I just read a compelling column by Davis Balestracci in Quality Digest. It offers a great example of how using statistical software to get a graphical view of your data can help you accelerate the process of improving quality.
The column talks about the benefits of graphing process output over time—in other words, creating a run chart. Here are a few key paragraphs from his conclusion:
...merely plotting a process’s output over time is one of the most simple, elegant, and awesome tools for gaining deep understanding of any situation. Even before plotting, one must ask questions, clarify objectives, contemplate action, and review current use of the data.
Questioning from this statistical-thinking perspective leads immediately to unexpected and deeper understanding of the process. The end results will also be valuable baselines for key processes and honest dialogue to determine meaningful goals and action.
Got that? He's rightly pointing out that often the benefits of doing statistical analysis begin before you even collect any data, let alone do any calculations: just considering what data you need to gather starts you thinking in terms that will help you better understand and improve what you're doing! That's pretty cool.
What Is a Run Chart?
The term "run chart" may sound techy, but it's really very simple: it just shows you a graphic representation of your process data, arranged sequentially by time. You can use run charts to look for evidence of special-cause variation in your process. Special-cause variation refers to outcomes that vary more than we would normally expect. This type of variation causes recognizable patterns or shifts in the data that you can easily detect on a run chart. Identifying and eliminating special-cause variation keeps a process in control.
When you use statistical software like Minitab to create a run chart, it plots your individual observations in the order you collected them, and draws a horizontal reference line at the median. In Minitab, creating run charts also performs two tests for randomness that give you information on non-random variation. When you see trends, mixtures, oscillation, and clustering patterns, it's a pretty good indication that you're seeing some special-cause variation.
And once you've found it, you can do something about it.
In my next post, I'll lead you through the creation of a run chart from start to finish in Minitab.