Carillon Technologies Limited

 Pareto Analysis

Pareto Principle

The principle, also known as the 80/20 rule, was named by the statistician J. M. Juran in the late 1940's after Vilfredo Pareto, an Italian economist (1848-1923).

 

The principle reflects the notion of "vital few" versus the "trivial many."

The principle holds true in industry, distribution of wealth, biology, and many other natural phenomena.

The principle holds that 80% of the effect is caused by 20% of the causes.

  

What does a Pareto Chart Look Like?

This Chart is an example of scrap causes for an automotive Instrument Panel

  

STEPS FOR DOING A PARETO ANALYSIS

 

Step #1

Select a specific product to work on and write down all the types of defects you can think of.
List these possible defects on your check sheet.
  • Problem: Defective corn flakes.
  • Possible defects:
  • Broken
  • Cracked
  • Burnt
  • Undercooked
  • Chipped
  • Undersize
  • Oversize

    • Step #2

    Collect data that shows how frequently each defect occurs (see the section on data collection check sheets).

    On the problem items check sheet, record the number of defects found.

    Inspection Method: Scoop a cup of corn flakes before they are packaged and count the number of times each type of defect occurs.

    Step #3a

    On the Pareto chart work sheet, re-list the defects in order of frequency. Start with the one that occurs most frequently and end with the least frequent one. The final list is shown below:

     

     Defect

     Frequency

     Chipped

     45

     Broken

     30

     Burnt

     10

     Undersize

     9

     Undercooked

     7

     Cracked

     3

     Oversize

      1

     Total =

     105

    		  

    Step #4

    Calculate the percentage each defect is of the total and list them on the work sheet. Use this formula to make the calculation:

    For example, for Chipped:

    Record these percentages on the Pareto chart work sheet.  

    Step #5

    Next, determine the cumulative percent. To do this add the individual percent for each defect to the percents of all the defects above it.

    For example, the cumulative percent for chipped is 42.9% (the same as the percent, since this is first on the list). The cumulative percent for broken is 71.5% (found by adding 42.9% for chipped plus 28.6% for broken).

    This cumulative number shows the importance of the top few defects versus all the others.

    Now complete the cumulative percent column on the work sheet.

    Step #6

    From Step #5 you can see that the two main discrepancies are chipped and broken flakes. These two defects account for 71.5% of the total bad corn flakes.

    You should also complete the Pareto chart at the bottom of the work sheet.

    This chart can be an important tool to use when you present your findings to others.

    The next few steps show how to complete this chart.

    Step #7

    On the chart draw a bar graph showing the individual percent (not the cumulative percent) of the problem for each defect.

    Use the letters on the side of the work sheet and the bottom of the chart to keep track of which defects goes with which bar.

    For example, over the "A" on this chart draw a bar that has a top at 42.9%. This represents the defect chipped.

    Over the "B" draw a bar up to 28.6% for the broken, and so on for the rest.

    		  

    Step #8

    You can now chart the cumulative percent. To do this, plot the cumulative percent for each defect along the right hand edge of the column above the defect.

    For example, on the right-hand edge of the column above A put a point at 42.9%. On the right edge of the column above B put a point at 71.5%, and so on for the rest of the defects.

    		  

    Step #9

      The final, and most important, step in a Pareto analysis is to take corrective action. You have identified the two most important defects chipped and broken flakes.
      You must now find the root causes for these defects and eliminate or at least reduce the number of occurrences.

     

    Pareto Summary

     

    Pareto and quality costs indicate which problems to investigate, but do not indicate the cause of the problems.

    The discovery of the problem cause, and the development of the best solutions for that problem, are obtained through statistical analysis of data.

    Return to Charts Page