The Shell Oil Company, which began about 1912, had been for decades a household name as a quality oil company in the United States. However, by the late 1970s much of its prestige as a premiere company had disappeared. How could Shell regain its high status? In the 1990s, Shell undertook an extensive research effort to find out what it needed to do to improve its image. As a first step, Shell hired Responsive Research and the Opinion Research Corporation to conduct a series of focus groups and personal interviews among various segments of the population. Included in these were youths, minorities, resident in neighborhoods near Shell plants, legislators, academics, and present and past employees of Shell. The researchers learned that people believe that top companies are integral parts of the communities in which the companies are located rather than separate entities. These studies and others led to the development of materials that Shell used to explain their core values to the general public.

Discussion

1. Suppose you were asked to develop a sampling plan to determine what a “premiere company” is to the general public. What sampling plan would you use? What is the target population? What would you use for a frame? Which of the four types of random sampling discussed in this chapter would you use? Could you use a combination of two or more of the types (two-stage sampling)? If so, how?

2. It appears that at least one of the research companies hired by Shell used some stratification in their sampling. What are some of the variables on which they are stratified? If you were truly interested in ascertaining opinions from a variety of segments of the population with regard to opinions on “premiere” companies or about Shell, what strata might make sense? Name at least five and justify why you would include them.

3. Suppose that in 1979 only 12% of the general adult U.S. public believed that Shell was a “premiere” company. Suppose further that you randomly selected 350 people from the general adult U.S. public this year and 25% said that Shell was a “premiere” company. If only 12% of the general adult U.S. public still believes that Shell is a “premiere” company, how likely is it that the 25% figure is a chance result in sampling 350 people?

4. PERT Survey Research conducted quantitative surveys in an effort to measure the effectiveness of various campaigns. Suppose on their survey instrument, they used a continuous scale of from 0 to 10 where 0 denotes that the campaign is not effective at all, 10 denotes that the campaign is extremely effective, and other values fall in between to measure the effectiveness. Suppose also that a particular campaign received an average of 3.6 on the scale with a standard deviation of 1.4 early in the tests. Later, after the campaign had been critiqued and improved, a survey of 35 people was taken and a sample mean of 4.0 was recorded. What is the probability of this sample mean or one greater occurring if the actual population mean is still just 3.6? Based on this probability, do you think that a sample mean of 4.0 is just a chance fluctuation on the 3.6 population mean, or do you think that perhaps it indicates the population mean is now greater than 3.6? Support your conclusion. Suppose a sample mean of 5.0 is attained. What is the likelihood of this result occurring by chance when the population mean is 3.6? Suppose this higher mean value actually occurs after the campaign has been improved. What does it indicate?