New method accurately estimates treatment effects in diverse heteroscedastic settings.
The article presents a new method to estimate the average treatment effect in a way that works well even when the data has different levels of variability. This method is shown to be accurate and reliable through computer simulations.