Introduction to Control Charts
IntroductionControl Charts
Sources of Variation
As we know, no two products or services are exactly alike because the processes used to produce them contain many sources of variation, even if the processes are working as intended. For example, the diameter of two crankshafts may vary because of differences in tool wear, material hardness, operator skill, or temperature during the period in which they were produced. Similarly, the time required to process two credit card applications varies because of the load on the credit card department, the financial background of the applicant and the skill and attributes of the employees. Nothing can be done to eliminate variation in process output completely, but management can investigate the causes of variation to minimise it.
Common Causes
There are two basic categories of variation in output:
Common Causes and Assignable Causes
Common causes of variation are purely random, unpredictable sources of variation that are unavoidable with the current process. For example, a machine that fills cereal boxes will not put exactly the same amount of cereal in each box. If you weighed a large number of boxes filled by the machine and plotted the results in a scatter diagram, the data would tend to form a pattern that can be described as a distribution. The mean, spread and the shape may characterise such a distribution.

Mean is the sum of the observations divided by the total number of observations

Spread is the measure of the dispersion of observations about the mean. Two measures commonly used in practice are the range and the standard deviation. The range is the difference between the largest observation in a sample and the smallest. Standard deviation is the square root of the variance of the distribution. An estimate of the population standard deviation based on sample is given by
= Σ
n
x x σ i
I x = Observations of a quality characteristic (such as weight)
x = Mean
n = total no. of observations
σ = Standard deviation of a sample.
Two common shapes of process distribution are symmetric and skewed. A Symmetric distribution has the same number of observations above and below the mean. A skewed distribution has a preponderance of observations either the above or below the mean.