Statistical Intuition without Coding (or Teachers)

Abstract Teaching political methodology classes typically requires a set of technical, instructor-led lectures on sophisticated statistical concepts (such as probability modeling, inference, and proper interpretation), followed by chances for students to iteratively adjust methodological specifications, parameters, and data sets so they can understand how each combination affects the results. Iteration is essential to learning complicated concepts, but until now has required simultaneous mastery of a statistical programming language (such as R), which makes learning both harder. Teaching R the semester before would make methods classes easier but also delay research experiences and demotivate our students eager to begin substantive research. We address both problems through a new type of interactive teaching tool that lets students iterate while learning the big conceptual picture and all its separate parts, without having to simultaneously become programmers. We make this tool available for use in classes now (via one click in a web browser) and as an example of a new type of more friendly methods instruction for students and instructors alike.