science and and B.A. in Economics.
I research how the connections between voters influence political behavior, in particular political coordination. In my current project, I measure the face-to-face social networks of more than 4,000 people and the online networks of millions. Onto these networks I layer experiments to examine how these networks shape political coordination behavior.
In other work, I examine the role of cognition in decision making, but with experimental subjects who hold influential positions in government and industry.
Political Science 204b: Introduction to Data and Regression This is the introductory course for first-year PhD students in the Political Science program. We focus on the mechanics and assumptions that underlie the most commonly used tool in the discipline. (Instructor) [Syllabus][Data Craft|HW][Sample Inference|HW][Regression|HW][LA Review|Regression2|HW][Joint Inference|HW][Interactions & Dummies|HW|code][Outliers & Assumptions|Code|HW][Model Problems][Causality Intro][Time-Seriesx]
Political Science 270: Math and Probability Foundations A (re)introduction to core math and probability concepts for first year political science PhD students. Topics covered in the course include calculus, optimization, linear algebra, and probability. (Instructor) [ Syllabus ] [ Discrete Probability | Conditional Probability | Continuous Probability | Random Variables | Special Distributions ]
Political Science 271: Quantitative Methods II This course moves beyond the linear model of PS 204 to cover the theoretical basis for maximum likelihood estimation (MLE). Particular focus is placed on the application of these models to political science data. Ordinal and count models, duration and survival models, time-series cross-sectional data, as well as likelihood and bayesian concepts will be covered. This is your swiss-army knife. (Lab Instructor) [Course Page]
Partial Effects Plot There are (too?) many calls out there to visualize the effects of choosen RHS-variables on LHS-variables, but for what I was working on, this was quicker. This function takes a stored model from a lm(...) call and produces a scatter plot of the RHS and LHS variabes, the best fit line, and confidence envalope. The function is easily extendable to handle different models. Because prediction requires you set variables, a word of caution: RHS variable correlation is poorly handled (at R's predict.lm(...) level).
Odds ratio for fit models from geeglm I was writing a paper for a journal that requested reports for odds ratios for binomial models. Golly, it was frustrating to tinker with the model, and then produce all the OR anew.
texReg output for geeglm objects A method for extracting model output from texReg, which is nicely extensible. Immediate inclusion of that sweet clustered model wherever you want it is now possible. I'll leave it put if you want to source(...) this at the top of your script. Otherwise, you can do as you like.