The subjective root cause analysis tools are used to systematically identify and subjectively weigh the correlation of the inputs (x’s) to the problem, Y.
Determine the most important known and hidden root causes (when data is not necessarily available) to analyze their importance and relationships in the ANALYZE phase. These tools take all the controllable inputs (causes), and filter them down to the vital few that have the most impact on closing the gap written in the project contract.
Use these tools in this order, if there are few x’s then the Correlation Matrix can be skipped. Each tool narrows the focus to the primary inputs, x’s, to the problem. The primary inputs, x’s, are referred to as the Key Process Input Variables, KPIV’s.
1) FISHBONE – includes all inputs, x1, x2, x3, x4, ……..xn.
2) CORRELATION MATRIX – also called Cause/Effect Matrix and Prioritization Matrix
3) FMEA – Failure Mode and Effects Analysis
There are often times times when not all data can be gathered for a number of reasons, the data may be flawed, or there are hidden causes that need to be discussed within the team. Dialogue and knowledge sharing between the team members helps understand and weigh the impact root causes have on the output.
One helpful and simple procedure to help a team drill down to the root cause is the 5-WHY tool. This is used to drill down and reveal the root causes of unplanned downtime, form errors, injuries and accidents, quality defects, late deliveries, and more.
There are several methods to help find the root causes of variation. These are more traditional tools that are commonly used in problem solving techniques due to the ability to break complex problems into small scopes and make them manageable so everyone can participate.
The goal is to capture all of the inputs, big or small. Narrow them down throughout the subjective tools and scoring system with cross functional representation and then take the few key inputs (KPIV's) and begin to ANALYZE them.
Problem solving techniques, such as A3, include subjective root cause tools, process mapping and value stream mapping, and most importantly emphasize clear understanding of the problem, gap, corrective action, and accountability to control.
A3 is a key component of Toyota problem solving process that applies to all levels of the workforce. The Plan-Do-Check-Act (PDCA) cycle is fully encompassed in this simple tool. The entire scope of the cycle from problem statement all the way to completed improvements are hand written on one large sheet of paper.
The subjective tools are invaluable since they bring out dialogue and information that isn't necessary found in the data. The team members and stakeholders closest to the process know the details that are behind data.....and they know what the data isn't saying.
Your job, as a Six Sigma Project Manager, is to mine and correctly interpret the data and show the team the results. They will provide the practical (reality) discussion around it.
However, this dialogue and interaction between team members can make it very time consuming to finish these tools. And you should not complete them yourself. That could lead to incorrect assessments and lose the trust of your team members.
The dialogue between team members can also lead to some arguments and disagreements on scores. You'll need to funnel that energy into positive momentum and be prepared to be the tie-breaker.
Be careful determining the quantity of x's and y's that are analyzed in the Correlation Matrix. This can make a significant difference in the amount of time it takes to complete this tool.
These subjective tools become even more important if there is a lack of data or poor data that simply can not be used. Just be prepared to move quickly and be a referee. Keep the discussion limited without being too restrictive. Take breaks, incorporate some humor, but most importantly get it done.
Six Sigma Modules
The following presentations are available to download
Green Belt Program 1,000+ Slides
Cause & Effect Matrix
Central Limit Theorem
1-Way Anova Test
Correlation and Regression