How to Extract Insight from Data in the DMAIC Analyze Phase
Last Updated October 14, 2020
Six Sigma provides proven tools that help organizations reduce defects and variation to make operations more effective. The Six Sigma methodology utilizes the acronym DMAIC, which stands for Define, Measure, Analyze, Improve and Control.
DMAIC gives project teams a five-phase process to follow as they move through a process improvement project. DMAIC is designed specifically to work with existing processes. The DMAIC phases are as follows:
- Define – Define the problem that needs solving
- Measure – Assess the extent of the issue and quantify it with data
- Analyze – Use a data-driven approach to find the root cause of the problem
- Improve – Put changes into place that eliminate the root cause;
- Control – Maintain the gains you’ve made with the changes
For each step, Six Sigma supplies tools to achieve milestones that mark a successful path to completion of a project. In the Define phase, the milestones or tollgates are creating a project charter, defining customer needs, and mapping the process. In the Measure phase, teams identify process performance indicators, create a data collection plan, and measure process baseline performance.
In the next phase, Analyze, teams begin to extract information from the data collected in the Measure phase.
Tollgates for the Analyze Stage
Each stage of DMAIC has a series of tollgates that mark critical milestones teams must achieve for a project to end successfully. Villanova University’s Six Sigma Green Belt course teaches students about the tollgates for each phase of DMAIC, including the following three steps for the Analyze phase.
- Find potential root causes, identify anything that might be contributing to the problem. The project team should not rule out any possible root cause.
- Narrow the root causes by identifying those most likely to contribute to the problem.
- Find critical root causes – those where reducing or eliminating them makes an immediate, noticeable difference. These critical root causes are the focus of the next phase, Improve.
The Analyze phase takes all the data collected in the Measure phase and uses it to gain insight into the root causes of problems. Students in Villanova’s Lean Six Sigma Black Belt course learn that the Analyze phase involves data analysis of inputs, processes and outputs, all in the name of finding the cause of problems.
Certain Six Sigma tools may prove especially helpful in this phase, including the following.
Cause and Effect Diagram or Fishbone Diagram
This diagram provides a way to quickly visualize how one task in an operation impacts the next. Typically, teams list a problem on the right side of the diagram, and then the team progresses to the left backward through a process, looking for potential causes. Causes typically involve the 6M’s – manpower (people), methods, measurement, mother nature (environment), materials and/or machinery.
The Five Whys
The Five Whys strategy also allows teams to find root causes of issues. Teams state a problem and then ask a series of “why” questions that dig deeper into the core of any issue. For example, the problem statement might be “We always deliver late to client X” or “We spend many hours each month fixing design flaws.” A series of why questions can unearth root causes, rather than wasting a team’s time dealing with symptoms of the problem.
This may have been done in an earlier phase, but it comes back into use now as teams track down root causes. The process map shows all the activities involved in delivering the final output (typically a product or service). It offers detail on where all the inputs enter the process and where handoffs are made from one person to the next. It also allows teams to see which inputs are controllable and which ones are not, as well as which ones add value and which ones do not. By having a visual map of an entire process, teams can quickly find operational roadblocks.
In the Analyze phase, the point is not to try out new approaches (that comes in the Improve phase). Rather, teams look to drill down and find what the true causes of variation are in a process. Regression analysis is one way of doing this, measuring the impact one variable in a process has on other variables, as well as on the final product. It can also help prove or disprove a hypothesis statement, which is an estimate based on the statistics of what is causing a problem.
If the data collected in the Measure phase is sound, the team will gain many insights using that data in the Analyze phase. They will look at every detail of a process and form hypotheses on what is causing an issue and use further analysis to prove or disprove those theories. All this leads to the Improve phase, where solutions are formed and tested.
This is an overview of tools and ideas put into play during the Analyze phase. Students in the Villanova’s Certificate in Lean Six Sigma program learn in much greater detail how the DMAIC process is handled and how to best apply Six Sigma tools to varying circumstances.