1)Determine if there is a difference in the mean high school graduation rate (PercentHSGrad) between urban and nonurban counties.

2)Determine if there is a difference in the proportion of urban and nonurban counties who have Low Lottery Sales.Use SAS code that includes the word "riskdiff" in some way.

3)Determine the linear model for predicting the population of a county if the median income was known.

4)For the model created in Question 3, determine the predicted population of a county with a median income of $20,000, $50,000, and $100,000.Find the 95% confidence interval for the mean of each of these, as well as the 95% confidence interval for the value of each of these.

5)Find the correlation coefficient between Bankruptcy Rate and Unemployment Rate for 2009.Determine the linear model that would be used to predict the Bankruptcy Rate if the Unemployment Rate were known.

6)2x2 Contingency Table between LottSales and ViolentCrime and run Chi-Square, Odds and Relative Risk tests.Use LottSales as your Rows.

7)Test the hypothesis that the mean of the percent of residents who registered as Democrat (PercentRegDem) is less than 40% at a 2% level of significance.

8)Create the 99% confidence interval for the mean of the Percent of residents who are on Disability (PctonDisability).

9)Create the 95% Confidence Interval for the percent of counties that are considered "URBAN". Test the hypothesis that this percent is not equal to 50% at a 5% level of significance.Find the power of this test.

STAT 3120 Final Exam Fall, 2017 PLEASE NOTE THAT WHILE THIS IS A TAKE HOME EXAM, AND YOU ARE FREE TO USE YOUR NOTES AND THE INTERNET, YOU CANNOT DISCUSS THE EXAM WITH ANY OTHER PERSONS - NOT JUST THOSE IN THE CLASS. This exam will utilize the \"homeless1\" dataset. After you import the data, please use the following code to create three categorical variables which will be needed for the analysis: Data homeless1; set homeless1; if population = 425 then Lottsales = "HIGH"; else Lottsales = "LOW"; if violentcrimerate = 300 then ViolentCrime = "HIGH"; else ViolentCrime = "LOW"; Run; After you run this code, I suggest that you run the descriptive statistics, visualizations on these new variables to ensure that you understand them. FOR EACH QUESTION, PLEASE INCLUDE YOUR CODE IN ADDITION TO YOUR ANSWER. YOU MUST ALSO INCLUDE ANY RELEVANT OUTPUT. YOU MUST USE SAS FOR THIS WORK. YOU WILL BRING THIS WITH YOU TO THE IN-CLASS PORTION OF THE EXAM, WEDNESDAY, DECEMBER 6. 1) Determine if there is a difference in the mean high school graduation rate (PercentHSGrad) between urban and nonurban counties. 2) Determine if there is a difference in the proportion of urban and nonurban counties who have Low Lottery Sales. Use SAS code that includes the word \"riskdiff\" in some way. 3) Determine the linear model for predicting the population of a county if the median income was known. 4) For the model created in Question 3, determine the predicted population of a county with a median income of $20,000, $50,000, and $100,000. Find the 95% confidence interval for the mean of each of these, as well as the 95% confidence interval for the value of each of these. 5) Find the correlation coefficient between Bankruptcy Rate and Unemployment Rate for 2009. Determine the linear model that would be used to predict the Bankruptcy Rate if the Unemployment Rate were known. 6) Create a 2x2 Contingency Table between LottSales and ViolentCrime and run Chi-Square, Odds and Relative Risk tests. Use LottSales as your Rows. 7) Test the hypothesis that the mean of the percent of residents who registered as Democrat (PercentRegDem) is less than 40% at a 2% level of significance. 8) Create the 99% confidence interval for the mean of the Percent of residents who are on Disability (PctonDisability). 9) Create the 95% Confidence Interval for the percent of counties that are considered \"URBAN\". Test the hypothesis that this percent is not equal to 50% at a 5% level of significance. Find the power of this test