|
Amran Malaysia
|
Tools for Finding Root Cause in 6 Sigma
Which tool is preferable in Six Sigma methodology to detect or find the root cause from special cause variation and then solve accordingly? And how to differentiate if the tool is suitable to use for common cause variation or special cause variation? Thanks for your help...
X
Sign up for free
Welcome to the Six Sigma best practices of 12manage.
Here we exchange knowledge and experiences in the field of Six Sigma.
❗Sign up now to gain access to 12manage. Completely free.
X
Continue for free
Please sign up and login to continue reading.
Here we exchange knowledge and experiences in the field of Six Sigma.
❗Sign up now to gain access to 12manage. Completely free.
|
|
|
|
|
Miranda, Netherlands
|
|
Check Data I would say that you could check the data using a histogram. If the RC mentioned does not appear regularly / more than once in a larger period of time, it would most likely be a special cause variation.
Next to that I would recommend to always ask the employees who participate in the process: if they have been working within that process for quite some time, they will be able to judge and advise if it's common cause or special cause variation.
To statistically see the difference, you can use a control chart (minitab).
Outliers (above or under the ucl/lcl) need to be investigated and prevent it from happening again.
Note: process should be stable and predictable in order to recognise valid outliers.
Hope this helps, good luck!
|
|
|
Pashupati Nath verma Professor, India
|
|
Cause-effect Diagram & Relationship Diagram Cause-effect diagram, also known as Fishbone diagram, proposed by Ishikawa, helps in identifying root causes of the quality failure.
Pareto analysis can suggest you which cause of quality failure is more serious and thus must be rectified first.
Relationship diagrams may also help you.
|
|
|
Amran Malaysia
|
|
Problem Solving Thank you Miranda. I would conclude the histogram and control chart are able to detect and differentiate the common and special cause. And to identify the root cause due to special cause can be done by asking the employee who participate the process, then solve accordingly. That is good recommendation as the corrective action always requires people’s experience to identify the source of problem. Next question in my mind is what would be a correct question to ask the employee in order to get correct answer to source of variation or root cause from special cause, because sometimes the employee is not aware the change in the process.
Hi Pashupati, I use Ishikawa and the tool definitely can help to identify huge number of possible cause but ample time is required to validate and verify the cause. Sometime it requires additional resources to test it. Often the special cause happening requires immediate corrective action, and thus using Ishikawa may require too much time to find and eliminate the root cause.
|
|
|
Mike Watson Manager, United Kingdom
|
|
Ishikawa Diagrams I find Ishikawa valuable and use it a lot in my work in the aircraft industry. However I would caution its use for definitive root cause analysis as your team will put forward ideas and hypotheses as to what they THINK are the causes.
The problem comes afterwards as some of them may be spurious and you must invest the time in exploring each one further before formally identifying it as a true root cause (or part contributor).
I usually use a 5Y on each to test the logic. It is a bit long winded, but don't forget that you will presumably be looking to invest in deploying proper corrective actions to prevent recurrence. It is important to be sure you have identified the root cause objectively.
|
|
|
Jagdish B Acharya Consultant, India
|
|
A Mix of Techniques Depending on Problem Statistics can point out that the problem is due to assignable causes but cannot tell anything about the causes. For that one has to have technical knowledge of the processes.
Ishikawa diagram throws many causes and usually is helpful in not missing any cause. Pareto helps in focussing on the vital indications for causal analysis. 5Why is by far the best in getting to a root cause faster.
I prefer to talk to all persons or look at reports of all sub processes and parts that result in a problem. If during the period of change anything has been changed, it may be the root cause. So going after changed variables often may lead to root cause.
|
|
|
Cristian Bleotu, Romania
|
|
Tools for Finding Root Causes Hi Amran, basically you've got the answer from miranda and jagdish:
1. Identify special cause variation using control chart (if exclude it from the data, the new control chart have to show a stable process)
2. Dig down for the special cause variation -identify root causes using existing data, information from the process actors and basic tools like 5 Why's / Ishikawa.
3. Apply measures so that those events never happen again (and track the results).
Cristian.
|
|
|
Gary Westerdale Director, United States
|
|
Software Tools for Finding Root Causes Two tools for root cause analysis are;
- Taproot
- Apollo
Both have methodologies and associated software.
|
|
Comments by date▼