The Truth Within: How to Interpret Data and Mislead with Statistics

Our world is full of data. This data is constantly being digested and presented to us through recommendation algorithms, feeds, headlines, and the news. Doing good science requires careful data analysis, and doing this poorly can let us draw ridiculous, erroneous, or misleading conclusions. In this semilab, we will uncover how to use probability to separate fact from fiction, how to read the news with a critical eye, and how various statistical effects can obscure the truth and lead a careless scientist down wrong paths. We will cover Bayes' Theorem, Simpson's Paradox, how to evaluate predictions and correlations, and other surprises in probability.