‡7RJLYHRUGLQDU\=LPEDEZHDQVDQRSSRUWXQLW\WRSDUWLFLSDWHLQWKHWUDQVLWLRQDOMXVWLFH process. ‡7RDVVHVVWKHRYHUDOOH[SRVXUHRIRUGLQDU\=LPEDEZHDQVWRSROLWLFDOYLROHQFH ‡7RXQGHUVWDQGWKHSULRULWLHVDQGQHHGVRIFLYLOLDQVDIIHFWHGE\SROLWLFDOFRQIOLFWLQWKH context of transitional justice. ‡7RFDSWXUHWKHYDULRXVUHFRPPHQGDWLRQVDERXWSHDFHDQGVRFLDOUHFRQVWUXFWLRQLQWKH country. ‡7RGRFXPHQWWKHSHRSOH¶VDWWLWXGHVDQGRSLQLRQVDERXWWUDQVLWLRQDOMXVWLFHPHFKDQLVPV This survey captured the people’s perceptions in all ten provinces, and opinions in both urban and rural areas of Zimbabwe were assessed. There are difficulties and inherent limitations in conducting a population-based opinion survey at this juncture in Zimbabwe. The Forum pursued the survey, however, in order to lobby and encourage the Zimbabwean government and international actors to develop transitional justice mechanisms which are supported by the people. How the research was conducted Research design The research method was based on a pre-determined questionnaire, which was administered to individual interviewees at household level. This method allowed the opinions of participants from a broad spectrum of the population to be elicited. The goal was to probe the opinions and attitudes of the general population about past gross human rights violations and possible responses to this history of abuse. Local staff with appropriate language and fieldwork skills were recruited and trained to assist in conducting the research. While the initial research target was 3 200, in the end 3 189 respondents were interviewed. Eighty-four (84) constituencies were sampled randomly from the ten administrative provinces of Zimbabwe using the Probability Proportionate to Population Size (PPPS) approach. This technique ensures that each area is fairly represented according to its actual population. In other words, the Forum conducted more interviews in areas with larger populations than in those with smaller populations. Within the 84 randomly selected constituencies, two approaches were used for the next phase of sampling. In rural areas, one village was selected randomly within the constituency, and a centrally located notable feature (e.g. a school or store) was chosen as the starting point; in urban areas, a centrally located crossroads was selected as the starting point. In both rural and urban locations, researchers were deployed along compass lines and sampled every fifth household until the assigned number of households had been interviewed. To allow an appropriate gender and age balance of respondents, one adult from each household was chosen randomly to participate in the survey as follows: the researcher would assign an alias to each adult present in the household and put slips of paper with those aliases into a hat. The researcher would then select one slip of paper from the hat to determine who would respond to the survey. Transitional Justice Survey – A Report of the People’s Perceptions and Recommendations 9