CONCEPT

Tampering

Deming's name for the most common management error in any system governed by variation: treating common-cause noise as if it were a special-cause signal, reacting to each instance individually, and thereby increasing the very variation the reaction was meant to reduce.

Tampering is the management equivalent of steering a car by overcorrecting—each correction a response to the last deviation, each response introducing a new deviation, the car tracing an increasingly erratic path because the driver is reacting to noise rather than holding a steady course. W. Edwards Deming identified tampering as the most pervasive and most destructive management error in any system governed by variation, and he demonstrated it with Walter Shewhart's canonical insight: variation in any process comes in two fundamentally different forms. Common-cause variation is inherent in the system's architecture—the random scatter of results that any process produces simply by being the process it is. Special-cause variation is produced by an assignable factor, something identifiable and removable that lies outside the system's normal behavior. These two types of variation require categorically different responses. Common-cause variation is addressed by redesigning the system—changing the architecture, the tools, the workflow, the incentive structure. Special-cause variation is addressed by

Keep reading with YOU ON AI