Polling in Iraq: Planning, Luck and Tragic Stories

In Maysan province on Feb. 27, said another, "before finishing my work the supervisor called me to say he will come to visit me to check my work. I waited for him for a long time, but he didn't come. I learned that a police patrol had arrested him after seeing him taking pictures. Because of the efforts of our field manager for this region, his telephone calls, and his personal relations with people in this governorate, the supervisor was released and returned back home safely."

SAMPLE -- The survey used an area probability sample based on 2005 population estimates by the Iraq Ministry of Planning. D3/KARL randomly assigned sampling points proportionate to the population of each of Iraq's 18 provinces, then proportionately to population size in each of the 102 districts within these provinces. Within districts, sampling points were assigned randomly among nearly 11,000 known settlements or neighborhoods. At each stage, the sample was stratified by urban or rural locale.

Maps (or a grid plan where reliable maps were unavailable) were used to select a random starting point for interviewers at each sampling point. Interviewers traveled to the starting point, randomly selected households (e.g., third house on the left), then randomly selected respondents within each household. (See separate methodological summary.)

In-person surveys of this type customarily interview 10 or more respondents per sampling point; 2,000 interviews would require 200 sampling points. Some use fewer points; a typical election exit poll in California (whose area and population are similar to Iraq's) may have 35 to 50 sampling points. To maximize this survey's geographical distribution, D3/KARL used 458 sampling points across Iraq, with about five interviews per point.

INTERVIEWS -- The questionnaire was prepared by ABC News in consultation with its media partners, translated into Arabic and Kurdish (using the Sorani dialect) by D3/KARL, then backchecked by translators working for ABC News. Interviews lasted about 30 minutes on average.

The survey had a cooperation rate of 62 percent -- far higher than usual in telephone surveys in the United States, albeit lower than ABC's previous surveys in Iraq, conducted in less violent times.

Photos from interviews show the range of homes visited -- from small villages to urban centers, from gated homes to shabby apartments.

"The situation of the neighborhood was good, as services are available, but there is no electricity line and there are only private electricity generators shared by people," one interviewer noted. "There is a shortage in fuel and it is very expensive, so people bring cooking and car fuel from the Iranian borders."

Said another, in Irbil, "I noted that the services in this region are average, but there is a shortage in electricity and it is not available before noon. Electricity comes for three hours only and there is no hygiene drainage [sewer] in the region."

Another, also in Irbil, sounded a common concern of farmers everywhere: "They complain that their harvests are being sold at prices that are too low because of the presence of vegetables and fruits from all neighboring countries of Iraq," the interviewer noted. Moreover, "People told me that most young people join the Peshmerga in order to get salaries and they don't work with their parents on the farm."

  • 1
  • |
  • 2
  • |
  • 3
Join the Discussion
blog comments powered by Disqus
You Might Also Like...