Counterfactual Analysis Workshop

Counterfactual Analysis Workshop!

Date: 03/05/2024

Location: Room G32, 7 George Square, Edinburgh, EH8 9JZ

Description:

Counterfactual analysis encompasses methods that adjust for confounding variables in assessing the association between a treatment/event and an outcome. This often involves matching or weighting individuals based on confounding variables across different levels of exposure to the treatment/event. Successfully matched groups enable comparison of outcomes, facilitating more confident attribution of the outcome differences to the treatment/event. In addressing complex problems where experimental manipulation is impractical, large-scale data collection efforts (Generation Scotland, UK Household Longitudinal Study, UK Biobank, Millennium Cohort Study, etc.) and social media (Reddit, Twitter/X, etc.) offer vast observational data. However, these data are susceptible to reverse causation, misidentification, confounding, and social selection. Counterfactual analysis mitigates these vulnerabilities, enabling the identification of active ingredients—potentially influential factors—for positive societal and cultural change.

The principles of counterfactual analysis, though not novel, remain relatively obscure and challenging to apply. Our team has developed DigiCAT, a free, easy-to-use tool designed to facilitate counterfactual analysis in line with best practices. This workshop will provide participants with a foundational understanding of counterfactual analysis and walk-through examples utilising DigiCAT to address research inquiries. Delivered in-person by the DigiCAT team at the University of Edinburgh's central campus (George Square), the workshop will provide an introduction to counterfactual analysis and demonstrate its implementation using DigiCAT.

Who should attend:

Welcome to anyone interested in learning more about Counterfactual Analysis.

Registration:

Limited spaces available, so reserve your spot now!

Before the workshop:

Feel free to check out our tutorials (https://uoe-digicat.github.io/)

Download DigiCAT using r code:

install.packages("remotes")

remotes::install.github("uoe-digicat/DigiCAT")

DigiCAT::run.DigiCAT()