Question: For numbers 5 and 6, you will be using the Watershed dataset Watershed0.xlsx (specifically, you will need conpolut, region, and advisory). You will need to recode regions into East and Midwest. This is similar to how we recoded ethnicity into white and non-white. The simplest way to do it is to sort the data by region and then make a new variable that is east if the watershed is in the Northeast or Southeast and Midwest if the watershed is in the Midwest. You are going to replicate part of the analysis in the text (see pages 188-190). First, analyze the normality of the variable Conpolut and compare that to the normality of the square root of Conpolut. To get the square root of Conpolut, you can use the square root function in Excel [=sqrt(CELL WITH CONPOLUT)]. To compare the normality, use the histogram tool in the Data Analysis Toolpak. For the rest of the question, you will use the square root of Conpolut because its distribution is closer to normal. To check whether the variances are equal, use the Excel tool discussed in the lecture. Then replicate the t-test for comparing the pollution levels of the East with those of the Midwest. Include a write up describing what you would conclude based on the test and the difference in means.