Repository for "The Economic Consequences of UN Peacekeeping Operations: Causal Analysis for Conflict Management and Peace Research"

💬 Talk on "Sensitivity Analysis for Inverse Probability Weighting Estimators via the Percentile Bootstrap" (Q. Zhao et al., 2017), for S. Pimentel's "Observational Study Design and Causal Inference" seminar at Berkeley, Spring 2018

An implementation of g-methods

A questionnaire containing 40+ questions is given to hundreds of people. People are interviewed about their feelings and hobbies with a goal to find the causal relationship between depression and cognitive impairment, where some questions are related to depression, some to cognitive impairment, and others are confounding. In psychological survey…

Optimization of KNN and Linear regression algorithms

Code for assessing the causal effects of chemotherapy Received Dose Intensity (RDI) on survival outcomes in osteosarcoma patients using a Target Trial Emulation approach.

Inverse probability weighting for non-binary exposures. Simple example in Excel and SAS.

Conducted a inverse probability weighted survival analysis in State Veteran Home facilities which investigated the hazard of being infected by COVID among varying vaccination doses and types. Involved complex data linkages to person-level and facility-level data and creating a nested control group of those who did not receive the bivalent booster.

💬 Talk on causal inference and variable importance with stochastic interventions under two-phase sampling

Non-parametric variable selection and inference via the outcome-adaptive Random Forest (OARF). Uses the IPTW estimator to estimate the ATE while the propensity score is estimated via OARF. This leads to smaller variance and bias. Only variables that are confounders or predictive of the outcome are selected for the propensity score.

Positivity violations in marginal structural survival models with time-dependent confounding: a simulation study on IPTW-estimator performance.

Tools for using marginal structural models (MSMs) to answer causal questions in developmental science.

IPW- and CBPS-type propensity score reweighting, with various extensions (Stata package)

The R package trajmsm is based on the paper Marginal Structural Models with Latent Class Growth Analysis of Treatment Trajectories: https://doi.org/10.48550/arXiv.2105.12720.

R package for estimating balancing weights using optimization

📦 🎲 R/medshift: Causal Mediation Analysis for Stochastic Interventions

Targeted maximum likelihood estimation (TMLE) enables the integration of machine learning approaches in comparative effectiveness studies. It is a doubly robust method, making use of both the outcome model and propensity score model to generate an unbiased estimate as long as at least one of the models is correctly specified.

Use regression, inverse probability weighting, and matching to close confounding backdoors and find causation in observational data

📦 R/medoutcon: Efficient Causal Mediation Analysis with Natural and Interventional Direct/Indirect Effects

📦 R/haldensify: Highly Adaptive Lasso Conditional Density Estimation