The course aims to advance these skills with an emphasis on producing publication-ready figures, including high-resolution plots that use color-blind-friendly palettes, to ensure that data-driven evidence is communicated as clearly and inclusively as possible. (7/n)

We might ask whether we still need data visualization training when we have powerful LLMs to help us. Certainly, LLMs can help to optimize our code. But without a profound understanding of the produced code, we run the risk of creating figures that may look nice but that misrepresent our data. (6/n)

In order to create effectively communicate data, we need to understand the inherent biases and dynamics of human visual perceptions. These insights on the fundamentals of data visualization are broadly applicable beyond the field of (social) science. (5/n)

As put by Kieran Healy: "Data visualization is not simply a matter of competing standards of good taste. Some approaches work better for reasons that have less to do with one’s sense of what looks good and more to do with how human visual perception works." (4/n)

Apart from providing code for various figures including cleveland plots, ridgeline plots, or dynamic plots, the class seeks to provide a theoretical foundation. Effective data visualization requires a profound understanding of the dynamics of human visual perception. (3/n)

The course will introduce a variety of data visualization techniques including model-based visualizations, focusing on the effective presentation of empirical results through coefficient plots, simulated expected value plots, and event study plots. (2/n)