Botswana, India, Northern Ireland and Zimbabwe agreed to join the project in 1995, following discussion with the Intercultural Centre for Intercultural Studies at the London University Institute of Education which had been in touch with other nations also. In addition to a willingness to participate, four other factors had to be considered: Commonwealth representativeness, support or concern for human rights in the country, convenience in managing the project, and financial cost.

Inevitably there was an arbitrary and fortuitous element in the final selection. With regard to Commonwealth representativeness, however, it was felt deisrable to include at least one developed country, and at least one with a relatively small population. It was necessary to obtain support from countries in different world regions, because the Commonwealth is a trans-regional association. Participation from India was particularly welcome, because over half the people in the 53 nation Commonwealth are citizens of that country. It was not, unfortunately, possible to include any Caribbean or Pacific nation.

While all Commonwealth countries assert their support for human rights, and none are totally immune from problems, it was felt that something more than the interest of an Education Ministry would be helpful in justifying inclusion in this study. While other states could also meet such criteria, it is worth observing certain positive elements in the four participants.

For example, Botswana has a particularly strong record in Africa of plural, multi-party democracy since independence; it was partly for that reason that it hosted a meeting of African Commonwealth leaders and political parties, to discuss democracy, in February 1997. India has recently established a National Human Rights Commission, has a strong culture of human rights NGOs, and is in the forefront of legal redress through public interest litigation based on its constitution. Northern Ireland, which has suffered from over 25 years of internal conflict, has the only compulsory curricular commitment in this field in Britain (Education for Mutual Understanding). Zimbabwe, the home of the Harare Commonwealth Declaration, 1991, has an inter-ministerial commitment at government level and strong NGOs.

Convenience and financial cost also came into the equation. At one stage the Institute of Education was exploring the involvement of five countries (almost a tenth of the Commonwealth membership) but lack of sufficient funding precluded this possibility. The then Overseas Development Administration in Britain and the Commonwealth Secretariat were only able to support work with developing countries, which meant that developed ones would have to pay for their own participation. There was a possible advantage in that two of the countries in the study are geographical neighbours.

In spite of the element of arbitrariness about the final selection, it was felt that inclusion of these four countries would give a sufficiently broad base to permit conclusions of wider significance. It was expected that other Commonwealth nations would wish to build on this work.

The international committee set up to oversee the project initially agreed to select a random sample of 100 students of 14 and 100 of 16 in five schools. An equal number of males and females were to be chosen. Because of the Standards system of forms or classes, in use in Botswana and Zimbabwe, this age difference could not be exact. But in principle the groups were separated by two years of schooling whose impact could be assessed. The students were told that their questionnaire responses would be kept confidential.

The choice of schools was left to the project team in each country, but the aim was to have a rough cross-section of institutions which would have differing characteristics and include main school types. Factors which might be aimed at included rural v urban, mixed v single sex, high status v lower prestige. If all the sample students had been drawn from the same type of school it might have given a misleading national picture, and obscured within-country differences.

The research phase took place in 1996, following trialling in Zimbabwe, Botswana and Britain. Schools were chosen as follows:

Botswana: Bokamoso Community Junior School in Gaborone, the capital, represented government urban schools. Legae Academy in Gaborone represented the exclusive private schools which charge for tuition and are often quite expensive. Capital Continuation Classes in Gaborone is a less privileged private night school catering in this study for students aged 17 to 37. Mosetlha Community Junior School is a government school in Bobonong, a rural village in central Botswana over 400km from the capital and 90km from the nearest town, with little influence from national media. Shakawe Community Junior School is a government school in a remote area, over 2000km from the capital. In Botswana the student questionnaires were delivered to classes of 40-43 students to fill out and then randomised in the office to produce ten male and ten female responses per class. Hence there were exactly 100 male and 100 female respondents.

India: Due to the huge population and expanse of India it was decided to select eight schools, four in rural and four in urban areas in four different states, Karnataka, Orissa, Rajasthan and Madhya Pradesh. All schools are affiliated to the Central Board of Secondary Education, and are coeducational, and no private schools are included in the sample. All follow broadly the same courses of study at Upper Primary (classes VI-VIII) and Secondary (classes IX-X) stages, are subject to the same public exams at the end of the Secondary stage, and both English and Hindi are in use as languages for study as well as media of instruction. The total sample numbered 312 (151 girls and 161 boys). The schools were: D.M. School, Mysore, Karnataka (urban); Jawahar Navodaya Vidyalaya, Shivaragudda, Mandya District, Karnataka (rural); D.M. School, Bhubaneswar, Orissa (urban); Jawahar Navodaya Vidyalaya, Munduli, District Cuttack, Orissa (rural); D.M. School, Ajmer, Rajasthan (urban); Jawahar Navodaya Vidyalaya, Nandla, District Ajmer, Rajasthan (rural); D.M. School, Bhopal, Madhya Pradesh (urban); Jawahar Navodaya Vidyalaya, Pawarkheda, District Hoshangabad, Madhya Pradesh (rural).

Northern Ireland: Due to the sensitive situation in Northern Ireland, at a time when the then cease-fire was breaking down, the research team promised anonymity to the five schools participating in the project. School 1 is a controlled (Protestant) high school with around 450 male pupils, serving a deprived working class urban area with strong loyalist/Unionist/Protestant affiliations. School 2 is an elite coeducational grammar school (Protestant) for 1208 pupils in a large town with a rural catchment. School 3 is a coeducational Integrated school (ie serving both Protestant and Catholic communities) in a disadvantaged urban area, which is regarded as innovative and effective. School 4 is a coeducational (Catholic) secondary established by a merger of two schools three years ago, where almost half of students are eligible for free school meals, located in a town but with a large rural catchment. School 5 is an urban maintained (Catholic) girls' secondary school of 933 pupils in a strongly Nationalist/Republican area, which has been considered highly innovative and won many awards. The total sample involved 108 14 year olds and 106 16 year olds, of whom 108 were male and 106 female.

Zimbabwe: The research team selected five schools of differing characteristics in the Harare Region (Educational Province); due to local circumstances they obtained 97 responses from the Form II students (the younger group), of whom 54 were males and 43 females, and 92 from the Form IV's (the older group), of whom 50 were males and 42 females. The schools were: Prince Edward, a government-owned boys' boarding school set up in the colonial era for white pupils, now a well-resourced multiracial school recruiting from well-to-do families; Girls High, a government-owned girls' boarding school, similar in history and present circumstances to Prince Edward; Zengeza 1 High, a coeducational day school set up just prior to independence in Chitungwiza, a not so affluent dormitory for Harare, now with large student numbers (2101) and short of classrooms and teaching materials; Domboramwari Secondary, a church-owned coeducational day school set up in the 1990s in a shanty, pert-urban suburb, serving low income families and poorly endowed with resources; and Mufakose 2 High, a government coeducational day school of 2220, opened after independence in a high density suburb, serving low income families and short of resources.