Reproducible Output
Reproducible Output keeps figure-facing text tied to the same reviewed export state that produced the result. Instead of writing Methods or legends from memory, researchers start from structured drafts assembled from the Data Contract, disclosure state, and provenance metadata.
What is Reproducible Output?
Writing Methods and legends after the analysis is complete creates two risks: memory errors and selective omission. Researchers may forget which transformation, correction, or sample-size context mattered for the final figure, or they may understate caveats that were visible during export review.
Licklider addresses this by generating reviewable drafts from the same state used during export, so the figure, its text, and its provenance stay aligned.