Online Program

Return to main conference page
Wednesday, September 25
Wed, Sep 25, 10:45 AM - 12:00 PM
Thurgood Marshall North
How Do We See Data Today? Advanced Data Visualization and Analytical Techniques

Animating Your Data (301046)

*Zachary Skrivanek, Eli Lilly and Company 

Pharmaceutical companies need to be able to effectively communicate the results of clinical trials. Visual communication is the most effective way to tell a compelling story and statisticians can play a role in this. They need to make sure that the visualization is accurate, not misleading, but still impactful and memorable. Typically, static graphics of clinical results consist of longitudinal line plots of the statistically adjusted mean for continuous data, Kaplan-Meijer plots for time to event data and bar charts for binary data. These graphics do not give any insight into the patient level response, not letting the reader know if the effect was sustained or if the effect is oscillating and only the average response is maintained over time. Furthermore, they are just plain boring and uninteresting. With conventional graphical methods, statisticians typically encode data to location, color, shape and size. But with modern technology we can easily encode to time as well. Not only does this answer some of the shortcomings in conventional methods mentioned earlier, but it can provide a much more compelling and transparent story. I will present some examples of using interpolation to implement Frame by Frame animation of clinical data. The examples will be based on R/Shiny and Javascript language.