These examples of a diverse set of research problems illustrate that one way or another, it is usually possible to use random assignment to evaluate educational programs. There is no one formula for randomization, but with enough resources and cooperation from policymakers, random assignment is possible.