Bootstrap Confidence Intervals for Ordinary Least Squares Factor Loadings and Correlations in Exploratory Factor...

Area of Science:

• Psychometrics
• Statistical Modeling
• Data Analysis

Background:

• Exploratory Factor Analysis (EFA) is crucial for understanding complex data structures.
• Accurate confidence intervals for factor loadings and correlations are essential for reliable EFA results.
• Traditional methods for confidence intervals in EFA may have limitations.

Purpose of the Study:

• To evaluate the performance of bootstrap confidence intervals for rotated factor loadings and factor correlations in Ordinary Least Squares (OLS) EFA.
• To compare various bootstrap interval types (SE-based, percentile, bias-corrected, BCa, hybrid) under different simulation conditions.

Main Methods:

• Utilized simulation studies with varying sample sizes, model fit (perfect/imperfect), and data distributions (normal/elliptical).
• Applied Ordinary Least Squares Exploratory Factor Analysis (OLS EFA).
• Investigated five types of bootstrap confidence intervals.

Main Results:

• Bootstrap confidence intervals demonstrated effective coverage performance.
• The bootstrap method proved particularly effective for assigning confidence intervals at moderately large sample sizes.
• Illustrated the application of bootstrap confidence intervals using a real-world personality dataset.

Conclusions:

• The bootstrap is a reliable and effective technique for constructing confidence intervals in OLS EFA.
• Results support the use of bootstrap methods for enhancing the interpretability and reliability of factor analysis findings.
• The study provides practical guidance on selecting appropriate bootstrap intervals for EFA.