Our previous blog outlined the second phase of the DMAIC (Define, Measure, Analyse, Improve, Control) process in Six Sigma: Measure.  

In Measure, you should have collected as much data as possible to create a thoroughly detailed and holistic picture of the processes related to the problem you identified in the Define stage.

Once you’ve collected enough data, you’re ready for phase three of DMAIC: Analyse.

About ‘Analyse’

True to its name, Analyse is about analysing the data collected in Measure. It aims to identify the root cause of performance issues using a series of statistical tools and processes.

Despite being the most important part of DMAIC, Analyse is also one of the most skipped phases. The human urge to jump ahead and try to fix a problem is too strong.

“Analyse is about digging into the data and looking for patterns, which enable us to identify the root cause, and the lever we need to pull to fix the problem,” said University of South Australia’s Six Sigma lecturer Dr Neil Davidson, who has more than 15 years of Six Sigma implementation experience.

“People don’t tend to do that well. Often they’ll end up fixing the symptoms rather than the underlying problem.”

This phase is important, because it moves you from doing corrective actions to performing one preventative action – ultimately saving the business time and money.

Analyse: Key steps

1. Identify datasets to analyse

Ask yourself, “What are the variables in the process that create defects?”

Look for correlations between these variables. If you’re looking for defect rates in the manufacturing process, for example, you might look at correlations between inputs like machine adjustments, raw materials, micro process timings, operators on shift, sensor data, and the actual output.

2. Statistical analysis

In order to calculate the level of correlation between the variables (thereby helping you to assess how likely it is that the variable is causing the defect, to use the example above), you then need to perform various methods of statistical analysis. This level of analysis typically uses root cause analysis, fishbone diagrams, linear regression, and hypothesis testing.

Here’s a more detailed list of tools and templates for Six Sigma.

3. List your causes

By determining which variables have a high level of correlation with the problem you are investigating, you have effectively produced a list of ‘causes’ for variation in your process that are backed by solid data.

People involved in Analyse

This phase requires a highly skilled statistician or Six Sigma–qualified individual to perform the data analysis. The project leader is also involved. This could be a team leader or manager from the process area.

It’s important to have someone who knows the processes inside out, as well as a data analyst, as they’ll complement each other and ensure data is being analysed in the right way.

Cautions for ‘Analyse’

A common mistake people make in this stage is trying to find a solution.

“It's really important to stop people jumping the gun,” Dr Davidson explained. 

“If people start thinking solutions up in Measure or Analyse, they go down rabbit holes or get blinkers on and only follow one track. They don’t really solve the problem. They solve what they think is a convenient problem.”

To avoid this trap, it comes back to discipline. Keep the team on track by bringing them back to the goal of each stage. With regards to Analyse, it’s not about looking for solutions. it’s about analysing the data to find the causes.

Summary

“To use a medical analogy, the Measure stage is about knowing you’ve got a headache. Analyse is about finding out why. Try to resist taking a painkiller until you’ve found out the root cause of the problem – you never know, it could be a brain tumour.” – Dr Neil Davidson, University of South Australia Six Sigma lecturer and consultant.

By the end of Analyse, you should have a list of issues or root causes that need to be addressed, which, if fixed, will increase the sigma rating.

Resist the urge to think of solutions for this list of causes. These are identified in the next blog: 

DMAIC phase four: Improve. 

When a machine stops, it can quickly escalate to calling in external help – sometimes unnecessarily. The Breakdown Checklist is designed to get you back online faster. It will get your team thinking about what caused the breakdown and assess the need for external advice. Download the free downtime checklist here.

SAGE Automation delivers agile, scalable and secure solutions that don’t just solve current problems, they preempt and deter future ones, helping your organisation thrive. With years of experience working in defence, infrastructure, resources, utilities and manufacturing we have the expertise you need to custom-build or perform manufacturing maintenance on your equipment for maximum ROI.