1. Specify the population definition
The population that they will use is car owners and non car owners. The use of non car owners is prudent because the development of the new models of cars is essentially a 5 year project. Mr. Rogers and Mr. Thomas feel that these non car owners may move into the car owner category within that 5 year time period. As far as the boundaries are concerned, they will include all U.S. households.

2. If a probability sampling method is to be used, what would be a reasonable sample frame for
a. Telephone survey
i. All those who have a telephone ii. Number is available in the yellow pages iii. Likely to be home during business hours of the day
b. Mail survey
i. All those who regularly check their mail ii. All those who live within a certain mile radius of a car dealership
c. An online survey
i. People who buy online and use credit cards ii. All those who use online banking iii. All those who own a computer and use it regularly

3. What are the practical problems involved with drawing a simple random sample of American households (regardless of the survey method)?
By drawing a simple random sample, you are going to get a ton of different inconsistent data. To use all of American households, is extremely broad and may return information all over the place. At that point you just wasted all that time trying to find out information from your sample and now you have a plethora of surveys completed with the possibility of inconsistent data.

4. If random digit dialing was used for the sample plan, what are the advantages and disadvantages of this sample method?
Advantages:
- You would possibly be reaching numbers that are not available in the yellow pages
- Cost efficient way of reaching a ton of people
- Embodies the requirements necessary to obtain a probability sample
- Unbiased estimates of the surveyed