Plan an audit
996102e0
Audit planner
Use this planner to turn study conditions you already know into a transparent navigation path through GazeAudit. It does not choose thresholds, exclude participants, declare validity, or decide which scientific assumptions are defensible for your study.
Select the conditions that actually apply.
Choose one or more statements. The planner will deduplicate the corresponding method families, preserve their canonical order, and explain why each method entered the route.
Your selected audit path
Ordered by the governed method catalog, not by the order you clicked the choices.
Continue from methods into an end-to-end workflow
These handoffs are governed navigation links. They do not add methods, infer risks, or change your selected plan.
Choose the study conditions that are relevant above. If you already know the methodological question, go directly to the Method explorer.
Exporting a plan
The audit brief and JSON plan are portable navigation records. They preserve the conditions you selected, the deduplicated method route, the reasons those methods entered the route, workflow handoffs, the current plan URL, and the documentation build revision when available. They deliberately do not contain invented thresholds, exclusions, diagnoses, or case-study outcomes for your study.
The JSON export has a stable schema_version field and is deterministic for the same selection on the same documentation build. No timestamp is inserted, so two collaborators can compare the substantive plan rather than an incidental export time.
How the planner is governed
The planner has three machine-readable inputs. planner-index.json contains only the navigation rules shown on this page, method-index.json contains the ten governed method families, and planner-workflow-index.json contains the three end-to-end workflow handoffs. CI verifies that every planner rule references an existing method ID, every planner-reachable method maps to exactly one workflow handoff, and every generated guide, example, plot, evidence, and workflow link resolves.
The planner deliberately does not infer a missing risk from another answer. For example, selecting sampling-rate sensitivity does not automatically add missingness sensitivity. Likewise, selecting an empirical case-related method does not copy that case’s outcome onto your study. Workflow handoffs are downstream navigation only; they do not alter the plan encoded in the URL.
What to record in a study
A useful audit record should preserve why a method was included, the exact GazeAudit version or commit, the declared thresholds or perturbation levels, the executed specifications, and the resulting artifacts. Use the publication-audit guide when you need deterministic provenance for a manuscript or archive.