The MAGIC criteria are a set of guidelines put forth by Robert Abelson in his book Statistics as Principled Argument. In this book he posits that the goal of statistical analysis should be to make compelling claims about the world[1] and he presents the MAGIC criteria as a way to do that.