Kontrollkaarte

Kontrollkaarte, known in English as control charts, are a tool in statistical process control used to monitor a process's variation over time. Each chart plots a statistic calculated from samples taken sequentially from the process, typically the sample mean (X) and a measure of dispersion such as the range (R) or standard deviation. A central line represents the process target or estimated mean, while upper and lower control limits define the expected range of natural variation, commonly set at ±3 standard deviations from the mean.

There are several common forms: X-bar and R charts, X-bar and S charts (for standard deviation), and individuals and moving-range charts when samples are single observations. The choice depends on data type, sampling frequency, and desired sensitivity.

To construct a kontrollkaart, collect sequential samples with size n, compute the statistic for each sample, plot it, determine control limits from historical data or process capability, and reassess when out-of-control signals occur.

Interpretation focuses on detecting non-random patterns that indicate a departure from stable, in-control variation. Points outside control limits or runs of points on one side of the center line suggest assignable causes requiring investigation. A stable process typically shows most points within limits and no systematic patterns.

Kontrollkaarte originated with Walter A. Shewhart in the 1920s and were later popularized by Deming. They are used across manufacturing, service delivery, healthcare, and software processes to improve quality and predictability.