World Values Survey 2007, Wave 5

Four full days, in addition to conducting of pilot interviews, were spent training the interviewers. Training occurred in English and Kinyarwanda, as such ensuring all interviewers understood the specific intricacies of each question, even those who had only functional English. Despite this, a large amount of time was still spent defining words and explaining the meanings of questions. The interviewer training consisted of basic research background and objectives, questionnaire training and general interviewer skills, to assist interviewers to approach respondents in a manner that would reduce refusals as well as gain the respondents trust to facilitate the garnering of more personal answers. Interviewers were also trained in sampling procedures including understanding how to read the maps illustrating the suburb and street in each selected enumeration area, the Random Walk procedure for dwelling selection, the left hand rule, using the Kish Grid for respondent selection and how to proceed when an identified dwelling is a block of flats, informal settlement or rural settlement without formal street names or household numbers. The training was conducted by both Tracy Hammond from Markinor, as well as Kenn Dirangu from DMC Research and Moses Odhiambo from The Steadman Group. The training was attended by all interviewers who would be working on the job. A team was trained and this team traveled around Rwanda conducting all interviews required to meet the sample requirements. Each attending interviewer and field supervisor conducted one pilot interview implementing all taught sampling procedures and questionnaire instructions. This allowed insight into areas interviewers and respondents were having difficulty with, as well as ensuring all interviewer instructions on the questionnaire were followed ( such as the One mention only instruction etc). To ensure this was achieved each pilot interview was attended by either Tracy Hammond or an executive from The Steadman Group. Interviewers implemented the questionnaire exceptionally well, with hardly any errors being made. Of the few errors that were made, these were immediately corrected. The same was true of the sampling procedures. Unlike the other African countries where pilots were usually conducted in the front of respondents houses from where they ran some sort of informal business or another, in Rwanda respondents in urban areas tended to herd the interviewer inside their house, so as to keep confidential the reason for the visit. In contrast in the rural areas respondents specifically placed interviewers outside of their houses, so as to ensure all knew what was occurring and were satisfied that nothing wrong or secretive was occurring . The average interview length was 68 minutes BUT pilot interviews conducted with illiterate respondents took significantly longer (average length: 112 minutes). The major reason for this was that illiterate respondents were unable to refer to the show cards while considering their answers they had to be reminded of the response options available. V12 - V21 Respondents had difficulty understanding that they could choose UP TO 5 responses for this question, attempting to force 5 responses OR rather only provide one. Extra time thus needed to be spent on this question ensuring the respondent understood the answer options available to them. V24 V33 Respondents had difficulty understanding the word organisation even when translated, as this was viewed as being a corporate company aimed at generating profit i.e. a business. Interviewers specifically had to explain that we were referring to groups and associations as well as their traditional understanding of organisations. V24 Respondents had differing interpretations of what religious organisation meant. Some believed this referred to the priest only while others believed it referred to the community that attended this church as well. Interviewers had to explain that we were referring all these groups of people as an inclusive entity. V163 V164 Respondents were in general afraid to answer these questions, in case there were repercussions as a result of their particular viewpoint. V165 Respondents although they knew what the specific goals were, did not have an understanding of the term Millennium Development Goals V46 V47; V68;V114; V116 V123; V162 V163; V178; V192;V221; V244 V246; V253 Respondents had GREAT difficulty interpreting scales with opposing statements on either side of a 10 point scale. They tended to give an answer of agreement or not for either statements separately rather than selecting a number to indicate their answer on the continuum between the two statements. A large amount of time had to be spent in each interview explaining (over and over again!) that a score below 5 indicated agreement in varying degrees of strength with the statement on the left, 5 meant a lack of agreement or neutral feeling towards both statements, and a score between 6 and 10 indicated varying degrees of agreement with the statement on the right. Attempts at utilising the counting stones scale assistance technique failed as respondents were too confused by the fact that there were two statements involved in each question. V231 V233 A secret ballot was utilised to collect data relating to respondents political party choices The Rwandan Government would only allow this questionnaire to be administered if the following questions were removed: V132, V138, V147B, V148 V151, V156, V161, V163, V164, V221, V231 V234.