A defined workflow keeps analytical decisions and outputs traceable.
Provide the research question, dataset, expected methods, software preference and deadline.
The analytical specification defines variables, models, diagnostics, robustness tests and outputs.
Data preparation and estimation follow a documented code-based workflow.
Key outputs, sample sizes, model diagnostics and reported statistics are checked.
Revisions address analytical issues within the agreed scope. New research questions, new datasets or materially different models are treated as additional work.