HW1 Solutions

Each answer worth 5 points, 65 total points

1. Read 1.12 but don't answer the book's questions. Instead answer these:
   1. What was the target population?
       
      Answer: All men and women
       
   2. What was the sampled population?
       
      Answer: Men and women from Gainesville, FL (not the 300 people, that is the sample not the sampled population).
       
   3. How many subjects were in the sample?
       
      Answer: 300
       
   4. How many variables were measured?
       
      Answer: 20. The variables are whether or not each object was recalled correctly.
       
   5. Does the study involve inferential or descriptive statistics? Why?
       
      Answer: Inferential. A sample was collected to make an inference about a population.
       
2. Referring to the section Calculating Excess Deaths in the Cato Report Lies, Damned Lies and 400,000 Smoking Related Deaths answer the following questions:
   1. Was the CPS sample representative of the target population (all Americans)? Why or why not?
       
      Answer: No. The sample was healthier, more educated and more white than the target population.
       
   2. The sample resulted in an over-estimate of "smoking-related deaths". Why?
       
      Answer: It underestimated the death rate of non-smokers.
       
   3. What kind of sample methodology was used in the CPS survey?
       
      Answer: A sample of convenience.
       
3. What is the difference between a sample of convenience and a random sample? Which is harder to collect? Why?
    
   Answer: A sample of convenience is collected without any statistical design. A random sample is collected in such a way that every member of the population has the same probability of being included in the sample. A random sample is harder because it require more planning to implement.
    
4. We want to estimate WSU students opinions on several subjects. A sample of WSU students is selected by posting fliers in the Smaug asking for volunteers to take our survey. Since 100 students volunteered and we only needed 30 we put the hundred volunteers names in a hat and selected a final sample 30.
   1. Is the sample of 30 a random sample of WSU students?
       
      Answer: No, students who volunteered had equal probabilities of being in the sample but non-volunteers had no chance of being in the sample.
       
   2. Is the sample of 30 a random sample of students who volunteered for the survey?
       
      Answer: Yes
       
   3. Is the sample of 30 likely to be representative of all WSU students? Why or why not?
       
      Answer: No. There is still likely to be volunteer error here.
       
5. Summarize the "within household selection process" described in How Polls are conducted
    
   Answer: They want to randomly select an adult in the household so they often ask the person who answers the phone to list all the adults in the household and they randomly select one name. Sometimes the ask for the adult with the latest birthday in the year as a shorthand method.

E-mail Mr. Callahan at stat110@edcallahan.com with questions or comments about this web site or about the class itself.