I'm looking for a suitable control chart to use to monitor the time is takes to respond to a corrective action request.
Historical data (2008) parameters are as follows:

Population - 177
Mean - 6.66 days
Median - 4.93 days
Std Dev - 6.896 days
Range - 4.62 days
Dist - RH Skew (No positive fit in Minitab)

Target - 3 days

There are a number of outliers present in the dataset.

I've looked at the data using various control and time series charts including subgrouping by week /month and am still unsure what is best to use.

I'm also trying the 'baseline' the process using a capability indices but the outliers are proving problematic from a stability standpoint.



I have wonder if one of our Stats experts could give me some guidance as to:

the most effective way to 'baseline' this data and
what type of control chart would be most suitable to monitor this type of data is the future?
Many thanks in advance
Steve

Just a few comments about the problem. First, control charting time to respond to a CAR is a terrible statistic. You have a target of 3 days, so does it matter if you miss the target by 1 day or 10 days? If not, then you can use a p-chart to look at percent of CARs that meet target. You can do this on a monthly cycle and use the np limits. Second, time to respond to a CAR is a lag function. So if the CAR was initiated today and I do not respond till September, the control limit would be out of control (theoretically) in September for an event that occurred in July. More importantly, unless you can normalize the time for the volume of CARs, the statistic is not very representative. If the number of CARs per day is charted, you might be able to make sure the process is operating. The 3 day limit is arbitrary. The system might not be capable of meeting the deadline with the current staff levels.

I think you need to look at your metrics and assess the best method for creating a dashboard. You can PM me if you have other questions.

Instead of a control chart, use a run chart. You can set the time frame by week/year or month/year. Average the number days to close during each time frame and plot that data. You may also want to plot the maximum time frame above the average line. On the same chart plot your goal of 3 days. This will allow you to show your improvements over time.

When you take improvement action to corrective action response time, make a note on the chart and record the action