The inspection paradox is a statistical illusion you’ve probably never heard of. It’s a common source of confusion, an occasional cause of error, and an opportunity for clever experimental design. And once you know about it, you see it everywhere.
When you drive on a highway, have you ever thought that everyone drives too fast or too slow, and no one else is a safe, reasonable driver like yourself? If so, you have been fooled by the inspection paradox, an often subtle form of biased sampling.
In this talk, I explain the inspection paradox with examples from social networks, transportation, education, incarceration, and other domains. Using simple Python code, I show how the effect can be quantified and how, with clever experimental design, you can take advantage of it.
This talk is accessible for people who know basic Python and minimal statistics.