Courses   <-  Sean Forman   <-  You Are Here
   Next: About this document ...
Section 2.3

MAT 1181, Sean Forman

Spring 2006



  1. To study the effects of exercise on the risk of heart disease, an investigator wishes to compare the incidences of heart disease in bus drivers and pedestrian police officers in New York City. He selects subjects so that the ages of the two groups are very similar and each subject has been on the job for at least ten years.

    1. What factor is the researcher studying?
    2. Answer: the effect of exercise on heart disease.

    3. What is the response variable?
    4. Degree of heart disease.

    5. Is this an observational study or a randomized experiment?
    6. Answer: Observational.

    7. List two possible confounding variables.
    8. Answer: Diet and stress.

  2. To determine the percent of men and women who smoke, the Surgeon General mailed questionnaires to 5,0000 people across the US.

    1. What is the torget population?
    2. All American adults

    3. What is the sample?
    4. 5,000 american adults

    5. List four important variables for the Surgeon General to study. Classify each as categorical or numerical (discrete or continuous).
    6. Amount of smoking (numerical contin.), age (numerical discrete), gender (categorical), income (numerical discrete)

    7. What inference would be made from such a study?
    8. Answer: That the results of the survey match the actual results for the population.

  3. A study of 100 countries showed that countries with higher numbers of television sets per household also had higher life expectancies.

    1. Is this a randomized experiment or an observational study?
    2. Study.

    3. What are the members of the population of interest? What is the sample?
    4. All countries, 100 countries

    5. What are the variables of interest in the study? Which is the explanatory factor or treatment and which is the response variable?
    6. Explanatory: tv ownership, resonse: life expectancies.

    7. Can we safely conclude that television extends life?
    8. No, observational studies alone do not show cause and effect and there are LOTS of confounding variables.



Next: About this document ...
sforman@sju.edu