Key points at a glance
- Define the binary outcome.
- Separate odds from probability.
- Check data and model stability.
Code the outcome clearly
For a binary outcome, logistic regression models the log-odds of an event. State what is coded as 1 and identify predictor reference categories. It is not a universal replacement whenever some variable is non-normal.
Interpret an odds ratio correctly
Example: An odds ratio of 1.5 for study time means 50% higher odds per stated unit under the model, not automatically 50% higher probability. The probability change depends on the starting level. Report the confidence interval and unit.
Check data and model
Examine missing values, rare events, strongly correlated predictors and the functional form of continuous predictors. Too many predictors for few events can make estimates unstable. Discuss whether the sample supports your proposed model.
Show limits in the write-up
Report coding, sample, coefficients or odds ratios, uncertainty and a suitable model check. Regression on observational data does not by itself establish causation. Predicted probabilities at clearly stated baseline values can aid interpretation.
Final submission context
Formal details can feel like a separate writing task, but they matter just as much for a printed submission. Anything missing, misplaced or inconsistently formatted in the document will also appear in the bound copy.
Use this guide together with your cover page, table of contents, page numbers, source notes and appendices. Prepare the final PDF for printing and binding only after the complete file has been checked.
Practical check before PDF export
- Do headings, chapter structure and page numbers match?
- Are sources, figures and tables included completely?
- Are required elements such as declarations, appendices or the cover page included where required?
- Did you open and check the final PDF after exporting it?
Checklist
- Event and references defined.
- Event count and missingness checked.
- Odds ratios and intervals reported.
- Causality not overstated.
Common mistakes
- Equating odds and probability.
- Using too many predictors for few events.
- Treating regression as proof of causation.
Prepare the finished thesis for submission
Once the content and university requirements are checked, review your PDF and the available printing and binding options.
Review printing optionsRelated content
Explore related guidance on thesis structure, sources, formatting and final submission.
Frequently asked questions
When logistic rather than linear regression?
A suitably binary outcome often points to logistic regression; match the final method to the question.
What does OR = 1 mean?
No difference in modeled odds for that comparison.
Must predictors be normally distributed?
That is not a general requirement of logistic regression.