- New JobMondelezCI Engineer
If data is stable, can we conclude that this is normal? how do the two compliment and how will they differ?
Stability and normality are unrelated concepts.Stability assesses whether the data is stable over time and does not exhibit any trends. Controls charts are often used to assess stability and it is true that some of them require the data to be normal. For non-normal data a Run Chart can be used to assess stability.Normality assesses whether the shape of the distribution of the data matches a theoritical bell shaped distribution.Hope this helps.Rgds,Adrianhttp://adrian.smith.name
Normality does not imply stability and stability does not imply normality. They neither compliment nor differ they are two independent concepts.
When in a DMAIC project you are assessing your Y metric you test it for four parameters-stability (run chart), shape (normality), spread (Cp) and centering (Cpk)
Stability implies that there are no special causes in your process. Normality implies that mean is the correct measure of your central tendency and std deviation the correct measure of your variation
A process having special causes can be normal (in shape) and vice versa
The forum ‘General’ is closed to new topics and replies.