In the classic approach to root cause analysis (RCA), we form a team of subject matter experts (SMEs) and set out to complete our interviews by dividing and conquering. The team splits up and each of us interviews a number of the subjects on the list. Then we return to our conference room, compare notes, and convey what we heard during each interview.
After three decades as a root cause practitioner in the commercial nuclear industry, the Nuclear Regulatory Commission, the Department of Energy and the nuclear weapons complex, a few issues came to light after evaluating why RCA teams experience a great deal of pain and anguish in bringing together all the information we gathered.
- Individuals, if left unchecked, introduce their biases into their answers. It takes additional follow-up to validate many of the statements and untangle much of the dissonance.
- Management interviews are often biased towards the "party line."
- Some individuals try to steer or dominate the conversation and turn it into a gripe session.
- RCA team members will inject their own biases into the conversation, even if the subject is not in their area of expertise.
- From an efficiency standpoint, individual interviews take a long time and generate copious notes that have to be later transcribed or somehow incorporated into the analysis.
When the team gets back together to share their notes the information has been biased by the individual, their managers, and even the RCA team member. The discussions that then ensue can be difficult, as each team member reports out on their observations and findings. In many cases, weeks and months pass while an RCA team slugs through this arduous process.
As my clients started asking me to lead more RCAs at the National Laboratories and the nuclear weapons complex, we always seemed to be short on time. This led to a sharp review of our current approach to interviews, and these are just a few of the adjustments you can make to eliminate bias and streamline the interview process at the same time.
- No more individual interviews (unless it deals with sensitive subject matter). Replace the one-on-one interviews with "facilitated causal analysis sessions" of small representative groups, and make sure you start with the workers before moving on to supervisors and managers. This alone will eliminate much of the bias introduced by individuals and managers.
- Do not separate the RCA team when conducing these sessions; they all need to hear the same thing at the same time because humans will hear what they want to hear, and they may capture the wrong context of the conversation. That will also go a long way to eliminating the biases introduced by team members.
- Do not take copious notes. Each bullet (note) you write down is like a small piece of a giant jigsaw puzzle. Imagine what happens when 5 RCA team members each bring about 200 notes from their interviews.
- Instead of notes, document their answers on your final product; the chart that is capturing your causal analysis. If you are using multiple tools to document your analysis, consider the BlueDragon Framework, which incorporates all the information you need in one place (i.e. timeline, admin requirements, physical barriers, multiple perspectives to pursue, lines of inquiry, cause & effect and more).
- During the sessions, validate each answer on your Why Staircases as you progress through the analysis by making sure that the representative group agrees with the answer you are about to write down on your chart...any dissenting opinions must be flushed out and reconciled before proceeding.
- If there are dissenting professional opinions between two groups, bring them back and iron it out together...only one right answer goes on the chart.
Also, to improve efficiency, leave the organization's SMEs off the RCA team...instead, bring them in as needed to provide some needed input up front, and to participate in the facilitated causal analysis sessions. Build the team with RCA experts that can facilitate the investigation of complex, human-centric problem (i.e. problems that cannot be solved with math and science alone). Clients love this approach because their valuable SMEs are not tied up for weeks and months on a root cause.
The end result is that my average time to complete the analysis portion of a root cause has dropped from 6 to 12 weeks with a team of 4 to 12 subject matter experts (depending on the complexity of the issue), to an average of 5 days with one assistant. I still harness the subject matter expertise of 50 to 150 of the organization's SMEs, but I usually only need them for a 2-hour window. These RCAs are some of the more difficult issues or events requiring the use of outside SMEs like myself.
If you would like additional information on dramatically improving the effectiveness of your root cause analysis, including eliminating bias and shortening their duration, visit us at: https://dle-services.com/bluedragon