Supporting citations use verified paraphrases and PDF-sheet locators from the canonical local copies.
A flavor-development trial becomes interpretable when one planned factor changes while component identity, quantities, proof, vessel, temperature, contact time, sampling, and evaluation conditions are recorded and otherwise held stable.
Reviewed 2026-08-29 against the cited source pages during the complete text-and-visual corpus pass.
A one-factor comparison is one possible design, not a mandatory minimum or universal foundation. Interactions and constrained mixtures may require factorial or mixture designs from the outset; replication and evaluation controls remain necessary.
Controlled parameters separate explanations, while extraction and finishing examples show why ingredient, time, proof, and treatment must be explicit.
Created from a deduplicated cross-source synthesis; retain qualifications when drafting. Fresh Molecular Gastronomy audit confirms PDF79 controlled-parameter passage but not a universal one-variable rule; other supporting sources and relations preserved, not independently reverified by this source audit. Fresh Richardson audit:257–258and279 provide direct practical comparison/record support;162alone was indirect. New evidence linked; no universal one-variable rule or physical validation inferred.