In many of the course lectures we will discuss how you can spot bullshit, call bullshit, and avoid becoming the victim of bullshit. Here we present a set of instructional essays on various aspects of bullshit detection and refutation. Many of the examples we draw upon are classic examples that others have brought to light in their articles, essays, blogs, and other sources.
Visualization: Spotting misleading axes. Data graphics tell stories. Fairly subtle choices on the part of their creators can influence the stories they tell, sometimes in misleading fashion. We look at how the ranges shown on axes can be misleading, and explore the classic issue of when the y-axis of a graph needs to include zero.
How do you know a paper is legit? Any scientific paper can be wrong, but you greatly decrease the chances of being misled if you know how to distinguish legitimate articles from untrustworthy ones. We discuss how to draw this distinction, and along the way provide a brief overview of how the scientific publication process works.
Visualization: Proportional ink. Many data graphics, including bar charts and pie charts, use the sizes of shaded areas to represent data values. We describe what we call the principle of proportional ink: in such charts, the amount of ink used to represent a value should be directly proportional the value itself. Unfortunately, this principle is commonly violated. We explore a number of examples.
Which face is real? Recent developments in artificial intelligence have made it possible to rapidly generate photorealistic images of people who don't even exist. While these are indistinguishable from real faces at a glance, you can learn to tell the difference with just a bit of practice.