Skills & Training

I served as the Co-Director for the Causal Inference between 2018 and 2022, where I oversaw research and development in causal inference. Given the data and settings at RAND, this typically involved using quasi-experimental methodology where methods aim to estimate causality based on natural experiments and statistical approaches that leverage different sources of exogeneity. Both while at RAND and continuing while at LinkedIn, I have done research using many different quasi-experimental methods, such as difference-in-difference, instrumental variables, weighting and matching estimators, and many other methods.

I designed and oversaw a randomized controlled trial of job training in New Orleans, where nearly 600 individuals were randomized into treatment or control. I have additionally led the design and launch of other experiments, including a field experiment of simulated business enterprises in high schools as well as experiments on LinkedIn platform.

While at RAND, I oversaw the weighting and sampling of the American Life Panel, a longitudinal panel of respondents. There, I generated survey weights for nearly 500 surveys and helped with sampling design in many cases. I also led and helped with the design of several survey instruments for different projects. 

I have led the economic modeling in many projects, covering cases such as dynamic programming, peer effects, cost analysis, and microsimulations. 

I co-taught a course on machine learning and causal inference. I have also used machine learning in various aspects of projects, including non-parametric methodology, neural networks, and random forests. While I am in many ways still an ML novice, I am an eager learner! 

• Mastery: R, Stata, SQL

• Experience: Matlab, Python, Scala

• Experience that is not recent: Perl, C, Fortran, SAS