APS
2024 APS Annual Convention · 2024
Navigating the Nuances of Causality: Comparing and Combining Conditional and Standard Direction Dependence Analysis in Multi-Group Scenarios
- Xintong Li
University of Missouri - Columbia - Adam Ueltschy
Westran R-I
Abstract
5 Mote Carlo simulations were conducted comparing CDDA and DDA in causal model selection across subgroups. It was revealed that while DDA is more powerful when effect directions are the same or reversed, CDDA excels in detecting confounding. Therefore, a combined use of CDDA and DDA for complex causal analysis.
Other