Here’s a 5-Why example from a recent posting on Business Week “Modern Analyst” saying how wonderful 5-Why’s is.
Issue: Employees did not receive their pay stubs on pay day.
· Why? Because the printing system failed the day before pay day.
· Why? Because the system could not recover from a hardware fault.
· Why? Because the system uses outdated hardware that has no automatic redundant backup.
· Why? Because the system hasn’t been replaced as it hasn’t been identified as a high enough priority to allocate budget to its replacement in the current economic climate.
· Why? Because the organization does not have an enterprise planning methodology that weighs the risks of current operational systems failing versus the criticality of these systems and the impact of such a failure.
Well, what do you think? Good or bad example?
What do you think of their “root cause”?
I’ll wait for others to post to share my ideas…
I think it is good. They found out that outdated equipment needs to be replaced.
Mark:
Thanks for posting this example of 5 Why.
It is typical of similar analysis that I have seen clients conduct.
So that said, this example is representative.
At the 3rd Why, it seems to go from factual to editorial.
“Could not recover from a hardware fault” – appears to have multiple branches that are more to the point than those indicated.
My assumption is that pay stubs were printed withing days or hours, once the proper hardware was replaced, after the appropriate technical resource conducted an analysis, once the resouce was notified and arrived or loged in.
To the point of your question, this is a great example of 5 Why in action. It also highlights how easily 5 Why provides the wheel with which to grind axes without the focused facilitation of the process.
Jeff
I think this is a bad example. 5-why in general works better if you use it on a causal factor or a cause(s) of your event rather than on the event itself. I generally use it in combination with fishbone, fault tree type tools after I generate some ideas of causes contributing to the event. Then I dig down into those causes by using 5-why’s for each of the significant causes identified.
I also feel that the 3rd why is not necessarily the most appropriate fit. The hardware being outdated didn’t cause it to fail. Was the asset properly maintained? Were the users properly trainined in its operation? There are lots of out-dated assets that are reliable if properly used and maintained.
I’m not a fan of the 5 Why’s for this reason. Without a strong systematic approach, end results can too often become op-ed pieces rather than a thorough investigation. Its either a bad example of a 5 Why’s or a good example of why not to use them. As others have said, the third point doesn’t adequately address what it should and then its off the rails.
I don’t see any fact-based information that connects each why with it’s next answer. That leaves me skeptical. Seems that the answers to why are entirely based on the training & experience of the person asking why. Even if the answer was correct, it could not possibly be a uniform result among multiple inquiries.
Im my experience the process resulted in the usual “blame the company, they should have got it right…” The first 2 whys are ok, afterwards there is no evidence to back up the statements as other people have indicated.
Furthermore going down the list, it is assumed by the person conducting the exercise that a “enterprise planning methodology” will magically identify the printing pay stubs on time it the most important thing in the company and should have lots of resources/money allocated to it.
Another key point missing is a ranking / analysis of the impact of the problem, and a decison made as to if something should actually be done about it. Is the late printing of pay stubs a problem that we really should address? or should we put our availaible resources to something more important?
I think it is a good example of forcing an “investigation” to the “root cause” that you want. Others have identified the 3rd WHY as questionable. I would propose that the first on is also questionable.
Saying that the printing failure on the previous day might be supported as a contributing factor, but there was apparently inadequate actions taken at that time. There may not be other ways to get the pay stubs to the employees, but the resetting of the machine after a hardware fault may have been done in time to prevent this “event”.
Was this the first and only time the printing system failed? How would our investigator know if it was a hardware fault?
There are many potential branches to explore. It would be inconceivable that the correct and only one would be found in this way.
Because it is unstructured, it leaves the “investigator” free to take any path and represent it as THE cause and effect. This allows the unimaginative to appear to do a thorough job. Another consequence is the resulting lack of confidence in any RCA method after seeing examples of use that demonstrate self-serving on the part of the facilitator.
As JR Ewing says, “Once you let go of integrity, the rest is easy”.
Chris inadvertently made the point as to the 5 why approach has some merit. He noted that some thought at least one of the 5 “why”s was questionable. It takes at least 4 or 5 “why”s to be answerred before you get to anything approaching a root cause. The last why – lack of an organizational enterprse management system – is deserving 5 “why”s of its own. At what point you want to draw the line on asking “why” is not always clear – just asking why 5 times doesn’t guarantee anything but the beginning of some penetration beyond the top level (proximate) cause. Another problem with the 5 why approach is that it tends to be single string – it does not branch out into separate lines of inquiry. A true root cause analysis methodology tends to look at a number of srings of causes.
OK – This has run most of the week – let me post my comments. (And please note that my comments do repeat some of the comments above.)
First, I think this is an excellent example of someone creating a 5-Why staircase that proves the point they wanted to prove. I frequently see people try to do this when they perform a root cause analysis. 5-Whys is especially susceptible to this “I already know the answer” error.
Second, this example doesn’t provide any information for someone to be able to peer review the analysts work. Who can tell for sure if this analysis is based on facts?
Third, there is only a single causal chain. Was their more than one causal factor and multiple root causes? Again, a single causal chain is a frequent problem I see when people use 5-Whys.
I think this is a “good” example of typical 5-Why analyses. Good … in that it is representative of what people do. But NOT good if you are really trying to solve a problem.
What would TapRooT® do different?
1) We would have a time line with facts (a SnapCharT®) to explain the details of what happened. This would provide the evidence needed to identify Causal Factors (and there would probably be more than one) and to find root causes.
2) If we decided it was needed, we could use Equifactor® as an equipment troubleshooting tool. If we frequently analyzed printer problems, we could develop a custom printer troubleshooting table to institutionalize our expert’s knowledge. We could also use heinz Bloch’s failure Modes and Failure Aganets to better understand what happened.
3) Each of the Causal Factors would have it’s own root cause analysis using the Root Cause Tree®. These analyses are less prone to jumping to conclusions.
4) Each of the root causes could have effective corrective actions by using the Corrective Action Helper® Module and SMARTER.
That’s my ideas…
Thanks to everyone who has commented so far and to those who post further comments in the future.
Best Regards,
Mark
Mark Paradies
President
System Improvements
An excellent example of starting from the wrong premise. This is not an issue, or a problem. The employees did not recieve the stubs, on pay day. So who cares? Why is this an issue. I can conclude that staff were paid, only the confirmation stubs were not printed?
Start with a problem statement or deviation statement; and then move forward.
The 2nd why then becomes, not that the printing sytem failed, but to the larger issue of notification method, why not email a pdf of the cheque to each employee? What alternative notirfications are there?
The 3rd why becomes, why was there no response plan?
Probably because , it was a simple failure, and by itself essentially free of consequences. Its makes for an example.
i read its nice example i like it. i am student of braindumps please post me