Deming and Shewhart Common cause and special cause (statistics)

within frequency probability framework, there no process whereby probability can attached future occurrence of special cause. 1 might naively ask whether bayesian approach allow such probability specified. existence of special-cause variation led keynes , deming interest in bayesian probability, no formal synthesis emerged work. statisticians of shewhart-deming school take view special causes not embedded in either experience or in current thinking (that s why come surprise; prior probability has been neglected—in effect, assigned value zero) subjective probability doomed hopelessly badly calibrated in practice.

it apparent leibniz quote above there implications sampling. deming observed in forecasting activity, population of future events while sampling frame is, inevitably, subset of historical events. deming held disjoint nature of population , sampling frame inherently problematic once existence of special-cause variation admitted, rejecting general use of probability , conventional statistics in such situations. articulated difficulty distinction between analytic , enumerative statistical studies.

shewhart argued that, processes subject special-cause variation inherently unpredictable, usual techniques of probability not used separate special-cause common-cause variation. developed control chart statistical heuristic distinguish 2 types of variation. both deming , shewhart advocated control chart means of assessing process s state of statistical control , foundation forecasting.