Introduction to projectLSA

Overview

projectLSA provides a unified Shiny-based environment for conducting Latent Structure Analysis (LSA), including:

  • Latent Profile Analysis (LPA)
  • Latent Class Analysis (LCA)
  • Latent Trait Analysis (LTA / IRT)
  • Exploratory Factor Analysis (EFA)
  • Confirmatory Factor Analysis (CFA)

The package is designed for users who prefer a graphical workflow without writing code, while still leveraging robust statistical methodologies implemented in well-established R packages.

Installation

install.packages("projectLSA")     # CRAN version (once released)
library(projectLSA)

Launching the Application

library(projectLSA)
run_projectLSA()

This will open the full Shiny interface, where you can upload data, choose an analysis module, and generate results.

Modules Included

1. Latent Profile Analysis (LPA)

  • Fit multiple models with varying profile numbers
  • Compare AIC, BIC, entropy, and class sizes
  • Visualize best model with custom profile names

2. Latent Class Analysis (LCA)

  • Designed for categorical indicators
  • Simulated and uploaded datasets supported
  • Class probability tables and interactive visualizations

3. Latent Trait Analysis (LTA / IRT)

  • Supports dichotomous and polytomous items
  • Rasch / 2PL / 3PL and graded models
  • Item information curves, test information, and multi-dimensional visualizations

4. Exploratory Factor Analysis (EFA)

  • KMO test, Bartlett test, and parallel analysis
  • Loading matrices, rotation, and factor scores
  • Auto-generated interpretation summaries

5. Confirmatory Factor Analysis (CFA)

  • Lavaan syntax editor
  • Fit indices, loadings, modification indices
  • SEM path diagram with customizable styles

Example Workflow

Below is a simple workflow using the built-in datasets.

library(projectLSA)

# Launch the GUI
run_projectLSA()

Once inside the GUI:

  1. Choose a module (e.g., LPA)
  2. Upload your dataset or select a built-in dataset
  3. Choose variables and model settings
  4. Fit the models and explore the outputs

Built-in Example Datasets

The package includes several example datasets:

  • pisaUSA15 — student motivation indicators
  • curry_mac — moral relevance & judgment (simulated)
  • id_edu — longitudinal educational identity (simulated)

These are accessible from within the Shiny interface.

Reproducibility and Reporting

projectLSA provides:

  • Exportable tables (CSV, Excel)
  • Downloadable graphics (PNG)
  • Reproducible summaries and model comparisons

This ensures results produced through the GUI can be published or documented with confidence.

Citation

Please cite this package as:

Djidu, H., Retnawati, H., Hadi, S., & Haryanto (2025). projectLSA: An R Shiny application for latent structure analysis with a graphical user interface.

Session Info

sessionInfo()