Some Skills and Experience Sharing in Doing 5 Whys Analysis

                Problem: Superman was late for his work this morning
                Why 1:  Why was Superman later for his work this morning?
                              Because he woke up late.
                Why 2: Why did he wake up late?
                             Because the alarm clock did not ring.
                Why 3: Why didn’t the alarm clock ring?
                            Because the battery was dead?
                Why 4: Why was the battery dead?
                 Because it was not replaced promptly before the end of its lifetime.
                Why 5: Why wasn’t the battery been replaced promptly before the end of its lifetime?
                 Because there was no established plan to replace the battery regularly.   

     Above is an example of 5 whys analysis (please see this article for introduction of 5 why analysis). It is a powerful tool for root cause analysis, but if it's not properly used, it may not lead to the right root cause. Below are suggestions that the users of 5 whys analysis need to keep in mind in order to effectively use this tool.

Always Try to Find the Systematic Cause
    One of the widely mistakes in 5 whys analysis is that the users stop too soon before the real root cause is identified. One can use the following criteria to determine whether the cause he has identified is the real root cause or not:
    If an action taken to resolve the cause can effectively prevent the recurrence of the problem, the cause is the root cause.
    Generally the causes of any problem occurred in a company can be classified as technical causes and systematic causes. Technical causes are the technical issues with 5M1E (direct issues with machine, material, method, men, measurement and environment, e.g. malfunction of the machine, defect materials and incorrect processing parameters), which result in the occurrence of the problem. While systematic causes are issues in the management procedures (e.g. ineffective procedure to prevent the machines from malfunction) and the management team (e.g. insufficient allocation of resources) which result in the technical causes of the problem. Normally, when doing 5 whys analysis, one should not stop at the technical cause and should continue until the systematic cause is identified. It is important to realize that resolving the technical cause can only restore the situation to normal. It is just a corrective action and cannot prevent the recurrence of the same problem. In the “Superman is late for his work” example, the technical issue is the battery was dead. If Superman thinks this was the root cause and stops asking himself further "why"s, he’ll just replace the battery and move on with his life. Inevitably, however, the battery will die again and he’ll probably wake up late again, and the problem recurs. Therefore he should not stop at the technical issue. He should do further analysis until he finds out the systematic cause: he did not have a plan to regularly replace the batter before it reaches its lifetime.

Do NOT Attribute the Root Cause to People
    Another common mistake of using the 5 whys analysis is that people, but not the system, is attributed as the root cause of the problems. For example, operators being careless is often regarded as the root cause for defective products that the operators produce in the production line. Such analysis is meaningless, as no corresponding action exists to ensure that operators will be careful all the time in the future to prevent the recurrence of the problem. If human error is the technical cause of the problem, the person doing the 5 why analysis may further ask himself: Is the training provided to the operators sufficient? Are the operators properly evaluated after the training to ensure they’re capable of performing their tasks? Is there error proofing method or reminding system existed to prevent the human error? Accordingly, the root cause can be attributed to insufficient training, insufficient evaluation, no error-proofing method or no reminding system.

Keep Clear Cause-and-Effect Relation Between Each Why and Its Answer
    To lead to the right root cause with the 5 whys analysis, it must be kept in mind to remain the tight cause-and-effect relationship between the question and the answer of each "why", so that there in turn exists a cause-and-effect relationship between the root cause and the occurred problem. It is not rare to see that cause-and-effect relation is not clearly observed in the analysis. Particularly, I have personally seen many cases, in which the root cause of the problem were just re-statement of the problem in different words. 

Ask the Right Question 
    In 5 whys analysis, the answer to the current "why" is the question for the next "why". One may lose the cause-and-effect relationship between the root cause and the occurred problem if incorrect question is asked. Such a mistake is quite common when the answer to the current "why" is too long or multiple reasons are identified, and hence the user cannot find the proper question for the next “why”. It is suggested to make the answer short in each “why”, and the question of each “why” should be exactly same as the answer to the previous "why", so that it does not cause confusion and the cause-and-effect logic can be well maintained (Note: In cases that multiple reasons exist for a “why”, it is probably not a good idea to use a single 5 why analysis to find the root cause. Instead, one may use other methodologies, such as Fault Tree Analysis).

Avoid the Leap of Logic
    Another common mistake in 5 whys analysis is the leap of logic, i.e. the answer to a "why" is not the direct reason. In the “Superman is late for his work” example, the answer to the 1st "why" is “Because he woke up late”. Some people may answer “Because the alarm clock did not ring”. Alarm clock did not ring, however, was not the direct reason of being late. Waking up late was.
    When identifying the root cause of real issues occurred in a company, the leap of logic may occur when the user of 5 why analysis overlooks some steps in a process, e.g. the communication between owners of the different process steps. Let’s take a look at following example:
                Problem: a defective product is produced
                Why 1:  Why was the defective product produced?
                              Because the machine suddenly broke down.
                Why 2:  Why did the machine suddenly break down?
                              Because the machine is low of cooling oil.
                Why 3:  Why is the machine low of the cooling oil?
                              Because the operator did not refill the machine with the cooling oil.
                Why 4:  Why didn’t the operator refill the machine with the cooling oil?
                              Because the operator was not aware that the machine needs to be refilled with
                              cooling oil regularly.
                Why 5:  Why was the operator not aware that the machine needs to be refilled with
                              cooling oil regularly?
                              Because the operator did not receive such a plan.
               Why 6:  Why didn’t the operator receive such a plan?
                              Because the equipment engineer did not establish the plan.
When answering the 4th “why”, some people may write the answer of the 6th “why” instead. If so, there's a leap of logic. Therefore, the user of 5 why analysis should not overlook any step in a process which should be conducted in order to effectively prevent the occurrence of a problem.