Get Your Data Collection Right

Posted by on Apr 30, 2012 in Lean Six Sigma | 0 comments

Data can be either qualitative or quantitative.  Qualitative data characterize things that are sorted by type, such as fruit (oranges, pears, apples, …), defects (scratches, burrs, dents, …), or operators (Jim, Harry, Martha, …).  Qualitative data are usually summarized by counting the number of occurrences of each type of event.  Quantitative data, on the other hand, characterizes things by size which requires a system of measurement.  Examples of quantitative data are length, time, and weight.

There are 5 steps in the data collection process:

1.  Clarify your data-collection goals.  Make sure the data you collect will give you the answers you      need.  You can use these questions to help you identify your data-collection goals:

  • Why are you collecting data?
  • What questions do you want to answer?
  • What will you do with the data?
  • How will the data help you?
  • What patterns or relationships might you want to explore?

Consider how you can possibly stratify the data.  Think of the following categories:

  • Who – which people, groups, departments, etc. are involved
  • What – relevant machines, equipment, products, services, etc.
  • Where – the physical location of the defect or problem
  • When – time of day, day of week, or step of the process involved
2.  Develop operational definitions and procedures.  Operational definitions tell you how you should measure something.  They help you specify who should collect the data using what instrument, what data-collection form, and whether they should measure every item or just a sample of items.
3.  Validate the measurement system.  Conduct a Measurement System Analysis to make sure your measurement system is capable.
4.  Begin data collection.  This involves training the data collectors, piloting and error-proofing the data-collection process, and deciding how you will display your data.
5.  Continue improving your measurement consistency.  Check to make sure that everyone is following the data-collection procedures.
Remember, the results you get and the decisions you make depend on the accuracy and integrity of your data.  Make sure you take the time to plan your data collection process and validate your measurement system.

Leave a Reply

Your email address will not be published.