So far, we’ve investigated statistical inferences assuming that the underlying distribution of the data is known. Furthermore, we’ve conducted statistical tests by estimating certain parameters of the underlying distribution. For example, if we assume that our sample(s) is (are) being drawn from an underlying normal distribution, then we estimate the population mean of group i, μi, by taking the sample mean, xi. Since we have to estimate underlying population parameters from our data, we can refer to the overall class of statistical methods as parametric methods. When we do not assume any underlying distributional form (and, therefore, do not estimate any population parameters), we refer to nonparametric or distribution-free statistical methods. To begin our discussion, it is worthwhile to review some basic data types.