In this unit you will gain a stronger understanding of how science works, become a more sophisticated consumer of research, and likely become a better researcher. Science is typically done not to benefit a small group of people, but rather to add to the knowledge base of the world. This ethos supports practices that, if followed, lead to robust, reproducible results. A reproducible study is one for which the methods are described in enough detail for others to follow, the analyses of the data are straightforward to re-run, and conducting the study again (where possible) yields results that support the claim of the original report. In areas of science where reproducibility has been evaluated, such as cancer biology and experimental psychology, replication success rates have generally been lower than 50%, and in clinical medicine, published outcomes show substantial biases. In this unit, examples are drawn from biology, health science, social science, education, and engineering to illustrate the practices that lead to reproducible science, as well as the underlying principles. Building on your basic knowledge of statistics, you will learn about statistical issues that lead to irreproducible science, such as poor statistical power and questionable research practices. You will also become familiar with how to apply practices that foster robust results, including practices that are increasingly expected by journals and research funders.