PLS-SEM in SmartPLS
SmartPLS was built for partial least squares structural equation modeling (PLS-SEM). You draw the path model, run the algorithm and assess measurement and structural model in one application.
Overview
PLS-SEM estimates composite-based path models and is widely used when the goal is to explain and predict key target constructs. SmartPLS covers the complete workflow, from estimation and significance testing to model assessment, prediction and the analysis of group differences.
What SmartPLS provides
- Estimation: the PLS-SEM algorithm, consistent PLS (PLSc-SEM) and weighted PLS.
- Inference: bootstrapping with percentile and bias-corrected and accelerated (BCa) confidence intervals, and permutation tests.
- Model assessment: reliability and validity criteria, discriminant validity with HTMT, and model fit.
- Prediction: PLSpredict and the cross-validated predictive ability test (CVPAT).
- Extended relationships: mediation, moderation, nonlinear effects, higher-order constructs (repeated indicator and two-stage approaches) and Gaussian copulas for endogeneity.
- Groups and segments: measurement invariance (MICOM), multigroup analysis, FIMIX-PLS, PLS-POS and importance-performance map analysis (IPMA).
Typical uses
- Explaining and predicting key target constructs in management, marketing, information systems and the social sciences.
- Models with composite constructs, mediators and moderators.
- Theses and dissertations that need a transparent, well-documented analysis.