On Wednesday, we talked about sample bias, or ways to really screw up the results of a survey or study. So how can researchers avoid this problem? By being random. There are several kinds of samples from simple random samples to
That’s So Random: Getting sampling right
