| Eco309 Economic Statistics | |
| Homework 1 |
Iona College - Economics
|
| 1. Suppose that a TV manufacturer is interested in
learning how satisfied recent customers are with their TV purchases.
A survey of 50 customers is being planned out of the 4000 that had returned
warranty registration cards. The TV manufacturer is also interested
in assessing whether there are differences in the level of satisfaction
for different TV models.
A. Why are probability based samples superior to non-probability based samples for obtaining information on population characteristics? B. Describe how you could draw a simple random sample of 50 persons out of the population of 4000. Use the random number Table E.1 (in the back of the text), and assume that the starting digit for your sample would be at row 15, column 01. List the first five sample numbers. How would you use these numbers? C. Describe how a stratified random sample of 50 persons out of the population of 4000 could be drawn. Advantages/ disadvantages? D. Describe how a systematic random sample of 50 persons out of the population of 4000 could be drawn. Advantages/ disadvantages? D. Describe how a cluster sample of 50 persons out of the population of 4000 could be drawn. Advantages/disadvantages? F. If the TV manufacturer used the "best" of the sampling methods described above, what kind of survey errors might still arise that would make the sample statistics poor measures of the population's characteristics? |
| 2. The MBADATA.xls
Excel spreadsheet contains program information for nearly all (N=624) US
MBA programs in two separate years (1988 & 1995). The SIZE95
variable contains 1995 enrollment data for each of the programs.
Use PHSTAT2 or EXCEL to list the 1995 enrollment level for a simple random
sample of 5 schools.
PHSTAT2 programming: A. Open the file and then click on <PHSTAT> <SAMPLING> <RANDOM SAMPLE GEN> and then enter the desired sample size and the range of values to select from. Can you identify which schools were included in the sample? B. If you wanted to generate a list of random numbers that could be used to select particular cases from the data file, follow the above steps but instead of entering a range of values to select from, enter the population's size. Try this option. EXCEL programming: A. Open the file and then click on <DATA><DATA ANALYSIS><SAMPLING><OK>. Then click on <RANDOM> and enter the desired sample size in the "Number of Samples" field. Can you identify which schools were included in the sample? NOTE: If <DATA ANALYSIS> is not available on the <DATA> menu, then you'll have to enable the "Analysis Tool Pak" Add-Ins in EXCEL. To do so, click on the Office Button (top left) , then <Excel Options><Add-Ins><Manage Excel Options><Go> and check mark the Analysis Tool Pak boxes and then click on OK. After you click on <OK> the <DATA ANALYSIS> option should appear on the <DATA> menu. |
| 3. A Federal Reserve Bank of St. Louis study, entitled "A
Recession for the Educated?", examines trends in the unemployment rate
of persons in the US with differing educational levels during the last
recession (in 2001).
A. According to the study, was the increase in unemployment during the last economic slowdown greater or smaller for college grads or those w/o high school diplomas? B. Using the numbers in the text and the chart, fill in the 1994, October 2000 and October 2001 unemployment rates for the three educational groups in the table below. According to your table, which groups suffered greater/smaller increases in unemployment from October 2000 to October 2001 (the last recession)? Educational
Unemployment Rate (%):
College graduate H.S. graduate < H.S. graduate
|