Some researchers have argued that marriage decreases the likelihood of engaging in criminal behavior more for males than females (on average). Others argue that the effect of marriage varies across subgroups because of differences in cultural expectations. Design a sampling strategy that will allow you to compare the relationships between gender and marriage for three ethnic/racial groups in one US city. Prior research suggests that comparing African Americans, Hispanic Americans, and Anglo Americans might be a productive strategy, although you have not made up your mind as of yet. Assume that you can talk to only one member of a couple (one interview per phone call and phone number) and get the data you require.
You have enough money to complete 1500 telephone interviews.
Identify the city you will study. Choose your city wisely given that you need to stratify the sample by race and gender.
Go to the US Census webpage to collect demographic information on the city (not an MSA or CMSA) you will study. Cities must be in the United States and have populations of between 300,000 and 4,000,000.
Go to the US Census webpage (http://www.census.gov/) and access the 2010 census (the decennial census). Within the 2010 census, find a city that meets the population parameters of this assignment. While you are searching for data, do note that the census refers to cities, towns, villages, etc. as Place.
For all the data you use from the Census, you must make a copy of the data (it is usually in table format) and submit it with your assignment. Within your assignment, you must explain in writing, in the appropriate section of your paper, which data/table you used from the Census and why you chose that data/table (i.e. why you chose that particular table with the race data for your city).
For a simple random sample using random digit dialing, how many phone numbers will you need? For your calculations, assume that 50% of the telephone numbers will be working numbers, 70% of the numbers will be residential (non-business) numbers, and your response rate is 70% (which means 70% will agree to participate). For the purposes of this study, do not concern yourself with landlines versus cell phones. Given the focus of your study, what else must you factor into your calculations to determine how many phone numbers you will need? You will need the answer to this question as well as the aforementioned assumptions to determine how many phone numbers you will need to call.
You must also determine the total cost of the survey. If a completed interview costs $75 and a refusal or not-in-universe case (i.e. business number, non-married) costs $7, and a non-working number costs $1, how much will this survey cost?
This study requires you to stratify the sample. Groups that stratify the sample by race will be eligible for a B on the assignment. Groups that stratify the sample by race and gender will be eligible for an A on the assignment.
Keep in mind that when stratifying the sample, the percentages must be made proportionate to 100% and sub-samples need to have an N of at least 100.
When doing your calculations, DO NOT round until you have determined the last number in that calculation. Then, round up for .5 or higher and down for .4 and lower. Keep in mind that you cannot have half a person, half of a phone number, or half of a survey.
Your assignment must be double-spaced with each section of the sampling design clearly designated (i.e. use headings). When you submit this assignment, you must also submit copies of the census tables and/or data you used to complete the assignment.
Links:
https://censusreporter.org/profiles/05000US36001-albany-county-ny/
