Sure, that would be okay, thanks.
18 different charts? 18 sample points over 28 days
- Thread starter ricen
- Start date
Here is a copy of Steve Prevette's EXCELLENT article.
Steve - I thank you for this one. I've actually used it in all of my classes at my previous employers and it has been very effective over the years. I am always amazed at the number of people who mis-use Excel's forecast line and 'rolling averages' just because they were taught to in some 'business stats class'.
Steve - I thank you for this one. I've actually used it in all of my classes at my previous employers and it has been very effective over the years. I am always amazed at the number of people who mis-use Excel's forecast line and 'rolling averages' just because they were taught to in some 'business stats class'.
Through some off line work, I did develop this chart as an example. It takes the 18 sample points over 28 days and plots the highest conductivity value measured each day. I used the logic that if any individual sample point exceeded the limit for conductivity, it would be a problem. From there took each daily max value and plotted it as a X-chart, using 2.66 times the average moving range as the 3 standard deviation value. This gave a control chart with the UCL well below the limit, but one day was below the LCL. I did recalc the average and control limits while removing the low point (making the ASSUMPTION there was something indeed different about it). I think this represent a reasonable approach to monitoring the daily results - plot the worst case result and do a simple X-mR chart. Can always get more sophisticated from there, if need be. File attached with permission of the OP.
Similar threads
