Home » Additional Features

Member Spotlight: Christopher Gotwalt

1 January 2011 3,441 views 2 Comments

Christopher Gotwalt

In a way, there is no such thing as a pure statistician. Every statistician works in at least one other discipline, whether genetics, manufacturing, agriculture, medicine, mathematics, or economics. A serendipitous sequence of events brought me to statistical computing and my position at SAS, where I’m the statistical applications manager for the JMP division.

I grew up in the Tampa Bay area and graduated from the University of Florida with a major in mathematics and a minor in physics. Like most college students, I spent a lot of time thinking about what kind of work I wanted to pursue after college. Because of my passion for mathematics and science, the many potential career paths I considered included electrical engineering, actuarial science, and teaching.

In my last year at UF, I took a couple of statistics classes and became excited about the prospect of a career in statistics because it is an inherently interdisciplinary mathematical discipline that offers many opportunities in industry. I decided that statistics was the career for me, and—after a whirlwind weekend in which I graduated from college on Saturday, got married on Sunday, and moved to Raleigh, North Carolina, on Monday—I began graduate studies in statistics at North Carolina State University on Tuesday.

Early in my graduate career, it wasn’t obvious what my secondary discipline would be. During my first year at NC State, I had a Hewlett-Packard calculator that came equipped with an LISP programming language
interpreter. I found myself spending hours programming it to do statistical calculations and replicating results SAS reported in my applied statistics class assignments.

One day in a computing lab class, the professor pointed out that a relatively recent graduate of NC State had developed the SAS procedure we were learning that day. A friend sitting next to me turned and said, “That’s what you should do, Chris! The thing you’re the most interested in is statistical computing, and SAS is just down the road from here.” He was right, and computer science became my second discipline. With the support and guidance of my PhD adviser, Dennis Boos, I began pursuing a career in statistical computing by taking steps in that direction via courses, internships, and doctoral research.

In 2001, I began an internship developing statistical software in C++ for JMP. That internship turned into a full-time position when I graduated in 2003. One of the things I love about working for JMP is that it is like a little company inside a big company and you get the best of both worlds. SAS is widely recognized as one of the best companies to work for in the country. We just celebrated being ranked No. 1 in that category by Fortune magazine. The JMP division, which develops interactive data visualization and analysis software, has a relatively small development staff, and I enjoy the opportunity to wear many hats and work in a variety of statistical areas.

Early in my career, I developed software tools for item analysis, a type of psychometric ability testing. Later on, I developed JMP platforms that analyze data using linear mixed models and generalized linear models. In 2006, I was developing tools for analyzing data from computer simulation experiments using Gaussian process models. More recently, I’ve led an initiative for JMP to expand its support of interactive modeling tools for the analysis of engineering reliability data.

Developing algorithms for optimal design of experiments has provided some of the most interesting and challenging work of my career so far. Optimal design boils down to an optimization problem with combinations of categorical and continuous inputs in which there can be various constraints on the input factors. I’ve enjoyed cutting my teeth on putting together numerical tools that solve that problem efficiently. Another area of particular interest has been design for nonlinear models, where it is necessary to take on a Bayesian approach that incorporates prior information on the model and parameters.

I took on management responsibilities in the summer of 2008, and now I lead the talented team that makes up the statistical software development group for JMP. I’m still what you would call a working manager. Currently, I’m developing software tools for data mining using neural networks and working on algorithms for Bayesian optimal design of accelerated life tests, a type of engineering reliability experiment. These differ from classical design problems because of the nonlinearity introduced by working with non-Gaussian distributions and the fact that you have to plan for the finite length of the study, which inevitably causes many of the experimental observations to be censored. I spend time interacting with our customers, both collecting their feedback and suggestions and occasionally providing light statistical consulting.

Outside of work, my wife, Jessica, and I are gradually renovating our 135-year-old house in downtown Raleigh, which has been an adventure in itself. In the evenings when I have time, I pick away at ragtime piano pieces and have spent some time learning transcriptions of old Jelly Roll Morton jazz piano rolls from the early 20th century. Last year, I got hooked on hot yoga, which I practice several times a week. I’ve found it to be a great way to stay in shape and reduce stress.

Working in the statistical software industry is challenging and rewarding. I am extremely honored to contribute to the software tools that companies use to improve the products we use every day, from the processors in our laptops to the drugs we take to keep us healthy.

1 Star2 Stars3 Stars4 Stars5 Stars (3 votes, average: 3.67 out of 5)
Loading...

2 Comments »

  • Anton Rasmussen said:

    Great piece… very inspiring for a soon-to-be statistics graduate student. Thank you!

  • JMP® Blog said:

    Amstat News Shines Spotlight on JMP Statistician…

    Christopher Gotwalt, the developer behind many innovative JMP software tools for data visualization and analysis, is featured in the Jan. 1 Member Spotlight column of Amstat News, the magazine of the American Statistical Association. Chris, the Statist…