APS
2022 APS Annual Convention · 2022
Elevated Type I Error Rates of the Bias-Corrected Bootstrap Confidence Interval in Missing Data Methods for the Indirect Effect
- Tristan Tibbe
University of California, Los Angeles - Amanda Montoya
University of California, Los Angeles - Craig Enders
University of California, Los Angeles
Abstract
The bias-corrected bootstrap confidence interval for indirect effects has been found to have inflated type I error rates in complete datasets. This simulation study reveals even worse inflation when this method is combined with multiple imputation for missing data, suggesting that its use with missing mediation data should be stopped.
Other
