If you were designing a quantitative survey of employees, what type of sample design would be best to obtain representative opinions from employees? How would you insure anonymity of individual responses on the quantitative survey, as well as a broad range of employee opinions?