|ROOT CAUSE ANALYSIS:
A CASE STUDYIntroductionIn March 2004, CWA Ltd., in collaboration with The Great Lakes Area Regional Resource Center (GLARRC), and the Michigan Department of Education, Office of Special Education and Early Intervention Services (OSE/EIS), designed and conducted a root cause analysis workshop with the engagement of thirty stakeholders. The participants to the workshop were representatives from the community of practitioners in the field of a monitoring process called Continuous Improvement Focused Monitoring (CIFM). These practitioners are responsible, among other things, for implementing the new No Child Left Behind (NLCB) legislation, passed by Congress in 2001, for the state of Michigan. The participants were initially engaged in a series of workshops for the purpose of designing the CIFM process, which they will then have to implement in the field with school districts throughout the state.After the designers completed the design of the CIFM process, it was decided to conduct a “root cause analysis workshop” with the engagement of the designers/participants. The purpose of this particular workshop was to try to anticipate any factors that might inhibit the successful implementation of the CIFM process in the field. The intention was to conduct an anticipatory root cause analysis, as opposed to one that is the result of an existing system problem(s).
Observer-Independent Root Cause Analysis
Generally speaking, observer-independent root cause analysis is a databased procedure for ascertaining and “analyzing” the causes of problems in an effort to determine what can be done to solve or prevent them. The goal of root cause analysis goes beyond merely “fixing” the problem. It seeks to actually prevent it from happening again.
What Root Cause Analysis Needs to Include
Over 30 years of research and development teach us that effective and reliable root cause analysis must provide three essential qualities:
1. The process in addition to facts must take advantage of people’s knowledge while preventing the biases of experts from controlling the direction of the investigation.
Methods which allow or even encourage the specialists/analysts themselves to choose which aspects of a problem to focus their search for solutions run a strong risk of failing to identify the best solutions. Traditionally management and/or knowledgeable stakeholders are better equipped to determine which solutions are the best, so it is desirable for them to have visibility of all of the available avenues toward prevention.
A process for identifying all factors contributing to a problem so that management can consider all possible avenues toward prevention is an important feature of the observer-independent root cause analysis methods.
2. The process must depict the facts of the case so that the causal relationships are clear and the causal relevance of those facts can be verified.
Root cause analysis needs a process which validates our thinking so that we can be sure we have included all of the relevant facts, and at the same time, only the relevant facts. It is the strong dependency on facts or data that makes this form of root cause analysis observer-independent, namely it minimizes dependence on “subjectivity” in favor of “objectivity.”
3. The process must also help us understand what actions must be taken to implement potential solutions and who in the organization needs to take those actions.
Once every possible avenue toward prevention is identified, the root cause analyst must understand what specific actions need to be taken. Is a new policy needed? If a policy already exists, then why wasn’t it effective, and what steps do we need to take to make it effective in the future? And who in our organization needs to take those steps? These issues are part of the process of identifying preventative measures and must be integrated into the root cause analysis system.
Observer-Dependent Root Cause Mapping
In the case of the CIFM root cause analysis for Michigan, it was recognized from the outset that this particular root cause analysis had to be “observer-dependent” and not “observer-independent.” The distinction between observer-dependent and observer-independent is founded in the context of the evolution of science from first phase science, e.g., Newtonian physics, to second phase, e.g., quantum physics, to third phase, e.g., second order cybernetics.
Newtonian physics was deliberately constructed in the 17th century to be observer-independent, and to be invariant in terms of space and time. For example, an observer in London seeing in 2000 an apple falling from a British tree, and a different observer in Los Angeles seeing in 2004 an apple falling from an American tree, will report the same phenomenon. However, in the arena of education, where the CIFM process is meant to be applied, observers/stakeholders are observing different phenomena in terms of school performance indicators at different locations at different times, all of which are equally valid in the context of their particular situation. The challenge for the conduct of observer-dependent root cause analyses in the field of education, or other social systems for that reason, is to enable these observers to construct high quality observations collectively, collaboratively, and systemically.
It appears that this very fundamental distinction between first and third phase science is ignored in a number of root cause analyses dealing with phenomena in the social systems design arena, such as the CIFM process, leading to erroneous results. In order to distinguish this third phase science form of root cause analysis from the traditional one founded on objectivity and facts, we have named it Root Cause Mapping (RCM).
In the case of the CIFM application of the RCM, the stakeholders were first engaged in a generative dialogue in response to the following triggering question:
In response to the above question the stakeholders identified forty-six inhibitors (red ideas). They proceeded to construct, through a strategic dialogue focusing on the relationships between pairs of inhibitors, a Root Cause Map displaying graphically the four inhibitors that were located at the roots of a tree-like pattern. This group work was completed during the first day of the workshop.
On the second day the stakeholders engaged again in a generative dialogue in response to a different triggering question, namely:
“What are Action Options for overcoming the Inhibitors by focusing at the roots of the
They proposed over forty preventative action options (blue ideas) for addressing the inhibitors. Voting individually and subjectively on the relative importance of these preventative measures, they identified fifteen that received three or more votes, with thirty people voting. Ten of the fifteen important preventative actions were subsequently superimposed by the group on the Root Cause Map, by engaging the group in a strategic dialogue of relational voting in order to discover the effectiveness of the actions in addressing the root causes of the map.
The deeper a preventative action is located in the superposition map of actions onto inhibitors, the stronger is its effectiveness in overcoming the inhibitors to the successful implementation of the CIFM process in the school districts of the state of Michigan. It was discovered that only four of the important preventative actions exerted strong leverage in overcoming the root cause inhibitors. The stakeholders must give preferential consideration in implementing those four effective preventative actions.
For more details about this case study, please click here to see a power point slide show describing the agenda for the Root Cause Mapping workshop, the methodology, and the products of the group work.