Box plots can be used throughout the phases of the Six Sigma methodology, but are most often found in the analysis phase. Before you can truly use a box plot, you must first understand how to do so correctly.

The first step toward using a box plot to analyze the data is to decide which characteristic you wish to examine. It must be measurable on a linear scale. This means the incremental value between the measurements must be the same. When it comes to order value boundary information such as: time, temperature, dimension and special relationships, this information can usually be measured in consistent incremental units.

If a particular characteristic being measured is consistently being produced, you will have to sample data. This, of course, gets into the more statistical testing, but it will still give you an accurate representation of the data you are viewing.

Box plots can be used to compare data, and are often shown side-by-side when this is the aim. You can even use multiple box plots to show different pieces of information or data sets. This is important because it may be necessary to break down the data so you will be able to take a closer look at trends and other pieces of information. You may need to compare various pieces of data with others and box plots will provide you the means to do just that.

When using a box plot, the data will be broken down into four quarters that are also called quartiles. You will use the box plot like a chart, but will be able to see the data represented in these four quartiles in a way that will make it easier to understand. Once you have the information you wish to analyze, and you know where it all falls within the four quartiles, you will need to find the boundaries between the quartiles. Once you have found these boundaries, you will have a clear view of where and how the information all falls into place. If a particular point is on the median, it is considered to be the last point in the quartile before, and the first in the quartile that follows.

When analyzing data, keep in mind that boxes tend to be long and thin. If there are multiple data sets present containing different numbers of data points in each set, make sure the width of the boxes correspond roughly with the relative quantity of data that is represented in each box.   Boxplots are an important tool in the Six Sigma Toolbox, and can be used to discover a wealth of information once you know how and when to use them correctly.