Choosing a regression family from the outcome and data structure
Model choice begins with the estimand, outcome support, dependence, sampling, and missingness, not a significance threshold.
Model choice begins with the estimand, outcome support, dependence, sampling, and missingness, not a significance threshold.
Define the estimand and outcome
Binary events, counts, ordered categories, continuous scores, and time-to-event outcomes imply different parameters and interpretations.
Map how observations depend on one another
Repeated observations, clusters, matching, survey sampling, or censoring change uncertainty and often the model itself.
Use diagnostics to test the working model
Functional form, residual patterns, influence, calibration, and sensitivity checks reveal where the approximation may be inadequate.
Failure modes to check
- Choosing a method because it is familiar rather than because it matches the estimand and data structure.
- Treating assumptions as a generic pass/fail checklist instead of evidence about a specific model.
- Changing the workflow after seeing results without recording the decision or its effect on uncertainty.
- Reporting software output without tracing the claim to final code, diagnostics, and limitations.
Continue the decision path
Explore the related statistical consulting service, or discuss the decision in your own study.
Related decision guides
Methodological note
This guide provides general educational information. The appropriate decision depends on the study question, design, data, evidence, ethics, and institutional requirements.