the interviewing in Canada (Quebec), Mexico, the United Kingdom, Germany and Japan, WBA partnered with Rosenthal Research based in Barcelona, Spain to coordinate the international sample purchase as well as programming and administration of the survey instrument. In survey research, the entire population is typically not interviewed, but rather a sample of that population is polled. Therefore, the data are subject to sampling error. The sampling error will vary depending on the sample size and the percentages being examined in the sample.
For the roughly 250 interviews conducted per country for the National Sleep Foundation's 2013 International Bedroom Poll, we can expect the values from our sample to be within about 6 percentage points of the true population value, 95 percent of the time. That means, if we get a value of 50 percent from our sample, we can be 95 percent sure that the population value is somewhere between 44 and 56 percent. This measure of the difference of the sample value from the true value, called the 'margin of error,' varies with the size of the sample and the results obtained.