A multi-vari chart shows both several sources of variation in addition to the most significant contributors to total variation.
When the output has a variable measurement.
When attempting to identify the biggest contributors to total variation.
When looking at sources of variation within a process.
1. IDENTIFY THE POSSIBLE SOURCES OF VARIATION. Construct a sampling tree with combinations of settings for three sources: A, B, and C.
2. CREATE A GRAPH. The y-axis should represent the output. Separate the x-axis in the following manner: create a section representing each setting of the A source (this should be the top level of the sampling tree). Divide each section further into subsections representing each setting of the B source. Plot points in a vertical line to represent the measurements at various settings of the C source.
3. CALCULATE THE MEAN of the values in the first line of points. Indicate this value on the line using another symbol. Repeat this process for each B group in the chart's first section.
4. CONNECT THE MEANS of the B groups that were calculated in the previous step.
5. CALCULATE THE MEAN of all values in the first section and MARK the value at the midline of the section using a third symbol.
6. REPEAT STEPS 3, 4, AND 5 for each section of the chart representing each setting of source A.
7. CONNECT THE SYMBOLS for each section's overall mean.
8. ANALYZE THE CHART. This may reveal patterns of variation in addition to the greatest contributors to the total variation.
Useful for identifying variation and factors contribution to the variation.
Requires some statistical experience.
Tague N. The tools. In: O'Mara P, editor. The quality toolbox. 2nd ed. Milwaukee, WI: ASQ Quality Press; 2005. p. 93-521.