Bayesian Regularization Methods for Factor Analysis: Simulation and Application Results

A 680.5 KB collection of simulation and application results comparing Bayesian regularization priors for factor analysis. The dataset, authored by Yifan Zhang and last updated in May 2026, likely contains performance metrics from studies evaluating graphical lasso, horseshoe, and spike-and-slab priors for estimating sparse latent factor correlations. It includes findings from a personality-inventory application demonstrating how partial regularization improves model interpretability and fit.

Use Cases

• Comparing the false-positive control and accuracy of spike-and-slab, horseshoe, and lasso priors based on simulation results described.
• Implementing partial regularization for multitrait-multimethod or extended bifactor models based on the methodological framework presented.
• Evaluating significance-based versus threshold-based model identification criteria for regularized estimators as discussed in the description.

Strengths

• Includes results from a specific personality-inventory application mentioned in the description.
• Compares three distinct Bayesian regularization priors (graphical lasso, horseshoe, spike-and-slab) as a unified framework.
• File size is precisely 680.5 KB, indicating a defined scope of content.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is small (680.5 KB), indicating limited scope.

Provenance

Source

figshare

Collection Method

Likely contains results from simulation studies and an empirical analysis, as described.

Freshness

Last updated 2026-05-27 18:29:01