Module 7: Reporting Regression Results
Purpose
Regression results must be reported in a way that a supervisor, reviewer, or reader can understand and verify.
Learning Objectives
After this module, learners should be able to:
- Create a regression results table.
- Write an interpretation paragraph.
- Report model fit and diagnostics.
- Avoid causal overclaiming.
Minimum Reporting Items
Report:
- SmartPLS version
- Dataset and final sample size
- Dependent variable
- Independent variables
- Control variables
- Standard error type
- Test type and significance level
- Coefficients
- Standard errors
- t-values
- p-values
- Confidence intervals
- R-square and adjusted R-square
- Diagnostic summary
Regression Table Structure
| Predictor | b | beta | SE | t | p | 95% CI |
|---|---|---|---|---|---|---|
| Usefulness | ||||||
| Ease of use | ||||||
| Trust | ||||||
| Service quality |
Results Paragraph Template
A multiple linear regression was conducted in SmartPLS 4 to examine predictors of [dependent variable]. The model explained [R2]% of the variance in [dependent variable] (R2 = ..., adjusted R2 = ...). Holding the other predictors constant, [predictor] had a positive and significant association with [dependent variable] (b = ..., beta = ..., t = ..., p = ..., 95% CI [..., ...]). The diagnostic plots did/did not indicate major violations of regression assumptions.
What Not to Write
Avoid:
- "The model proves..."
- "X has an impact" when the design only supports association.
- "The hypothesis is accepted" without explaining the effect size and direction.
- Reporting only p-values.
- Ignoring diagnostics.
Lab
Complete Lab 6: Write the final results section.
Template
Use Regression Results Report Template.
Next Module
Continue to Module 8: PLS-SEM Measurement Model.