|Country Report Cambodia - ICRC Worldwide Consultation on the Rules of War (International Committee of the Red Cross , 1999, 72 p.)|
The ICRCs worldwide consultation on the rules of war, which is the cornerstone of the People on War project, was carried out in 12 countries that have been ravaged by war over the past decades. In each case, the ICRC conducted a public opinion survey with a representative sample of the countrys population and organized in-depth interviews and focus groups with those involved in or directly affected by the conflict.
For comparative purposes, the views of people were also sought in France, Russian Federation, Switzerland, United Kingdom and United States on the basis of the opinion survey only.
The consultation was based on three principal research methods:
· A survey of 1,000 (in some cases 1,500) respondents representative of the countrys general population;
· Focus groups (between 8 and 12 depending on the country) allowing a professionally moderated and intensive discussion in small groups;
· In-depth, face-to-face interviews (about 20 in each country) with individuals with specific war experiences.
In almost every case, the ICRC and local Red Cross or Red Crescent staff conducted the interviews, organized the focus groups, including recruitment of participants, and helped with translation/interpreting. Greenberg Research, with a local partner company, developed the sample design for the survey, processed data in electronic form, provided moderators and prepared transcripts.
Questionnaire. The opinion survey questioned people on their war experiences and views on international humanitarian law. The survey was mainly standardized for all countries, though the wording was modified to reflect each context and to achieve consistent meaning. About 10 per cent of the questions were contextual and in many cases unique to the country. In an additional five countries, the questionnaire was designed to elicit peoples perceptions on war and humanitarian law.
The questionnaires were developed by Greenberg Research, in consultation with the ICRC, on the basis of interviews with humanitarian law experts in the United States and Europe. The survey and questions were pre-tested in Mozambique and Colombia.
Sample design. In each country, interviews were held with 1,000 to 1,500 respondents, selected by a stratified, multistage cluster sampling method. The sample was stratified to ensure representation (500 interviews) from each of the principal conflict-affected geographic areas or ethnic/religious groups. In Bosnia-Herzegovina, for example, this meant some 1,500 interviews (500 from Republika Srpska and 500 each from the Bosniac and Croat areas of the Federation of Bosnia-Herzegovina); in Israel, the occupied territories and the autonomous territories, this meant 1,000 interviews (500 in Israel and 500 in the occupied territories and the autonomous territories). These areas were divided into urban and rural geographic units (strata), to ensure representation of urban and rural populations.
The local partner randomly selected small geographic units within these strata. These units - 100 to 200 in each country - constituted the sampling points for the survey. In each geographic unit, 10 households (though fewer in some countries) were selected using a random route method appropriate to the country. In some cases, interviewers were provided with a map and a route; in others, interviewers were simply given a route and selection instructions.
Within households, respondents were selected using a Kish grid (a respondent selection key that employs a combination of random numbers, alphabet codes and the number of available members in a household to identify the appropriate respondent) or the birthday criterion (a respondent selection process that employs dates of birth to determine the appropriate respondent). Interviewers were to make three attempts to achieve a completed interview, including locating the respondent elsewhere. In nearly every country, non-response was below 10 per cent.
The demographic distribution of the surveyed respondents was compared with the best available census data on education, age, household type and occupation. Where the sample survey was sharply askew (e.g., too many college-educated or too many young respondents), statistical weights were applied to eliminate the bias.
Interviews carried out by phone reached 755 adults in France, 1,000 in Switzerland, 750 in the United Kingdom and 1,000 in the United States, and 1,000 face-to-face interviews were carried out in the Russian Federation.
Survey administration. In nearly all the countries, the survey was administered by the ICRC, with the assistance of Greenberg Research and a local research partner. Interviews were conducted by Red Cross or Red Crescent staff. Greenberg Research provided training, which typically took two days.
Parallel research. In three of the countries - Colombia, Bosnia-Herzegovina and the Philippines - Greenberg Research commissioned a parallel quantitative survey, administered by a local research company using professional interviewers, in order to identify patterns of bias. The results of the parallel studies were then compared with the results of the ICRC-administered surveys. The exercise found only a few areas of systematic bias. Those interviewed by the ICRC and Red Cross or Red Crescent staff, for example, were consistently more supportive of the ICRCs role and more aware of the Geneva Conventions and the rules of war. However, the parallel research found few systematic differences in opinions on international humanitarian law. The ICRC results closely resemble the parallel survey results on most other questions. (A technical report assessing the parallel research and Red Cross bias is available separately.)
Focus groups. The focus groups provided a relatively unstructured environment for people to discuss their war experiences freely, express their views on the appropriate limits to war and consider possible actions against those who exceed them. To be effective, the groups had to be as homogeneous as possible, that is, the participants all had to have similar characteristics. Thus, in general, the participants in a group came from the same area, were all male or all female and shared an important experience (e.g., families of missing persons, ex-soldiers, ex-fighters, prisoners, teachers or journalists). The discussions were frequently intense and emotional and provide a rich commentary on how the public approaches these issues.
In each country, 8 to 12 focus groups were organized - four in each of the principal conflict areas. The participants were recruited by Red Cross or Red Crescent staff, based on guidelines provided by Greenberg Research. The local research company provided a professional moderator, who facilitated the discussions using guidelines prepared by Greenberg Research.
The discussions were held in focus-group facilities, school classrooms, hotel rooms and even in the open air, if, for example, they involved guerrilla fighters. ICRC, Red Cross/Red Crescent and Greenberg Research staff observed and listened to the discussions from an adjoining location, with simultaneous translation in English. The focus group discussions were recorded and later transcribed in English.
In-depth interviews. To help interpret the full meaning of the survey responses, about 20 in-depth interviews were conducted with individuals who had had specific war experiences. The in-depth interview guidelines repeated questions from the public opinion survey, although they allowed for open-ended, rather than categorized responses. Interviewers were encouraged to probe and follow up on responses.
The in-depth interviews involved a broad range of people - officers, medical personnel, students (secondary school and university), journalists, former combatants, refugees, displaced persons, family members of missing persons, war invalids and others.
The interviews were recorded on tape, transcribed and translated into English by the local partner.