On June 27, 1996, an article appeared in the Wall Street Journal presenting the results of a nationwide poll regarding the White House procurement of FBI files on prominent Republicans and related ethical controversies. The article was headlined "White House Assertions on FBI Files Are Widely Rejected, Survey Shows." At the end of the article, the explanation of the sampling procedure, as shown in the box at the bottom of the preceding page, was given. Discuss the different aspects of sampling that appear in this explanation.
The Wall Street Journal/NBC News poll was based on nationwide telephone interviews of 2,010 adults, including 1,637 registered voters, conducted Thursday to Tuesday by the polling organizations of Peter Hart and Robert Teeter. Questions related to politics were asked only of registered voters; questions related to economics and health were asked of all adults. The sample was drawn from 520 randomly selected geographic points in the continental U.S. Each region was represented in proportion to its population. Households were selected by a method that gave all telephone numbers, listed and unlisted, an equal chance of being included. One adult, 18 years or older, was selected from each household by a procedure to provide the correct number of male and female respondents. Chances are 19 of 20 that if all adults with telephones in the U.S. had been surveyed, the finding would differ from these poll results by no more than 2.2 percentage points in either direction among all adults and 2.5 among registered voters. Sample tolerances for subgroups are larger.