Iraq Poll: Note on Methodology
National Iraq survey involved a random sample of 2,212 Iraqis aged 18 and older.
Sept. 10, 2007 — -- This survey was conducted for ABC News, the BBC and NHK by D3 Systems of Vienna, Va., and KA Research Ltd. of Istanbul, Turkey. Interviews were conducted in person, in Arabic or Kurdish, among a random national sample of 2,212 Iraqis aged 18 and older from Aug. 17 to Aug. 24, 2007.
Four-hundred-fifty-seven sampling points were distributed proportionate to population size in each of Iraq's 18 provinces, then in 101 of 102 districts within the provinces (excluding Sinjar district in Ninevah province, which was not accessible because of security concerns at the time of the survey), then by simple random sampling among Iraq's nearly 11,000 villages or neighborhoods, with urban/rural stratification at each stage.
Maps or grids were used to select random starting points within each sampling point, with household selection by random route/random interval and within-household selection by the "next-birthday" method. An average of five interviews were conducted per sampling point. Sixteen of the 457 sampling points were inaccessible for security reasons and were substituted with randomly selected replacements.
Interviews were conducted by 117 trained Iraqi interviewers with 30 supervisors. Sixty-nine percent of the interviews were supervised or reviewed by supervisors -- 44 percent by direct observation, 16 percent by revisits and nine percent by phone. All questionnaires were subject to further quality-control checks.
In addition to the national sample, oversamples were drawn in Anbar province, Sadr City, Basra city and Kirkuk city to allow for more reliable analysis in those areas. Population data came from 2005 estimates by the Iraq Ministry of Planning. The sample was weighted by sex, age, education, urban/rural status and population of province.
The survey had a contact rate of 93 percent and a cooperation rate of 65 percent for a net response rate of 60 percent. Including an estimated design effect of 1.51, the results have a margin of sampling error of 2.5 percentage points at the 95 percent confidence level.
