Regression in SmartPLS 4.0

Lab 4: Regression Bootstrapping

Goal

Run regression bootstrapping in SmartPLS 4 and use bootstrap confidence intervals for final inference.

Initial check:

Subsamples: 1,000

Final reporting:

Subsamples: 10,000
Confidence interval: Percentile bootstrap
Test type: Two-tailed unless directional hypothesis was pre-specified
Significance level: 0.05
Seed: Fixed seed for reproducibility

Steps

  1. Return to the model window.
  2. Select Calculate.
  3. Select Regression Bootstrapping.
  4. Choose the final settings.
  5. Run the calculation.
  6. Open the bootstrapping report.
  7. Export coefficient table and confidence intervals.

Interpretation

A coefficient is significant when its confidence interval does not include zero at the selected confidence level.

Deliverable

Submit a bootstrapping table:

Predictor b t p 95% CI Decision

Next Lab

Continue to Lab 5.