Award-winning paper by John Sall shows how plots and graphs help scale-up process characterization

[#1]

Over the years, I’ve modified my JMP elevator speech – you know, the summary of your product that you can say in the brief span of an elevator ride.

Here is my latest: JMP is computer software that people use to understand information and patterns about that information. It analyzes information patterns, might predict future patterns, and immediately displays graphs that help you understand, in a glance, what it found.

So, as I was exploring options for a new graphic for our Mastering JMP webinars, I chose this rendering of a JMP graph, because it shows, in a way that almost anyone could appreciate, the pattern for the result of a prediction. It shows:

• A regression analysis (the diagonal blue line)
• Each category of information (triangle, circle x, diamond)
• A graphical representation of the uncertainty about the regression analysis results (the shaded, arched lighter blue bands around the diagonal blue line; they help us see the range where the diagonal blue line might actually be)

So, without deciphering any tables of statistics, from this simple graphic, anyone could see where the observations are for each of the categories (triangle, circle x, diamond). And, by looking the diagonal line's increase toward the upper right, they could see where the next observation would fall.

Over the past 35 years he has been developing JMP software, John Sall, SAS Executive Vice President, has consistently emphasized the importance of graphs and plots for showing and interpreting statistics and statistical results.

Recently, John specifically made the case for the importance of using effective summaries and graphs to portray the behavior of processes. For his paper, Scaling-up process characterization, he was awarded the 2019 Søren Bisgaard Award. The award is given annually for the article published in Quality Engineering that is considered to be of greatest potential for advancing the practice of quality improvement.

In the summary for his paper, John wrote:

“The goal of statistical analysis is to produce reports to effectively and efficiently portray the behavior of processes in context. When there are many processes to portray, effective summaries and graphs that give a fair picture of process health are needed. Engineers want to start with the overall picture and then dig into the details as justified by the summaries. Engineers want good default analyses so that they do not have to spend time making decisions about the methods and details to use in the analysis. A good report communicates an understandable portrait of what the data say.”

We thank John Sall, the American Society for Quality, Quality Engineering and Taylor and Francis for allowing us to share this paper. We hope you will read it and see how John describes and gives examples of using JMP graphs and reports to characterize process conformance, changes and stability.

Scaling-up process characterization is an Accepted Manuscript of an article published by Taylor & Francis in Quality Engineering, Volume 30, 2018, Issue 1: Special Issue on the Fifth Stu Hunter Research Conference on December 7, 2017. Available online:  https://doi.org/10.1080/08982112.2017.1361539