Last week, you finally landed the job you always wanted. You have been hired as a project director for a research firm that specializes in politics but works mostly for newspapers and television stations at the local level.
The firm has been hired by the Des Moines Register newspaper to conduct a poll in the state of Iowa, and you have been assigned as the project manager. The firm will conduct a telephone survey, N = 800, of likely 2020 general election voters in Iowa, asking about general election ballots for president. There will probably be some questions about U.S. Senator Ernst, who announced her intention to run for re-election some time ago.
Before the firm moves forward to start designing a questionnaire, they want a report on your recommended sampling methodology, including how the interviews would be distributed geographically throughout the state. They may also want to look at the data by congressional district, because that is how the caucus system is organized in the state.
In a short memo (no more than two pages), please explain what steps you will take to decide how to distribute interviews, what information you referenced to determine that, and a summary of the results. It is critical that you recommend a sampling frame(s). Journalists in Iowa are a lot more politically plugged in than in other parts of the country and they can smell a neophyte from afar.
Finally, rather than setting interview quotas by county, the paper would like to set quotas by region. Given the political focus of these surveys, make a recommendation for setting regional quotas, and write in your memo how many interviews would fall into each region.
Along with the memo, you will need to turn in the spreadsheet you used to make the calculations. The spreadsheet should contain all the data you collected to make your calculations, the formulae used to make the calculations, and a column at the far right containing your final distributions for the survey. That column should have the following header, so it can be easily identified: “All Voter N =”.
Remember, your job here is to design a sample based on a series of regional areas. You should make certain you are drawing from the correct distribution within each area, based on your key assumptions about the race about which you are trying to poll.
This assignment requires the use of a spreadsheet. If you are unfamiliar with the manipulation of a spreadsheet, let the instructor know that you will need additional assistance and he/she will set that up for you.
Your memorandum to the client should include answers to the following questions once your spreadsheet is completed:
- What is the regional distribution of men and women in your sample?
- What is the regional distribution of interviews by party in your sample?
- What happens when the phone bank is unable to finish the proper number of interviews in one of your regions?
- In this case, what if Region A is 13 interviews short and region B has too many interviews? What other information might you need to determine a solution?
You will turn in both a spreadsheet and a memorandum reviewing your recommendations for a sample distribution. The memo should be limited to two pages, but the spreadsheet must include all the calculations you made during the course of your decision-making.
Note: Spreadsheet software such as Microsoft Excel will often round numbers with decimal places, and as a result, the actual sum of interviews may not be exactly what you desire. Double-check the results and make judgment calls about where to place an additional interview or two if the sum is a couple short.
