Statistical governance & analysis plan templates

Make your analyses planful, auditable, and reproducible — not an afterthought.

Why this matters

Unplanned analytic choices are a leading source of irreproducible results. When teams don’t record decision rules, version code, or review analyses consistently, small choices can change conclusions and undermine trust. These templates help you capture the who, what, when, and why of every analysis so results are easier to defend, reproduce, and learn from.

What you'll understand and be able to do

Using the templates and checklist here, you will be able to:

• Write a concise Statistical Analysis Plan (SAP) that defines primary outcomes, comparisons, estimands, and handling of missing data before looking at the final results.
• Specify decision rules (e.g., stopping rules, subgroup criteria, inclusion or exclusion logic) so they’re transparent and reproducible.
• Adopt simple versioning practices and code‑review steps to track analytic changes and who approved them.
• Use a reporting checklist to confirm diagnostics, assumptions checks, and sensitivity analyses are documented.

Who benefits

These materials are practical for individual researchers, lab groups, data scientists, quality teams, and project managers across contexts — clinical and preclinical research, observational health studies, environmental monitoring, manufacturing quality control, product A/B testing, and educational research. Small teams and solo investigators can use the same lightweight practices larger groups apply to maintain rigor.

Practical examples

Examples of how the templates apply in different settings:

• Clinical research: preregister primary and secondary endpoints, define multiplicity adjustments, and record blinded review decisions.
• Manufacturing QC: specify acceptance criteria, control-chart analyses, and actions for out-of-spec results with documented decision gates.
• Product experiments: declare outcome metrics, sample-size rules, and stopping boundaries for A/B tests to avoid selective reporting.
• Academic/field studies: preserve versioned analysis code and a checklist of diagnostics before drafting results to support reproducibility and peer review.

How to use the resource

Start by adapting the Statistical Analysis Plan template to your project: declare outcomes, estimands, handling of missingness, and planned models. Use the decision‑rules template to capture branching logic and stopping criteria. Pair templates with the included Statistical Analysis & Reporting checklist to confirm assumptions checks and sensitivity analyses are complete. Where helpful, convert static checklists into interactive forms or saved audits so teams can record decisions and preserve a searchable record of analytic choices.