Guide to Statistics
153
14. Guide to Statistics
14.1 General
Understanding statistics is about understanding variation. Variation is present
everywhere – in nature as well as in industrial processes. Without variation
evolution would not be possible, but variation can also cause problems. This
applies in particular to industrial processes, where even a slight deviation from
the target value – a dimension, for instance – may have a serious effect on the
functionality of the finished product. Therefore, understanding variation, and in
some cases controlling it, is crucial.
A stable process is a process in which no particular causes of variation are
present (particular causes of variation are, for instance, dull tools, inaccurate
tolerance settings or faulty materials). Even though there are no obvious causes
of variation, small differences in the outcome of the process exist. This is due to
the so-called common causes of variation. Measuring the outcome (a dimension,
thickness, torque value…) of a stable process more often than not results in a
specific pattern; i.e. the measured variable follows a normal distribution.
Two parameters describe the normal distribution; the mean, µ, and the standard deviation, often denoted
σ (sigma). The mean is a measure of the process level; i.e. the middle value, and the standard deviation
describes the scatter of the data.
No.
To Next Page
Loading...
