7 Secrets of Root Cause Analysis
This article offers practical root cause analysis tips that are secrets – if you define “secrets” by how many people need to learn them. We don’t keep these tips under wraps, but it may sometimes seem that way.Why? Because we find so many people who don’t understand what we see as obvious. So, in an effort to help organizations improve their analyses, we thought we’d break down what we see as the seven most important secrets. Read on to discover each one.
Secret 1: Your root cause analysis is only as good as the information you collect. Or, garbage in equals garbage out.
Most root cause analysis systems operate as a standalone module. Information goes in and an answer comes out. The systems don’t help investigators collect accurate information. Further, some root cause analysis tools actually begin by developing a hypothesis and then use the information collected to verify or disprove the hypothesis. Extensive research has shown that once an investigator becomes invested in a particular hypothesis, their brain begins looking for facts to confirm it – and disregards facts that are counter to the hypothesis. The result? You find what you want to find. This is not a robust root cause analysis process.
We recommend a different approach: Organize the information you’ve collected into a timeline of what occurred. By visualizing the incident in a linear way, investigators can begin to better understand what happened and what information they are still missing.
And that’s it. That’s the first secret – to obtain accurate, necessary, complete information about the incident to understand what happened. If you instead try to analyze assumptions, you will end up guessing at root causes.
Secret 2: Your knowledge – or lack of it – can get in the way of a good root cause analysis.
Some people assume that an investigator who has seen a cause produce an effect can use that knowledge to diagnose future problems. An obvious issue with this approach is that inexperienced investigators don’t know all the cause-and-effect relationships that exist. They can’t find what they don’t know.
But even experienced investigators may be led astray by the assumptions behind cause-and-effect analysis. How? For starters, just as with amateur investigators, even the most experienced investigators don’t know all the cause-and-effect relationships that exist. Experienced investigators can also fall into the “favorite-cause-itis” trap, in which their biases lead them to find cause-and-effect relationships that they are familiar with. Why? Because that is what they automatically look for – they seek familiar patterns and may disregard counterevidence; this is also referred to as “confirmation bias,” a bias that all humans exhibit at one time or another. And the more experienced an investigator is, the more likely they are to fall into this trap.
To overcome these types of problems, be sure you don’t rely only on your knowledge when performing a root cause analysis. That’s where investing in a software or other expert system comes into play. Look for a system that will help you overcome favorite-cause-itis and confirmation bias; ensure the system you use will double-check your work and prevent you from overlooking potential root causes.
Secret 3: You must understand what happened before you can understand why it happened.
This secret seems obvious – of course you must understand what happened. But some investigators, as well as some root cause analysis tools, start by asking why an event occurred when they should be trying to understand what happened. Starting by asking why can lead an investigator to jump to conclusions regarding a cause because they didn’t first seek to understand what actually occurred. Thus, it’s important to collect and organize as much information as possible about the incident that’s being investigated. As more information is collected, the investigator might realize that the incident was more complex than they initially thought.
Secret 4: Interviews are not about asking questions.
The way you interview people is important (remember Secret 1 – your root cause analysis is only as good as the information you collect). Consider this: After an incident, what happens when you ask an interviewee, “Why did you do that?” Does the interviewee answer with information about the event, or do they begin to feel defensive and focus their attention on justifying their actions? For some people, the “why” question turns their focus from remembering an incident’s details to explaining why they chose to engage (or not engage) in certain actions.
The purpose of an interview is to collect information – not justification. Interrupting an interviewee who is recounting their memory of an incident can cause a loss of potentially critical information. Each time the interviewer interrupts the interviewee, the interviewee’s brain must shift gears. They may lose their train of thought and forget about certain details.
Secret 5: You can’t solve all human performance problems with discipline, training and procedures.
If you review most incident investigations, you’ll find three standard corrective actions: (1) disciplining employees who made errors; (2) training, which may be the most used – and misused – corrective action of all; and (3) procedures (i.e., if you don’t have one, write one, and if you already have one, add to it).
The misuse of these standard corrective actions is the reason that so many incident investigations don’t cause performance to improve. That’s because they don’t address the real root causes. You’re probably also aware of the dreaded “re-corrective action,” including retraining, reinforcing and reworking. But is that the smartest route? Remember, doing the same things over and over but expecting different results is how some people define the word “insanity.” If it didn’t work last time or the time before that, why will it work this time if you don’t make any changes?
Thus, the fifth secret is that you need more than just the three standard corrective actions to create performance improvement.
Secret 6: People can’t always identify effective corrective actions even if they correctly identify the root cause.
Why is this the case? It’s often because they have performed the work the same way for so long that they can’t imagine another way of doing it.
Here’s an example of a poor corrective action: Back in 1994, a team of students analyzed the root causes of a fairly simple incident. One of the causes was that the valves being operated were not labeled. So far, so good.
But here was their corrective action: “Tell operators to be more careful when operating valves without labels.” They couldn’t see that the valves could, in fact, be labeled. It was beyond their experience.
We realized we needed to do something differently to help people see various corrective action alternatives. This eventually led to us developing the Corrective Action Helper Guide, which offers suggestions that can aid investigators in discovering other ways to solve problems beyond those they have experience with.
Secret 7: All investigations do not need to be created equal – but some investigative steps can’t be skipped.
People sometimes cringe when it’s suggested that a root cause analysis be performed. They think this means hiring a team of well-paid experts who will spend months trying to come to a conclusion. And it might be true – some investigations take a long time and cost a lot of money. But that doesn’t mean that every root cause analysis will take too long or cost too much.
A root cause analysis should be scaled to the size of the problem and the risk of similar future incidents. Small risk equals a small investigation. If it’s a big risk? Then spend more time and more investigative effort. The difficult part of responding appropriately is projecting the risk of future incidents before the investigation starts. This is why you need to gather the facts, build a timeline and adjust your response as necessary if the facts change. Also keep in mind that your time is a precious resource. Any good root cause analysis process will indicate when an investigation should be stopped because the value of continuing it is less than the risk the incident presents.
About the Authors: Mark Paradies is president of System Improvements Inc. and co-creator of the TapRooT RCA System (www.taproot.com). He is a board-certified ergonomist as well as a human factors expert with a master’s degree in nuclear engineering.
Alex Paradies is director of projects and innovation for System Improvements Inc., a TapRooT instructor and a Lean Transformation leader.