“Unfortunately, too many people like to do their statistical work as they say their prayers – merely substitute in a formula found in a highly respected book written a long time ago” (Hotelling et al., Annals of Mathematical Statistics 19:95).Although this statement is now 60 years old, it still holds true. The teaching goal of this course, therefore, is to provide a deeper understanding of statistical methods, which can only be achieved by teaching students some of the basics of probability theory.We will start out with descriptive statistics, including correlation, regression, and effect size measures. Touching upon probability theory, we will then delve into the realms of inferential statistics. Rather than covering two dozen different hypothesis tests , we will exploit the fact that all these tests follow a common logic ("null-hypothesis significance testing (NHST)"). This logic can best be outlined using the binomial test, which is the test most amenable to students with limited knowledge of probability theory. Building on this foundation, we explain the usage and pitfalls of some of the most common forms of NHST (z-test, t-test, analysis of variance, chi square test, nonparametric tests), with a special emphasis on statistical power. In addition, we encourage students to approach their data with an exploratory attitude as outlined in Tukey’s seminal work on Exploratory Data Analysis. This is important because hypothesis tests yield highly processed data (p-values) that are prone to misinterpretation.

Students are expected to bring their own laptops running copies of either Microsoft Excel or Open Office Calc, which will be used for some exercises.