S
Dear all,
For my first post, I would like to address the issue of "normal" versus "between/within" capability study.
It all boils down to a) the presence of subgroups and b) the calculation of standard deviation
In an aeronautical context, I need to assess a process performance based on sample batches (3/4-samples) extracted from structural parts, hence I consider each part as coming from a different process stream. I also consider each sample batch as a subgroup.
The current method uses the minimum of 2 indices, PpK(ind) based on SD of individual values, and PpK(avg) based on SD of averages of each subgroup, to determine performance.
Here are my questions :
Does PpK(avg) make sense ? (SD calculated straightforwardly with (n-1))
Shouldn't one use "between/within" capability study ? (variation between subgroups is important)
How do you justify this ? And more importantly, how do you conduct it ? I now have several SD : "within", "between", and "between/within"
I was thinking of computing PpKs using "between/within" instead of "overall" SD, which is twice bigger, and also "between" SD to show variability in-between process streams...
I cannot find much documentation on the web on "between/within capability study", but you folks seem to have all the savvy needed, so please feel free to drop your 2 cents worth, Ill very much appreciate it.
For my first post, I would like to address the issue of "normal" versus "between/within" capability study.
It all boils down to a) the presence of subgroups and b) the calculation of standard deviation
In an aeronautical context, I need to assess a process performance based on sample batches (3/4-samples) extracted from structural parts, hence I consider each part as coming from a different process stream. I also consider each sample batch as a subgroup.
The current method uses the minimum of 2 indices, PpK(ind) based on SD of individual values, and PpK(avg) based on SD of averages of each subgroup, to determine performance.
Here are my questions :
Does PpK(avg) make sense ? (SD calculated straightforwardly with (n-1))
Shouldn't one use "between/within" capability study ? (variation between subgroups is important)
How do you justify this ? And more importantly, how do you conduct it ? I now have several SD : "within", "between", and "between/within"
I was thinking of computing PpKs using "between/within" instead of "overall" SD, which is twice bigger, and also "between" SD to show variability in-between process streams...
I cannot find much documentation on the web on "between/within capability study", but you folks seem to have all the savvy needed, so please feel free to drop your 2 cents worth, Ill very much appreciate it.