The Landscape of Error in Surgical Pathology

Error: In everyday language, error is getting things wrong, usually in relation to aims and purposes [67]. There is a different technical use of the term error in statistics [27]. For statisticians, error means differences in repeated measurements. These measurement differences arise from either random variation or bias. Random variation causes inconsistent differences between measurements; bias produces systematic differences between measuring methods or devices. Ordinary language and statistical error: Ordinary language and statistical uses of error have this in common: we make both errors and measurements. Study of the two kinds of error connects in this way: observers detect differences between random variation and events gone wrong systematically by measuring characteristics of the events that fail to achieve their purposes. Observers may then act rationally from their understanding of nonrandom variation to reduce and sometimes prevent practical errors. This way of connecting systematic event measurement with process improvement follows from the insight into production processes first articulated by the statistician Walter Shewhart, then extended and made famous by, Shewhart’s student W.E. Deming [13, 79]. The Shewhart–Deming approach investigates practical errors, failures of steps in a process to achieve their objectives, and AQ1