Video-on-Demand Library

13 January 2021

Bodo Kirsch presents ways to explain the possible impact of risk factors on a binary outcome with means of graphical displays and tools. Very different approaches have been sent in to solve the challenge of last month.

To get an overview of the data histograms and box and whisker plots are helpful. And as multiplots grouped by outcome the can give first hints on dependencies, but may be misleading. Different aspects of the dependency structure can be visualised with a heatmap correlation plot or a log-odds plot. A very nice idea was a probability plot including the individual data points as well as the predictive value within its confidence interval. The classification tree was presented as a very powerful display to visualise the impact of the predictors. Next was a nomogram for manual use followed by its interactive counterpart including interactions. Last but not least a presentation-ready break-down plot for a logistic regression model was shown as an example suitable for “non-technical” audience, too. The new challenge was introduced and will be about visualisation of a multidimensional quality of life outcome.

Wonderful Wednesdays are brought to you by the Visualisation SIG. The Wonderful Wednesday team includes: Bodo Kirsch, Alexander Schacht, Mark Baillie, Daniel Saure, Zachary Skrivanek, Lorenz Uhlmann, Rachel Phillips, Markus Vogler, David Carr, Steve Mallett, Abi Williams, Julia Igel, Gakava Lovemore.

Part of Collection(s)

Topic(s)

Upcoming Events

Latest Jobs