GfK Roper Consulting conducts market research for multinational companies who want to understand attitudes in different countries so they can market and advertise more effectively to different cultures. Every year they conduct a poll worldwide, which asks hundreds of questions of people in approximately 30 different countries. Respondents are asked a variety of questions about food. Some of the questions are simply yes/no (agree/disagree) questions: Please tell me whether you agree or disagree with each of these statements about your appearance: (Agree = 1; Disagree = 2; Don’t know = 9).

The way you look affects the way you feel. I am very interested in new skin care breakthroughs. People who don’t care about their appearance don’t care about themselves. Other questions are asked on a 5-point scale (Please tell me the extent to which you disagree or agree with it using the following scale: Disagree completely = 1; Disagree somewhat = 2; Neither disagree nor agree = 3; Agree somewhat = 4; Agree completely = 5; Don’t know = 9).

Examples of such questions include:

I read labels carefully to find out about ingredients, fat content, and/or calories.

I try to avoid eating fast food.

When it comes to food I’m always on the lookout for something new.

Think about designing a survey on such a global scale:

a. What is the population of interest?

b. Why might it be difficult to select an SRS from this sampling frame?

c. What are some potential sources of bias?

d. Why might it be difficult to ensure a representative number of men and women and all age groups in some countries?

e. What might be a reasonable sampling frame?