Question 2: Head Office Appearance A CEO at a large retailer wants to conduct a survey to estimate the proportion of employees who are satisfied with the appearance of the head office buildings and grounds. The CEO is considering three methods of obtaining a sample of 150 from the 2000 employees at the head office. (a) For ease and timeliness, the first method the CEO is considering is to select the first 150 employees who arrive at work one Friday morning. Identify the sampling method and explain why such a sampling method may or may not be biased in producing an estimate of the proportion of employees who are satisfied with the appearance of the buildings and grounds. (b) Because of the large number of employees at the head office, the second method the CEO is considering consists of using a computer with a random number generator to select a simple random sample of 150 employees from a list of 2000 employees' names. Describe how to implement such a method. (c) Because stratification can often provide a more precise estimate than a simple random sample, another method the CEO is considering consists of selecting a stratified random sample of 150 employees. The head office campus has two major buildings (an older building and a newer building) with male and female employees spread out across the campus. Over 98% of the head office staff consists of Tech" (information technology) and marketing / retail support, both spread out across the campus. Under what circumstance(s) would stratification by Tech and marketing/retail support be a more precise estimate of the proportion of employees who are satisfied with the appearance of the buildings and grounds than stratification by gender? (d) Assume the business systems provide an email address to each and every employee at head office. The CEO sends an email survey and receives positive support on buildings and grounds from the 30% that responded. Should the CEO simply accept these results? Briefly explain why or why not, and explain how bias, if any, would inflate or deflate the sample estimate vs. the true parameter