semEffect: Unified Effect Analysis and Visualization for Structural Equation Models in R

Description

Provides a unified interface to calculate, extract, and visualize standardized effect decompositions (direct, indirect, and total effects) for three major structural equation modeling (SEM) frameworks: 'lavaan', 'piecewiseSEM', and 'plspm'. The package simplifies comparative SEM analysis by offering consistent functions across different modeling paradigms. Key features include automated handling of zero-effect variables, generation of publication-ready 'ggplot2' visualizations, and returning both wide-format and long-format effect tables. It supports flexible effect filtering, accepts multi-model object inputs, and allows extensive customization of visualization parameters. The 'sem_modeling()' function serves as a central wrapper, enabling users to fit models using different packages with a unified syntax.

Citation

Please cite the package as:

Mei et al. (2025) semEffect: Unified Effect Analysis and Visualization for Structural Equation Models in R. R package version 1.3.0.

Installation

Option 1: Install from CRAN:

install.packages("semEffect")

Option 2: Install from GitHub:

install.packages("devtools")
devtools::install_github("PhDMeiwp/semEffect", dependencies = TRUE)

Option 3: Install from Gitee:

install.packages("remotes")
remotes::install_git("https://gitee.com/openResearch/semEffect.git")

Core Functions

1. sem_modeling() - Unified SEM Fitting Interface

Fit SEM models using different packages with consistent syntax:

sem_modeling(model, data, type = "lavaan", modes = NULL, ...)

2. get_effect() - Effect Extraction

Extract standardized effects from fitted SEM objects:

get_effect(object, target, delete_zero_effect = TRUE, zero_threshold = 1e-10)

3. plot_effect() - Effect Visualization

Create publication-ready visualizations of effect values:

plot_effect(object, target, total_only = FALSE, 
            total_color = "skyblue",
            color_palette = c("darkgreen", "skyblue", "orange"))

Usage Examples

Example 1: lavaan Framework

library(semEffect)
library(lavaan)

# Define model using lavaan syntax
model <- "
  # Measurement model
  ind60 =~ x1 + x2 + x3
  dem60 =~ y1 + y2 + y3 + y4
  dem65 =~ y5 + y6 + y7 + y8

  # Structural model
  dem60 ~ ind60
  dem65 ~ ind60 + dem60
"

# Fit model using sem_modeling
fit <- sem_modeling(model, data = PoliticalDemocracy, type = "lavaan")

# Extract effects for target variable
effects <- get_effect(fit, target = "dem65")

# View results
print(effects$effect_table)

# Create visualization
p <- plot_effect(fit, target = "dem65", total_only = FALSE)
print(p)

Example 2: piecewiseSEM Framework

library(semEffect)
library(piecewiseSEM)

# Create model list
# option 1
model_list <- "
    rich ~ cover
    cover ~ firesev
    firesev ~ age
"

#option 2
model_list <- list(
  lm(rich ~ cover, data = keeley),
  lm(cover ~ firesev, data = keeley),
  lm(firesev ~ age, data = keeley)
)

# Fit model using sem_modeling
pmod <- sem_modeling(model_list, data = keeley, type = "piecewiseSEM")

# Analyze effects with custom colors
effects <- get_effect(pmod, target = "rich")
plot_effect(pmod, target = "rich", 
            color_palette = c("darkgreen", "grey80", "purple"))

Example 3: plspm Framework with Lavaan-style Syntax

library(semEffect)
library(plspm)
data(satisfaction)

# Define model using lavaan-style syntax
model_plspm <- "
  # Measurement model
  IMAG =~ imag1 + imag2 + imag3 + imag4 + imag5
  EXPE =~ expe1 + expe2 + expe3 + expe4 + expe5
  QUAL =~ qual1 + qual2 + qual3 + qual4 + qual5
  VAL =~ val1 + val2 + val3 + val4
  SAT =~ sat1 + sat2 + sat3 + sat4
  LOY =~ loy1 + loy2 + loy3 + loy4

  # Structural model
  EXPE ~ IMAG
  QUAL ~ EXPE
  VAL ~ EXPE + QUAL
  SAT ~ IMAG + EXPE + QUAL + VAL
  LOY ~ SAT + IMAG
"

# Fit model (modes default to all "A" reflective)
pls_fit <- sem_modeling(model_plspm, data = satisfaction, type = "plspm")

# Extract and visualize effects
effects <- get_effect(pls_fit, target = "LOY")
plot_effect(pls_fit, target = "LOY", total_only = TRUE,
            total_color = RColorBrewer::brewer.pal(5, "Set2"))

Advanced Features

Customizing Visualizations

# Customize plot appearance
p <- plot_effect(fit, target = "dem65", total_only = FALSE)
p +
  ggplot2::coord_flip() +
  ggplot2::theme_minimal() +
  ggplot2::ggtitle("Standardized Effects for Dem65") +
  ggplot2::theme(legend.position = "bottom")

Handling Zero Effects

# Keep zero-effect variables in results
effects <- get_effect(fit, target = "dem65", delete_zero_effect = FALSE)

# Adjust zero threshold
effects <- get_effect(fit, target = "dem65", zero_threshold = 1e-5)

Model Comparison

# Compare different SEM frameworks on same data
lavaan_fit <- sem_modeling(model, data = PoliticalDemocracy, type = "lavaan")
plspm_fit <- sem_modeling(model, data = PoliticalDemocracy, type = "plspm")

# Compare effects
lavaan_effects <- get_effect(lavaan_fit, target = "dem65")
plspm_effects <- get_effect(plspm_fit, target = "dem65")

Contributing

We welcome contributions to the semEffect package:

• Bug reports and feature requests: File them on https://github.com/PhDMeiwp/semEffect/issues
• Pull requests: Submit via https://github.com/PhDMeiwp/semEffect/pulls
• Documentation improvements: Help us improve examples and documentation

License

This package is distributed under the GPL-3 license.

Contact

For questions and support, please use the https://github.com/PhDMeiwp/semEffect/issues or contact the maintainer directly.