Appropriate Control Chart

F

Fiascosaif

Dear experts,

I have been trying to create a control chart in excel but I am not certain if I have chosen the correct control chart. Can someone review my working out and advise if I am doing things correctly?

My data looks at harm caused from an incident to patients during their occupied bed days. I have decided to go for a u chart on the basis that the occupied bed days can vary significantly and the incidents could happen multiple times to a single patient.

I have attempted to plot this out in excel however I am confused as to the numbers I am getting for the u bar, UCL and LCL. They are all small numbers 0.01 etc. Can anyone explain why there are small numbers? Also is it possible to show the data per 1000 occupied bed days?

I have attached the excel spread sheet if anyone would like to view it.

Your help/guidance is much appreciated!

Saif
 

Attachments

  • Copy of attribute_charts.xls
    96.5 KB · Views: 236

Miner

Forum Moderator
Leader
Admin
The u chart is appropriate to your scenario. Your UCL and ubar appear to be correct, but the LCL should be zero. LCLs of u charts cannot be negative.

The reason the numbers are small is because the chart is not plotting the number of incidences, but the number of incidences divided by the opportunities (beds).
 
F

Fiascosaif

Many thanks for the clarification Miner

If I wanted to see it displayed as per 1000 occupied bed days, is that possible? I have attempted to multiply the defects / unit by 1000 but then then graphs UCL LCL and u bar stopped working. Will I need to calculate them separately and then use those figures in the u charts as opposed to what's currently in them?

Regards,

Saif
 

Miner

Forum Moderator
Leader
Admin
You would have to create the chart using the raw data then add a multiplier on you r scale to show defects per 1000.
 
F

Fiascosaif

That's the part I am struggling on. Every time I try to add a multiplier the control limits don't work anymore, as in they become very small/close together. and the graph becomes huge. Any advice would be great
 
Last edited:
F

Fiascosaif

Just to clarify what I meant by the UCL and LCL going small, I have attached what happens when I tried to add a multiplier.
 

Attachments

  • Copy of attribute_charts.xls
    97 KB · Views: 332
Top Bottom