Bayesian final project report template
Executive summary
State the scientific question, the main model, the main conclusion, and the main uncertainty caveat.
Data and problem setup
- Data source
- Units of analysis
- Key measurement issues
- Target estimand
Model specification
Write the model carefully, including priors and parameterization choices.
Prior predictive checks
Summarize what the priors implied before fitting and any revisions made.
Posterior computation
- Inference method
- Tuning choices
- Runtime
- Convergence diagnostics
Posterior results
Report:
- parameter summaries;
- transformed scientific quantities;
- predictive summaries.
Model criticism
Include posterior predictive checks, residual checks, or calibration checks as appropriate.
Sensitivity analysis
Document how the main conclusions changed under plausible alternative assumptions.
Predictive or decision evaluation
Explain how models were compared and why the reported metric matches the scientific goal.
Limitations
State clearly:
- remaining model assumptions;
- unresolved computational issues;
- threats to interpretation.
Reproducibility
- repository structure;
- environment information;
- instructions to rerun the analysis.