Random Variation
Random Variation, also called random sampling, random allocation, pseudo random allocation, and pseudo random exposure, is a situation where the observed variation of an independent variable does not necessarily follow a normally defined probability distribution. In other words, it is the deviation of the statistical distribution of the value of a variable from its mean or average value. This deviation can occur for any variable that is normally measured and used in scientific research, but especially so when dealing with quantities that are extremely large or unpredictable in nature. Random Variation is typically used in scientific and business studies where the level of variability can be extremely high; thus, the need to control for the deviation of statistical estimates from a normally distributed distribution.
Random Variation occurs frequently in the process control portion of the supply chain management system. The process control function includes such processes as product development, pricing, marketing, customer service, warehouse operations, as well as waste management and accounting. Any deviation from an expected path may lead to a process outage, costly overtime, or worst case scenario-a company may even go out of business entirely! Random variations in any of these key processes lead to delays in the completion of work and increased costs and/or errors. Random Variation is therefore an exceedingly important tool in controlling the variability in the variables being used in the SDCA process.
A random sample (also called a "black swan" effect) is an unexpected and undesirable outcome that can occur when a random number generator (RNG) is used to generate numbers or data sets in a system where the sample size is too small. Most computer systems will contain a random number generator (RNG), which generates numbers and probabilities based on the input parameters. Some examples of common RNGs used in the SDCA process include random number generators used in medical experiments; random number generators that are used in telecommunications; and random number generators that are used in the Internet. This form of statistical distribution can be extremely valuable in the analysis of statistical distributions used in safety-critical applications because it allows for the quick detection of errors and the correction of statistical errors or poor performance.
CIOP Webinar Series
Learn the Global Best Practices in Supply Chain Management, Procurement, and Inventory Management.
