The purpose of this exercise is to demonstrate that while a random sample from a population is not guaranteed to give a completely accurate representation of any characteristic of the population, a random sample tends to give a fairly accurate representation of the population across a wide variety of characteristics.

1. (5) Refer to the SPSS data file containing information about the 328 staff survey results (STAFFSURVEY.sav).  For all 328 surveys, obtain descriptive measures for years of service (Service), construct a histogram for years of service (Service), and construct a bar graph for the age of the staff member (Age). (3 points)
2. (6) Draw 5 different simple random samples, each of size 25, from the 328 staff survey results.  Use SPSS to draw the samples.  Set the starting point 3409  To do this, go to the Transform menu, select Random Number Generators and look under Active Generator Initialization.  Select Set Starting Point.  Choose Fixed Value and type in the last four digits of your ID number.  After each sample is drawn, the data file can be reverted to its original form by highlighting the filter column and right-clicking to clear it. If you elect to output a dataset for each random sample of n = 25, you wont need to revert the file back to its original form.
3. (6) For each sample, obtain descriptive statistics min, max, mean, standard deviation for the years of service (Service), construct a histogram of the years of service (Service), and construct a bar graph for the age of the staff member (Age).  You do not need to show the staff members selected, but you do need to show the statistics and graphs for each sample.
4. Using this exercise as an example, explain (briefly, please) what is meant by the terms
   1. (1) Sampling variability
   2. (1) Sampling error
5. (1) Do you think that this exercise has illustrated the purpose stated above?  Explain briefly.