Random Encounters (RE) is a dynamic random number generator, cloaked in an RPG-style interface. It was created to improve education in statistics and experimental design the following three ways: (1) Rewarding students for applying new methods in creative or exploratory ways. Rather than demonstrating knowledge of statistical methods on a multiple-choice exam with pre-specified hypotheses and artificial data, students are taught to infer useful relationships in the virtual world. These inferences demonstrate knowledge to the educator, but more importantly, they demonstrate the usefulness of statistical methods to the student. (2) Providing students access to “big data,” and a personally significant context for interpreting results. Rather than practicing new skills on unrealistic data with arbitrarily named variables, students collect their own data, based on their own interests. This makes the numbers relatable. Furthermore, the data are exported in a format requiring organization before the test of inference. This allows students to practice an essential part of data analysis, which is almost always excluded from introductory courses. (3) Practice communicating quantitative results broadly, and to peers with shared interests. Rather than taking multiple-choice exams for course credit, students gain “peer recognition” by contributing to a growing body of evidence called The Royal Library. Contributions are simple-- an abstract stating the research question, null hypothesis, and non-technical conclusion; a description of the methods for collecting data and testing the null; a figure with caption that clearly supports the non-technical conclusion.
