SEM-PLS vs CB-SEM: Choosing the Right Statistical Method for Empirical Research

Q: What is the "10-times rule" in PLS-SEM and is it still valid?

The traditional 10-times rule suggested sample size should be 10 times the maximum number of structural paths. Modern methodological guidelines (Hair et al., 2022) recommend using formal statistical power analysis (via G*Power) targeting a minimum power of 0.80 at alpha = 0.05 rather than relying on the 10-times rule.

Q: When should I strictly choose CB-SEM over PLS-SEM?

Choose CB-SEM (AMOS, LISREL, lavaan) when the research goal is theory testing and confirmation of an established model, the sample size is large (N > 250), data meets multivariate normality assumptions, and all constructs are purely reflective.

Table of Contents

1. Foundational Differences
2. Methodological Decision Framework
3. Formative vs Reflective Models
4. Mandatory Reporting Metrics
5. Software Tools: SmartPLS vs AMOS vs lavaan

1. Foundational Differences

Structural Equation Modeling (SEM) is the dominant multivariate statistical methodology in management, psychology, marketing, and information systems. Two primary paradigms exist:

Covariance-Based SEM (CB-SEM): Minimizes the difference between the sample covariance matrix and the theoretical model-implied covariance matrix. Primary objective: Theory Confirmation and Model Fit.
Partial Least Squares SEM (PLS-SEM): Maximizes the explained variance (R² values) of dependent latent constructs. Primary objective: Exploratory Theory Development and Prediction.

2. Methodological Decision Framework

Criterion	Choose PLS-SEM	Choose CB-SEM
Research Goal	Predicting target constructs / exploratory	Testing, confirming, or comparing theories
Construct Type	Formative AND Reflective models	Strictly Reflective models
Data Distribution	Non-normal / skewed distributions accepted	Requires multivariate normal data
Sample Size	Effective with smaller samples (N=50-200)	Requires larger samples (N > 250)
Model Complexity	Handles high complexity (50+ variables)	Prone to non-convergence with high complexity

3. Formative vs Reflective Measurement Models

Misclassifying construct measurement models is a primary reason for journal desk rejections. In Reflective constructs, indicators are manifestations of the latent variable (dropping one indicator does not alter the construct meaning). In Formative constructs, indicators define and cause the construct (dropping an indicator omits a vital conceptual dimension).

4. Mandatory Reporting Metrics for Top-Tier Journals

When reporting PLS-SEM findings in Scopus/SCI journals, authors must present:

Measurement Model: Indicator Reliability (outer loadings > 0.708), Internal Consistency (CR > 0.70), Convergent Validity (AVE > 0.50), and Discriminant Validity via the HTMT criterion (HTMT < 0.85 or 0.90).
Structural Model: Collinearity (VIF < 3.3), Path Coefficients (β) with p-values via 5,000 bootstrap resamples, R² of endogenous constructs, f² effect sizes, and Q² predictive relevance.

5. Software Tools: SmartPLS vs AMOS vs lavaan

SmartPLS 4 and ADANCO lead the PLS-SEM domain, while IBM SPSS AMOS, Mplus, and R package lavaan dominate CB-SEM implementations.

Frequently Asked Questions

What is the "10-times rule" in PLS-SEM and is it still valid?

The traditional 10-times rule suggested sample size should be 10 times the maximum number of structural paths. Modern methodological guidelines (Hair et al., 2022) recommend using formal statistical power analysis (via G*Power) targeting a minimum power of 0.80 at alpha = 0.05 rather than relying on the 10-times rule.

When should I strictly choose CB-SEM over PLS-SEM?

Choose CB-SEM (AMOS, LISREL, lavaan) when the research goal is theory testing and confirmation of an established model, the sample size is large (N > 250), data meets multivariate normality assumptions, and all constructs are purely reflective.

VIGOORR Quantitative Analytics Lab

Statistical Methodologists & Econometricians

VIGOORR's editorial and advisory practice comprises experienced patent attorneys, PhD research methodologists, senior academic editors, and corporate commercial counsels with decades of combined practice across international institutions.