The previous section illustrated some very nice mathematical results for the distribution of the sample mean when sampling from an infinite process or a finite population. However, we also saw in Section 1 that the mean may not be the parameter we are most interested in and/or may not even be possible to calculate depending on the form of the data. In this section, you will look at a possible inference procedure when the median is of interest, as well as a newer methodology that works for any statistic of your choosing.
