Improved Inference in Mediation Analysis: Introducing the Model-based Constrained Optimization Procedure

This repository contains the supplemental materials for the manuscript entitled Improved Inference in Mediation Analysis: Introducing the Model-based Constrained Optimization Procedure. Here is a list of the documents:

• SuppMaterials-MBCO-rev1.Rmd: The document explains the R code used in the manuscript. We discuss in detail the steps required to analyze our empirical example and re-produce the results.
• memory_study.csv: The data file that contains the empirical data example. When using the data, please cite the relevant study by MacKinnon et al. (2018).
• rsq.R: This file contain the rsq() function that produces R^2^.
• simulation folder contains the simulation functions used in the study.
  • The main two files for the single-mediator and two-mediator model are run_sim_1m.R and run_sim_2m.R, respectively. The other files in the folder are read into these two main files during the simulation runs.

References

MacKinnon, D. P., Valente, M. J., & Wurpts, I. C. (2018). Benchmark validation of statistical models: Application to mediation analysis of imagery and memory. Psychological Methods, 23, 654–671. https://doi.org/10.1037/met0000174