The 5 Whys and Three-Way Five Whys

The 5 Whys (Five Whys) and More

One of the most well‐known RCA techniques is called the Five Whys. Toyota developed it in quality circles in Japan and it has been used across the motor industry and beyond. It arose from a ‘Lean manufacturing’ approach, seeking to make things ‘right first time’.

Thus, if things didn't turn out as expected on the production line, operators learned to ask why, why, etc., to diagnose the reasons for this, and as a result, to put in place solutions to fix problems over the long run. Other Lean techniques developed alongside the Five Whys include poka yoke (or mistake proofing) and a kaizen culture (of continuous improvement).

As you gain experience with why questions, it can become clear there are different ways to focus these ‘why’ questions. So, if something bad happens, you can ask:

• Why measures that might have prevented things from going wrong didn't work or work in time;
• Why measures that might have detected a threat before it went wrong didn't work or work in time;
• Why recovery measures that might have reduced the impact of a problem didn't work or work in time.

Thus, we arrive at the first RCA technique that goes beyond just asking why five times; it's called the ‘three‐way five whys’, where we ask ‘why’ in relation to prevention, detection and recovery and recognise the need to go beyond looking for just one root cause. A template for the three‐way five whys is provided in Diagram 1.4. It would be completed with answers to why something went off track, supported by appropriate evidence to underpin each of the conclusions so that they can be defended.

So, if a serious incident arises because of a leaking fuel pipe (say on an aircraft), it is quite possible it could arise because of all three of the following:

• A flaw in the fuel pipeline – which, if built with a better design, might have prevented the leak; and
• The absence of a fuel pressure monitor – which, if it was there, might have detected that fuel was leaking; and
• A missing secondary fuel pipeline – which might have been switched to if the first fuel pipeline failed.
• All three of these factors are likely to be important, because if any one of the measures was working fully effectively, they might have stopped or reduced the impact of an incident in the first place.

Doing RCA using the three‐way five whys can be likened to ‘digging into’ all the facts and looking for the important roots. Thus, the three‐way five whys can also be depicted as a ‘fault tree’ using Diagram 1.5.

You can see how an analogy with the roots of a plant is useful. If you don't find and pull out all the roots, the plant grows back. If you remove all the roots, it won't grow back. So, suppose you see repeating issues or similar issues arising. In that case, there may still be root causes that have not been identified or addressed. The technique can support both ‘single‐loop’ learning, which improves action within existing rules, and ‘double‐loop’ learning, which questions the governing rules or assumptions themselves. Tracing a deeper cause does not by itself make the learning double-loop.

The fault tree in Diagram 1.5 is one example of various ‘tree’ diagrams that can be used in RCA, e.g. issue trees, logic trees, etc. Tree diagrams are commonly used in oil and gas, aviation, and engineering environments, but can be used in a range of other contexts. They have power in the way that they can show how causes can ‘split’ (i.e. be the result of one thing as well as another). Further information about tree approaches appears in Appendix B.