|Title:||thinkCausal: Practical Tools for Understanding and Implementing Causal Inference Methods|
|Principal Investigator:||Hill, Jennifer||Awardee:||New York University|
|Program:||Statistical and Research Methodology in Education [Program Details]|
|Award Period:||3 years (07/01/2020 – 06/30/2023)||Award Amount:||$899,876|
|Type:||Methodological Innovation||Award Number:||R305D200019|
Co-Principal Investigator: Scott, Marc
Purpose: The purpose of this grant is to develop a highly scaffolded multi-purpose causal inference software package, thinkCausal, with the Bayesian Additive Regression Trees (BART) predictive algorithm as a foundation. This will allow education researchers from varied backgrounds to access and better understand these versatile estimation tools.
Project Activities: The core BART algorithms will be extended beyond what is available in current software packages to accommodate a wider range of data structures. The performance of the methodological extensions to the underlying BART software in R will be evaluated using simulations that extend the corpus of 7700 datasets created for the 2016 Causal Inference Data Analysis Challenge.
The new software will scaffold the education researcher through the data analytic process, from uploading data all the way through to graphical and tabular displays of results. The software will also have interactive educational components in order to provide researchers with the opportunity to gain a deeper understanding of the methods and their underlying assumptions. As part of the software development process, the research team will receive feedback on the software through their education research colleagues, through their advisory board research network and during training sessions at conferences. The researchers will also form an education-focused user group and a methods-focused user group via Google groups to provide a central forum for software experiences and issues of general interest.
Publications and Products
Dorie, V., Perrett, G., Hill, J. L., & Goodrich, B. (2022). Stan and BART for Causal Inference: Estimating Heterogeneous Treatment Effects Using the Power of Stan and the Flexibility of Machine Learning. Entropy, 24(12), 1782.