The first round of data collection was carried out in late 2005 and 2006. These probability surveys are valid* within a statistical margin of error, also called a 95% confidence interval. This means that if the survey is conducted 100 times using the exact same procedures, the margin of error would include the "true value" in 95 out of the 100 surveys. With a sample size of 1,000, the margin of error for a percentage at 50% is ±3 percentage points.
Because these surveys use a clustered sample design, the margin of error varies by question, and if a user is making critical decisions based on the margin of error, he or she should consider inflating the margin of error by the design effect. The design effect accounts for the potential of correlated responses, and increase in the margin of error, caused by the sample of clusters of households in PSU.
*Assuming other sources of error, such as nonresponse, by some members of the targeted sample are equal. Other errors that can affect survey validity include measurement error associated with the questionnaire, such as translation issues, and coverage error, where a part or parts of the target population aged 15 and older have a zero probability of being selected for the survey.
Contact Us
Acquire the power to understand and act on what the world is thinking. For more information, please e-mail worldpollpartners@gallup.com or call 202.715.3030 (Washington, D.C.) or +44 (0) 20 7950 4400 (London).
Based on the largest and most in-depth study of its kind, this book presents the remarkable findings of the Gallup Poll of the Muslim World.