Part B Breast cancer recurrence. Download "breast cancer recurrence.csv". The data contain follow-up information on 4901 breast cancer patients. The second variable "recurrent" indicates recurrent cancer. The value "1" means that cancer recurred the value "0" means that cancer did not recur. Use Stata to perform the following calculations: (a) Create a bar graph of the variable "recurrent" (make sure graph is appropriately labeled). (b) Create a frequency table of the variable "recurrent" (make sure table is appropriately labeled). Determine the recurrence rate in this entire sample. (c) Determine the recurrence rate and 95% confidence interval of the entire sample using the normal approximation. (Choose "wald" when selecting the confidence interval type or use the command: ci proportions recurrent, wald [Note: the "wald" option means we are using the normal approximation. STATA's default is the exact]). (d) Determine the recurrence rate and 95% confidence interval of the entire sample using the exact method (Choose "exact" when selecting confidence interval type (STATA default) or use command: ci proportion recurrent, exact) (e) Compare and contrast the recurrence rate and confidence intervals for the entire sample (normal approximation (c) versus exact binomial (d). Is the normal approximation appropriate to use for the entire sample? Explain. (f) Take the first 10 observations and determine the recurrence rate and the 95% confidence interval using the normal approximation. Choose "Wald" when selecting the confidence interval type or use the command: ci proportions recurrent in 1/10, wald (Note: the "1/10" in the command means we are only using observation 1-10 and not the entire sample. The Wald means we are using the normal approximation). (g) Use the proportion and the standard error from the output in (f) from the first 10 observations to manually calculate 95% confidence interval based on the formula using the normal approximation: 1.96 x SE. Compare your results to the STATA output from (c). What happens to the lower endpoint of the confidence interval from Stata. (h) Take the first 10 observations and determine the recurrence rate and the 95% exact binomial confidence interval. Choose "exact" when selecting the confidence interval type or use the command: ci proportions recurrent in 1/10, exact (i) Compare and contrast the recurrence rate and Confidence intervals calculated from the first 10 observations (parts (f) and (h)). Which confidence interval approach (normal or exact) is most appropriate for determining the confidence interval of the first 10 observations? Explain.