When to use this
Use this when the research question concerns association, prediction, or adjusted relationships.
Why it matters
Clear roles and diagnostics are needed to interpret coefficients responsibly.
Step-by-step instructions
1. Define the outcome and predictors from the research design.
2. Choose Correlation, Linear Regression, or Logistic Regression.
3. Select compatible variables.
4. Review missing-data handling and requested diagnostics.
5. Run the model.
6. Inspect coefficients, intervals, residual or fit diagnostics, and effective sample size.
Expected outputs
Coefficient tables, uncertainty intervals, model diagnostics, and saved version-bound results.
How to interpret the results
Coefficients estimate conditional relationships under the model. Statistical association alone does not establish causation.
Common mistakes
Do not select predictors only because they are significant. Check coding, multicollinearity, functional form, sample size, and theoretical justification.
Was this article helpful?
Your response helps the methodology team improve this documentation.