Problem Solving, Root Cause & The Five Whys
Finding the cause of a problem helps us make a higher quality decision about what action to take and makes development of an optimum corrective action more likely. But sometimes discovering the immediate cause of a problem doesn’t give us all the information we need.
Many problems are caused by a chain of events in which one problem leads to another, which leads to another. For this reason, tracking root cause can give us a more complete picture of how the observed problem came into being. With a complete understanding the root cause, (the entire chain of events), we can consciously decide where it would be most appropriate to take action.
This article provides a simple definition of root cause, one that has “emerged” from the review and facilitation of several thousand root cause investigations we’ve conducted with a particular client over the past five years. We will share insightful tips about how best to use the Five Whys procedure to help the troubleshooter avoid logical errors, minimize frustration, and obtain more satisfactory results.
Five Whys – Definition
The “Five Whys” is a troubleshooting procedure suggested by Masaaki Imai and it was made popular as part of the Toyota Production System in the 1970’s. Application of the procedure involves taking any problem and asking “Why – what caused this problem?” Then, when the cause is identified, asking “Why?” again (i.e. “what caused the cause?”)
The strategy, as commonly understood, is to ask “Why” about five times thereby uncovering links in a causal chain going backward in time.
Incorrect Definition of Root Cause
We have surveyed participants in our critical thinking workshops about their own working definitions of root cause.
They tell us:
root cause is “the cause which when removed corrects the problem.”
root cause is what started the sequence of events that led to the problem”
or words to that effect.
Also, as typically understood, the Five Whys procedure leads backward in time, step by step to the “root cause” of the problem (the earliest link in the causal chain.)
We have found the conception of root cause as the end of a chain is unsatisfactory when developing the best corrective actions. The intervening links between the first and last often contain vital information for helping to correct the problem. The entire chain of events is important, not just the last link in the chain.
Interruption or removal of any link may stop the ultimate effect from happening.
Correct definition of Root Cause
We define root cause as simply the uncovering of how the current problem came into being. For a simple causal chain, it is the entire chain. For a complex system of interlocking paths and events, again it is the entire thing. You know you are done gathering information when you see the complete picture of how this particular problem came into being and are ready to consider what to do about it.
The magic bullet is not to be found only at the end of a causal chain; in fact, there is no end, there is only a decision to stop gathering information. Examining only the tip of the root without considering everything else it’s attached to is a limiting approach. Having an awareness of the entire chain will help create a wide range of possible corrective actions from which to choose. Fixating on the earliest link of a causal chain or network is simply not effective as a general approach and there is no logical reason to limit ourselves in this way.
Problem Solving with the Five Whys
The BPI problem solving process is a tool for determining the cause of a problem. It is used to answer the question “Why?” one link at a time.
Problem Solving is a cause analysis tool that can be used when following the Five Why procedure to answer “Why?” at any step where the answer is unknown or contested.
Five Whys – Four Tips
The following are several tips for making better use of the Five Whys procedure.
- Verify the cause before proceeding. One way to improve the use of the “Five Why” procedure is to insist that all answers be verified (to the extent possible.) It is one thing to ask what caused something and it is another to actually know the answer! It is not enough to give an answer that seems right or popular. It is crucial that each link in the causal chain is the verified, true cause. False links will lead to ineffective corrective actions (at best) and disasters (at worst). (Use the BPI Problem Solving tool when cause analysis is required to verify a link.)
- Find the cause first then decide what to do to fix the problem.
It is common to confuse the cause of a problem with the world’s failure to implement your solution! For example, one might see the cause of the increase in absenteeism as a lack of an absenteeism policy. And, one might believe the cause of a child eating cookies before dinner to be the lack of discipline by the parents. Or, perhaps one thinks the cause of a defective part delivered to a customer was due to the fact that we do not have a final inspection point prior to shipping.
- Though a new policy may have an effect on the absenteeism behavior, a missing policy does not cause absenteeism.
- While discipline can affect a change in the child’s behavior, the lack of discipline is not the cause of the behavior. A child eats cookies because she is hungry and she thinks cookies are delicious.
- Installing an inspection station could prevent the delivery of defective parts after they are created. But, installing an inspection system will not stop the creation of defective parts.
The focus needs to be on what caused this problem, today. Then we can discuss what to do about it. What was the sequence of events that produced the observed problem with the existing system in place? Ask, “How did this problem come into being?”, and lay out the answer step by step.
This is an historical investigation – don’t change the system or anything else yet! Put a hold on ideas for how to prevent this problem in the future.
Logically, deciding what to do to fix a problem comes after a full understanding of how the problem was created in the first place.
Properly implemented, the Five Why procedure will often reveal higher level causes. For example, a problem with a part failing is caused by an improper fitting technique / this was caused by ignorance of the proper technique / this was caused by a manager decision to cancel training / this was caused by reduced staffing levels and so forth.
But, one problem we have observed with using the Five Why procedure is that people get stuck at one level. When they should proceed deeper down the causal chain (from problem, to a part, to a procedure, to the system, to a management decision), instead people get stuck providing more and more detail about one link.
After looking into this, we have determined that people get stuck because at some point they created one (or more) compound cause statement(s) in the Five Why series.
Compound Cause Statements Example:
(How the mind interprets WHY? in parentheses.)
Problem: SUV Model Z exhaust system rattle
- Why? (Why does the exhaust system rattle?) Because of a change of position of bracket results in vibration.
- Why? (Why does the bracket position cause vibration?) Because the bracket is too close to the pipe resulting in vibration.
- Why? (Why does being very close cause vibration?) Because the vibration from the pipe and vibration from the road are additive due to harmonics.
- Why? (Why are road plus pipe vibrations harmonic?) …STOP! (we are off into a technical rabbit hole thinking about “HOW” position created more vibrations and not progressing deeper.)
Asking “Why?” of a compound cause reverses the analysis to a cause-effect sequence and out of the effect-cause “5 Whys” pattern. The reverse sequence of causes is interrupted because the mind becomes confused about what “Why?” refers to.
Statement (A) contains a compound cause. Asking “Why?” tricks the mind into interpreting the question as “how did bracket position result in vibration?” The chronology is wrong. We’re not trying to explain “how” the bracket position caused vibration. We want to focus on the bracket positions and find out “What caused the bracket to be in that position?”
Always work going back in time from cause to its cause not forward explaining cause to effect.
Statement (B) also contains a compound cause. Asking “Why?” tricks the mind into interpreting this question as “Why (HOW) does being very close create vibration?”
The following shows the Five Why procedure without the trap of compound cause statements leading to (or towards) a systemic cause:
Problem: SUV Model Z exhaust system rattles
- Why? (Why does the exhaust system rattle?) Because of the exhaust pipe vibration.
- Why? (Why does the pipe vibrate?) Because the exhaust bracket is very close to the support.
- Why? (Why is bracket now very close to the support?) Because the line workers installed bracket in this location.
- Why? (Why did workers install bracket very close to support?) Because specifications stipulate this new location.
- Why? Unknown
Notice this analysis correctly moves backward in time, making clear the entire chain of events which led to the presently observed problem.
The 5-Whys procedure is not an ANALYSIS tool. You need a cause analysis tool to use with the 5-Whys. The BPI Problem Solving process is a powerful tool for determining cause. It answers the question “Why?” and can be used when needed to proceed down the causal chain, one link at a time. Please click Contact at top or bottom of this page and let us know what information you might want.