Celebrating JMP Superusers: Michael Anderson, Corning
Feb 7, 2020 12:59 PM
Michael Anderson, Senior Statistical & Metrology Development Engineer, Corning
Fun fact: I have been a runner for 32 years, but my first race ever was in 2012 at age 46! Currently, I log more than 2,500 miles per year.
JMP User since 1991
"When JMP arrived, it was a revelation," says superuser Michael Anderson of Corning.
What is your favorite feature in JMP?
That is like asking which child is my favorite! The reality is I use JMP in all phases of my work, so I have a favorite screening feature, experimental design feature and so on. It is funny, I was very late to the Graph Builder party and still use the Variability Chart quite often. Now Graph Builder is where I spend my time as soon as I get a new data set. On the experiment side, I would say that most of my experiments have been Definitive Screening Designs, as they are incredibly powerful and efficient for a development organization.
What was your first job (ever)?
There are a couple of ways to answer that question. As a young person (before “officially” working at age 16), I had a paper route and lugged trays of sodas up and down the stairs at what is now Patriots Stadium. The first normal scheduled job I had was as a store clerk at a long defunct New England chain called “Bradlees”. I was given a lot of responsibility and learned a ton about business operations, inventory and people!
What is your proudest professional moment?
In ~2000 I was asked to give a talk at what turned out to be a one-time measurement technology conference. I was able to talk about the new field of spectroscopic scatterometry. That stands out for being the culmination of a lot of hard work in an area that was both theoretical and practical. I was using my engineering, optics and statistics skills in a way that would bring new capabilities to the industry.
How did you get interested in statistics?
I took all of the available statistics classes during my engineering undergrad and really enjoyed them. But, it was during my first job discussing precision resistors with another engineer who didn’t understand distributions, truncation, segmentation and the calculation of likelihood that got me really energized. It became clear I wanted to further pursue statistics graduate classes and statistics as part of my career.
What do you like most about the work you do?
I love that I get to share my love of data analysis with people at all levels. I also enjoy getting to mentor younger engineers just starting out in statistics and JMP.
How are you currently using JMP?
I have JMP open all day, every day, and utilize the graphing, DOE, screening and modeling features extensively. In our product and process development area, we are constantly interacting with research and production groups. That means we do small sample analysis, exploratory experiments and use large and wide data sets. Functional Data Explorer is the most recent area that has seen a great increase in usage.
What was the first project you worked on using JMP?
I was new at a mid-sized company and was leading a high-visibility project and knew my reputation would be made from this one way or the other. I had recently started using JMP, and used it to look at some preliminary data and run a simple factorial experiment. Within two weeks, I was able to present, using three graphs, results that showed the problem and suggested a solution in a way that had been missed for nearly two years. The power of JMP and everything it enabled was firmly established for me and many others in that moment.
Is there anything you would like to say to JMP development or John Sall?
I would like to thank John Sall and all the incredible developers behind JMP for their vision and dedication. It is hard to think of what things were like in 1989 as an engineer and statistician. There were software packages that could do analysis, others that could do presentation graphics, and still others for specialized quality features. But each had significant downsides and learning curves. When JMP arrived, it was a revelation. How quick and easy it made getting past the early stages of analysis of our data. I was presenting at a conference in 1994 and using a Fit Y by X graph with some added fitting items. Afterwards someone asked if I had used JMP! We then got into a discussion about how much we loved JMP and how it had changed so much of our workflow from the moment we’d started using it.
How do you see the field of data science/analytics progressing in the next 20-50 years?
The death of human-based statistical analysis is greatly exaggerated. I was educated right at the end of learning to use a slide rule. I learned how to calculate ANOVA, regression, matrix math and so on by hand. At the time, there were some who bemoaned the use of computers to do these things, saying it would replace thinking statisticians with cold machines. Now we hear that AI and ML will obsolete statisticians. I simply don't believe it. With each new leap in technology, we have seen the ability of people to make better use of data being a key to learning and innovation.