manymome: An R package for computing the indirect effects, conditional effects, and conditional indirect effects,...

Area of Science:

• Behavioral Science
• Statistics
• Psychometrics

Background:

• Mediation, moderation, and moderated mediation are prevalent in behavioral research.
• Existing tools for indirect effects lack user-friendliness for standardized conditional indirect effects.
• Some current methods are limited to specific model types.

Purpose of the Study:

• To introduce manymome, an R package for estimating and forming confidence intervals for various effects.
• To provide a user-friendly tool for standardized and non-standardized indirect, conditional, and conditional indirect effects.
• To enable researchers to focus on model fitting before effect estimation.

Main Methods:

• Developed the manymome R package.
• Employed a two-step approach: parameter estimation (SEM via lavaan or regression via lm) followed by effect computation.
• Supported missing data handling when using structural equation modeling.

Main Results:

• manymome can estimate and form confidence intervals for indirect, conditional, and conditional indirect effects.
• The package accommodates standardized and non-standardized effects.
• No inherent limitations on the number of predictors, independent variables, moderators, or mediators.

Conclusions:

• manymome offers a flexible and accessible solution for complex effect estimation in behavioral research.
• The package simplifies the process, allowing researchers to prioritize model development.
• It addresses a gap in statistical software for standardized conditional indirect effects.