Statistics are essential to many practices of validation but often treated as “off-the-shelf” tools that regulators can deploy as needed. This paper uses the introduction at the U.S. FDA of two methods between 1930 and 1980—dose-response extrapolation for ascertaining toxicity and statistical designs for interpreting clinical trials—to show the negotiations between statistical knowledge and regulation. Statisticians were not proposing their discipline as a tool for objective decision-making so much as a set of methods for establishing a regulatory procedure. They were far more circumspect about the ability of statistical methods to provide objective, accurate assessments of fact for regulators.

This paper examines the fraught mediation of vision testing under transformation in the early 20th century, as illuminated by a 1926 malpractice lawsuit in colonial Shanghai. It explores how experience of dizziness, a seemingly mundane side effect of new glasses, became a focal point for clinical, commercial, and legal disputes over the validity of eyesight tests. Weaving together competing perspectives of an Indian customer, a Jewish ‘optometrist,’ British legal professionals, European physicians, a female patient witness, and local entrepreneurs, the paper uncovers enduring individual and social anxieties over whose judgment — examiner, instrument, or patient — should be prioritized or regulated to ensure a valid test. Drawing on disability media studies and the history of medicine, it highlights the discrepancies between the examiner’s observations, the machine’s findings, and the patient’s lived experience. Despite the growing epistemic power of medical professionals and diagnostic technologies in standardizing diagnostic practices, this study argues that uncertainties voiced by patients and consumers continually challenged the validity of vision testing in the complex interplay of knowledge and ignorance within scientific communication across China, the United States, and the British Empire.