TOPIC: Special Cause Variation- Control Charts

Hello All,

I came across a tricky question on Oliver Lehman 75q test.

Production process is intended to produce steel bolts with a length of 20 cm.
The Control Limits are 19.955cm and 20.045cm - the measures made at the end of the process yielded the following results:
20.033cm, 19.982cm, 19,995cm, 20.006cm, 19.968cm, 19.963cm, 19.958cm, 19.979cm, 19.959cm.

What should be done? Answer is: A SPECIAL CAUSE COULD BE INVESTIGATED, THE PROCESS SHOULD BE ADJUSTED.
Special Causes - are unpredictable
Common Causes/ Random Causes- are predictable
Is this a special cause b/c 19,955cm is not consistent with the balance of the metrics?

Hi Kristin,

Let's first discuss the difference between common cause and special causes.
On a control chart, common cause variations would have the pattern of

• all points within the control limits
• points distributed randomly on both sides of the average value

On a control chart, special cause variations would have the pattern of either:

• a point or more beyond the control limits
• some trends of the points (e.g. more than 5 consecutive points on one side of the average value)

Now, taking the numbers given in the question, the mean of the measurements comes to 19.983. If you see, 6 values are below the mean and 3 values are above the mean. Therefore, this hits the condition - "some trends of the points (e.g. more than 5 consecutive points on one side of the average value)" and as a result it is a special cause which could be investigaed and the process should be adjusted.

Hope this helps.

Hi Rahul,

Thank you for your detailed response, very helpful!
Will we have to make these calculations on the exam or will it be more of understanding what common causes vs special causes mean?
I'm asking b/c it can be time consuming.

Thanks.

Hi Rahul,
Based on these values, can you clarify this is how you calculated to get the solution.

20.033cm, 19.982cm, 19,995cm, 20.006cm, 19.968cm, 19.963cm, 19.958cm, 19.979cm, 19.959cm.
mean/ average= 19.983 (179.843/9) add all # values/divide by sum of #'s
below the mean: 19.982, 19.963, 19.958, 19.979, 19.959
above the mean: 20.033, 19.995, 20.066

Overall, more than 5 consecutive points on one side of the average value (below the mean) Therefore, a special cause could
be investigated, and the process should be adjusted.
Thank you!
Kristin

Hi Kristin,

Yes, that's exactly correct. And yes you can expect such question but the only calculation that is required is to take the sum and divide by the number of measurements (mean or average).

After that's its an eyeball check to see how many are below and above the mean.

Hope this helps.

Rahul-
That's great!
Thank you for your clarifications.
It sure did help me!

Thanks for the feedback Rahul. Apologies for posting in wrong section but Please concerning ITTOs and process flow for the 47 processes, please whats your advice? Your kind revert is most appreciated. Thanks. This email address is being protected from spambots. You need JavaScript enabled to view it.