Reflections on CSP: Why I Love the Conference on Statistical Practice

Eric Vance, director of the Laboratory for Interdisciplinary Statistical Analysis, University of Colorado-Boulder, and chair of the 2019 CSP Steering Committee

The 8th annual Conference on Statistical Practice (CSP) was held in New Orleans, Louisiana, February 14–16. There were 483 attendees, including me. Along with several other people, I have attended all eight CSPs. Why? Because I love the Conference on Statistical Practice. What in particular do I love about it? I love many things about CSP; I’ll discuss three of them here.

In all four themes of the conference and throughout all the short courses, concurrent session talks, posters, tutorials, and practical computing demonstrations, the focus is on the practice of statistics and data science. The Data Modeling and Analysis theme highlights techniques and best practices for—you guessed it—modeling and analyzing data. The Data Science and Big Data theme helps attendees stay current with state-of-the-art methods for solving problems with extremely large, unconventional, or complex data. The Software, Programming, and Data Visualization theme helps attendees integrate statistical software into their current processes. And my favorite theme—Communication, Collaboration, and Career Development—helps attendees develop professional skills to increase their impact and advance in their careers.

Second, CSP is the right size for face-to-face interactions. Acquaintances met during the first day of short courses can become lifelong friends and colleagues by the closing session. The CSP Mentoring Program matches beginning statisticians and data scientists with those with more experience, providing a two-way avenue to discuss issues beneficial for both mentee and mentor. Themed group dinners provide social opportunities organized around a statistical topic. Planned evening outings can be the highlight of the entire conference. This year, attendees could choose between joining a ghost tour of New Orleans or packing into a jazz hall on the first night of the conference and then watching a Mardi Gras parade through New Orleans’ historic French Quarter on the second night.

My favorite part of CSP 2019 was the keynote address. As chair of the CSP Steering Committee, I had the privilege of choosing the keynote speaker. Out of everyone I could have chosen, I asked Hadley Wickham, chief scientist at RStudio, because he is developing tools that help statisticians and data scientists do smart and relevant things with data. His R packages in the tidyverse make importing, tidying, exploring, visualizing, summarizing, and modeling data easier, faster, and more fun. His keynote, “What Is Data Science?” provided attendees with practical tools and tips to improve their practice of statistics and data science and become more impactful in their careers.